Image processing

PDF generated using the open source mwlib toolkit. See http://code.pediapress.com/ for more information.
PDF generated at: Thu, 20 May 2010 09:07:15 UTC
Contents
Articles
Image processing 1
Image processing 1
Digital image processing 3
Digital imaging 5
Medical imaging 6

Digital images 14
Quantization (signal processing) 14
Brightness 16
Luminance 17
Contrast (vision) 19
Color space 23
Color mapping 27
Color management 28
Digital Imaging and Communications in Medicine 32
JPEG 2000 40

Operations on images 53
Linear filter 53
Histogram 57
Image histogram 62
Color histogram 64
Affine transformation 66
Scaling (geometry) 70
Rotation (mathematics) 72
Color balance 77
Image registration 82
Segmentation (image processing) 85

References
Article Sources and Contributors 93
Image Sources, Licenses and Contributors 95

Article Licenses
License 96
 1

Image processing

Image processing
In electrical engineering and computer
science, image processing is any form of
signal processing for which the input is an
image, such as photographs or frames of
video; the output of image processing can be
either an image or a set of characteristics or
parameters related to the image. Most
image-processing techniques involve
treating the image as a two-dimensional
signal and applying standard
signal-processing techniques to it.

Image processing usually refers to digital

image processing, but optical and analog
image processing are also possible. This
Monochrome black/white image
article is about general techniques that apply
to all of them. The acquisition of images
(producing the input image in the first place) is referred to as imaging.

Typical operations
Among many other image processing
operations are:
Euclidean geometry transformations such
as enlargement, reduction, and rotation
Color corrections such as brightness and
contrast adjustments, color mapping, The red, green, and blue color channels of a photograph by Sergei Mikhailovich
Prokudin-Gorskii. The fourth image is a composite.
color balancing, quantization, or color
translation to a different color space
Digital compositing or optical compositing (combination of two or more images). Used in film-making to make a
"matte"
Interpolation, demosaicing, and recovery of a full image from a raw image format using a Bayer filter pattern
Image registration, the alignment of two or more images
Image differencing and morphing
Image recognition, for example, extract the text from the image using optical character recognition or checkbox
and bubble values using optical mark recognition
Image segmentation
High dynamic range imaging by combining multiple images
Geometric hashing for 2-D object recognition with affine invariance
Image processing 2

Applications
Computer vision
Optical sorting
Augmented Reality
Face detection
Feature detection
Lane departure warning system
Non-photorealistic rendering
Medical image processing
Microscope image processing
Morphological image processing
Remote sensing

See also
Imaging
Photo manipulation
List of image analysis software
Near sets
Multidimensional systems

Further reading
Tinku Acharya and Ajoy K. Ray. Image Processing - Principles and Applications [1]. Wiley InterScience.
Wilhelm Burger and Mark J. Burge (2007). Digital Image Processing: An Algorithmic Approach Using Java [2].
Springer. ISBN1846283795 and ISBN 3540309403.
R. Fisher, K Dawson-Howe, A. Fitzgibbon, C. Robertson, E. Trucco (2005). Dictionary of Computer Vision and
Image Processing. John Wiley. ISBN0-470-01526-8.
Bernd Jhne (2002). Digital Image Processing. Springer. ISBN3-540-67754-2.
Tim Morris (2004). Computer Vision and Image Processing. Palgrave Macmillan. ISBN0-333-99451-5.
Milan Sonka, Vaclav Hlavac and Roger Boyle (1999). Image Processing, Analysis, and Machine Vision. PWS
Publishing. ISBN0-534-95393-X.

External links
Lectures on Image Processing [3], by Alan Peters. Vanderbilt University. Updated 28 April 2008.
EURASIP Journal on Image and Video Processing [4] Open Access journal on Image Processing
Image processing algorithms, implementations and demonstrations [5]

References
[1] http:/ / books. google. co. in/ books?id=smBw4-xvfrIC& lpg=PP1& ots=FVFYaOATEF& dq=image%20processing%20ajoy%20ray&
pg=PP1#v=onepage& q=& f=false
[2] http:/ / www. imagingbook. com/
[3] http:/ / www. archive. org/ details/ Lectures_on_Image_Processing
[4] http:/ / www. hindawi. com/ journals/ ivp/
[5] http:/ / www. ipol. im/
Digital image processing 3

Digital image processing

Digital image processing is the use of computer algorithms to perform image processing on digital images. As a
subfield of digital signal processing, digital image processing has many advantages over analog image processing; it
allows a much wider range of algorithms to be applied to the input data, and can avoid problems such as the build-up
of noise and signal distortion during processing. Since images are defined over two dimensions (perhaps more)
digital image processing can be modeled in the form of Multidimensional Systems.

History
Many of the techniques of digital image processing, or digital picture processing as it was often called, were
developed in the 1960s at the Jet Propulsion Laboratory, MIT, Bell Labs, University of Maryland, and a few other
places, with application to satellite imagery, wirephoto standards conversion, medical imaging, videophone,
character recognition, and photo enhancement.[1] But the cost of processing was fairly high with the computing
equipment of that era. In the 1970s, digital image processing proliferated, when cheaper computers and dedicated
hardware became available. Images could then be processed in real time, for some dedicated problems such as
television standards conversion. As general-purpose computers became faster, they started to take over the role of
dedicated hardware for all but the most specialized and compute-intensive operations.
With the fast computers and signal processors available in the 2000s, digital image processing has become the most
common form of image processing, and is generally used because it is not only the most versatile method, but also
the cheapest.
Digital image processing technology for medical applications was inducted into the Space Foundation Space
Technology Hall of Fame in 1994.[2]

Tasks
Digital image processing allows the use of much more complex algorithms for image processing, and hence can
offer both more sophisticated performance at simple tasks, and the implementation of methods which would be
impossible by analog means.
In particular, digital image processing is the only practical technology for:
Classification
Feature extraction
Pattern recognition
Projection
Multi-scale signal analysis
Some techniques which are used in digital image processing include:
Pixelization
Linear filtering
Principal components analysis
Independent component analysis
Hidden Markov models
Partial differential equations
Self-organizing maps
Neural networks
Wavelets
Digital image processing 4

Applications

Digital camera images

Digital cameras generally include dedicated digital image processing chips to convert the raw data from the image
sensor into a color-corrected image in a standard image file format. Images from digital cameras often receive further
processing to improve their quality, a distinct advantage digital cameras have over film cameras. The digital image
processing is typically done by special software programs that can manipulate the images in many ways.
Many digital cameras also enable viewing of histograms of images, as an aid for the photographer to better
understand the rendered brightness range of each shot.

Film
Westworld (1973) was the first feature film to use digital image processing to pixellate photography to simulate an
android's point of view.[3]

See also
Computer graphics
Computer vision
Digitizing
Endrov
GPGPU
ImageJ
FIJI (software)
Homomorphic filtering
OpenCV
Standard test image
Super-resolution
Multidimensional systems

Further reading
Wilhelm Burger and Mark J. Burge (2007). Digital Image Processing: An Algorithmic Approach Using Java [2].
Springer. ISBN1846283795 and ISBN 3540309403.
R. Fisher, K Dawson-Howe, A. Fitzgibbon, C. Robertson, E. Trucco (2005). Dictionary of Computer Vision and
Image Processing. John Wiley. ISBN0-470-01526-8.
Bernd Jhne (2002). Digital Image Processing. Springer. ISBN3-540-67754-2.
Tim Morris (2004). Computer Vision and Image Processing. Palgrave Macmillan. ISBN0-333-99451-5.
Milan Sonka, Vaclav Hlavac and Roger Boyle (1999). Image Processing, Analysis, and Machine Vision. PWS
Publishing. ISBN0-534-95393-X.
Digital image processing 5

External links
Tutorial for image processing [4] (contains a Java applet)
Image processing algorithms, implementations and demonstrations [5]

References
[1] Azriel Rosenfeld, Picture Processing by Computer, New York: Academic Press, 1969
[2] "Space Technology Hall of Fame:Inducted Technologies/1994" (http:/ / www. spacetechhalloffame. org/
inductees_1994_Digital_Image_Processing. html). Space Foundation. 1994. . Retrieved 7 January 2010.
[3] A Brief, Early History of Computer Graphics in Film (http:/ / www. beanblossom. in. us/ larryy/ cgi. html), Larry Yaeger, 16 Aug 2002 (last
update), retrieved 24 March 2010
[4] http:/ / www. ph. tn. tudelft. nl/ Courses/ FIP/ frames/ fip. html

Digital imaging
Digital imaging or digital image acquisition is the creation of digital images, typically from a physical scene. The
term is often assumed to imply or include the processing, compression, storage, printing, and display of such images.

History
Digital imaging was developed in the 1960s and 1970s, largely to avoid the operational weaknesses of film cameras,
for scientific and military missions including the KH-11 program. As digital technology became cheaper in later
decades it replaced the old film methods for many purposes.

Descriptions
A digital image may be created directly from a physical scene by a camera or similar devices. Alternatively, it may
be obtained from another image in an analog medium, such as photographs, photographic film, or printed paper, by
an image scanner or similar device. Many technical imagessuch as those acquired with tomographic equipment,
side-scan sonar, or radio telescopesare actually obtained by complex processing of non-image data. This
digitalization of analog real-world data is known as digitizing, and involves sampling (discretization) and
quantization.
Finally, a digital image can also be computed from a geometric model or mathematical formula. In this case the
name image synthesis is more appropriate, and it is more often known as rendering.
Digital image authentication is an emerging issue for the providers and producers of high resolution digital images
such as health care organizations, law enforcement agencies and insurance companies. There are methods emerging
in forensic science to analyze a digital image and determine if it has been altered.
Digital imaging 6

See also
Digital image processing
Digital photography
Dynamic imaging
Image editing
Image retrieval
Graphics file format
Graphic image development
Society for Imaging Science and Technology, (IS&T)
Film recorder

External links
Cornell University. Digital imaging tutorial [1]
Digital Imaging Definitions/Glossary/Terms. Digital Imaging Definitions [2]
Dartmouth, Hany Farid. Digital Image Forensics [3]

References
[1] http:/ / www. library. cornell. edu/ preservation/ tutorial/ contents. html
[2] http:/ / www. msimaging. com/ definitions. asp
[3] http:/ / www. cs. dartmouth. edu/ farid/ publications/ sciam08. html

Medical imaging
Medical imaging is the technique and process used to create images of the human body (or parts and function
thereof) for clinical purposes (medical procedures seeking to reveal, diagnose or examine disease) or medical science
(including the study of normal anatomy and physiology). Although imaging of removed organs and tissues can be
performed for medical reasons, such procedures are not usually referred to as medical imaging, but rather are a part
of pathology.
As a discipline and in its widest sense, it is part of biological imaging and incorporates radiology (in the wider
sense), nuclear medicine, investigative radiological sciences, endoscopy, (medical) thermography, medical
photography and microscopy (e.g. for human pathological investigations).
Measurement and recording techniques which are not primarily designed to produce images, such as
electroencephalography (EEG), magnetoencephalography (MEG), Electrocardiography (EKG) and others, but which
produce data susceptible to be represented as maps (i.e. containing positional information), can be seen as forms of
medical imaging.

Overview
In the clinical context, medical imaging is generally equated to radiology or "clinical imaging" and the medical
practitioner responsible for interpreting (and sometimes acquiring) the images is a radiologist. Diagnostic
radiography designates the technical aspects of medical imaging and in particular the acquisition of medical images.
The radiographer or radiologic technologist is usually responsible for acquiring medical images of diagnostic
quality, although some radiological interventions are performed by radiologists. While radiology is an evaluation of
anatomy, nuclear medicine provides functional assessment.
Medical imaging 7

As a field of scientific investigation, medical imaging constitutes a sub-discipline of biomedical engineering,

medical physics or medicine depending on the context: Research and development in the area of instrumentation,
image acquisition (e.g. radiography), modelling and quantification are usually the preserve of biomedical
engineering, medical physics and computer science; Research into the application and interpretation of medical
images is usually the preserve of radiology and the medical sub-discipline relevant to medical condition or area of
medical science (neuroscience, cardiology, psychiatry, psychology, etc) under investigation. Many of the techniques
developed for medical imaging also have scientific and industrial applications.
Medical imaging is often perceived to designate the set of techniques that noninvasively produce images of the
internal aspect of the body. In this restricted sense, medical imaging can be seen as the solution of mathematical
inverse problems. This means that cause (the properties of living tissue) is inferred from effect (the observed signal).
In the case of ultrasonography the probe consists of ultrasonic pressure waves and echoes inside the tissue show the
internal structure. In the case of projection radiography, the probe is X-ray radiation which is absorbed at different
rates in different tissue types such as bone, muscle and fat.

Imaging technology

Radiography
Two forms of radiographic images are in use in medical imaging; projection radiography and fluoroscopy, with the
latter being useful for intraoperative and catheter guidance. These 2D techniques are still in wide use despite the
advance of 3D tomography due to the low cost, high resolution, and depending on application, lower radiation
dosages. This imaging modality utilizes a wide beam of x rays for image acquisition and is the first imaging
technique available in modern medicine.
Fluoroscopy produces real-time images of internal structures of the body in a similar fashion to radiography, but
employs a constant input of x-rays, at a lower dose rate. Contrast media, such as barium, iodine, and air are used
to visualize internal organs as they work. Fluoroscopy is also used in image-guided procedures when constant
feedback during a procedure is required. An image receptor is required to convert the radiation into an image after
it has passed through the area of interest. Early on this was a fluorescing screen, which gave way to an Image
Amplifier (IA) which was a large vacuum tube that had the receiving end coated with cesium iodide, and a mirror
at the opposite end. Eventually the mirror was replaced with a TV camera.
Projectional radiographs, more commonly known as x-rays, are often used to determine the type and extent of a
fracture as well as for detecting pathological changes in the lungs. With the use of radio-opaque contrast media,
such as barium, they can also be used to visualize the structure of the stomach and intestines - this can help
diagnose ulcers or certain types of colon cancer.
Medical imaging 8

Magnetic resonance imaging (MRI)

A magnetic resonance imaging instrument (MRI scanner), or "nuclear
magnetic resonance (NMR) imaging" scanner as it was originally
known, uses powerful magnets to polarise and excite hydrogen nuclei
(single proton) in water molecules in human tissue, producing a
detectable signal which is spatially encoded, resulting in images of the
body. MRI uses three electromagnetic fields: a very strong (on the
order of units of teslas) static magnetic field to polarize the hydrogen
nuclei, called the static field; a weaker time-varying (on the order of
1kHz) field(s) for spatial encoding, called the gradient field(s); and a
weak radio-frequency (RF) field for manipulation of the hydrogen
nuclei to produce measurable signals, collected through an RF antenna.

Like CT, MRI traditionally creates a two dimensional image of a thin

"slice" of the body and is therefore considered a tomographic imaging
technique. Modern MRI instruments are capable of producing images
in the form of 3D blocks, which may be considered a generalisation of
A brain MRI representation
the single-slice, tomographic, concept. Unlike CT, MRI does not
involve the use of ionizing radiation and is therefore not associated
with the same health hazards. For example, because MRI has only been in use since the early 1980s, there are no
known long-term effects of exposure to strong static fields (this is the subject of some debate; see 'Safety' in MRI)
and therefore there is no limit to the number of scans to which an individual can be subjected, in contrast with X-ray
and CT. However, there are well-identified health risks associated with tissue heating from exposure to the RF field
and the presence of implanted devices in the body, such as pace makers. These risks are strictly controlled as part of
the design of the instrument and the scanning protocols used.

Because CT and MRI are sensitive to different tissue properties, the appearance of the images obtained with the two
techniques differ markedly. In CT, X-rays must be blocked by some form of dense tissue to create an image, so the
image quality when looking at soft tissues will be poor. In MRI, while any nucleus with a net nuclear spin can be
used, the proton of the hydrogen atom remains the most widely used, especially in the clinical setting, because it is
so ubiquitous and returns a large signal. This nucleus, present in water molecules, allows the excellent soft-tissue
contrast achievable with MRI.

Nuclear medicine
Nuclear medicine encompasses both diagnostic imaging and treatment of disease, and may also be referred to as
molecular medicine or molecular imaging & therapeutics [1] . Nuclear medicine uses certain properties of isotopes
and the energetic particles emitted from radioactive material to diagnose or treat various pathology. Different from
the typical concept of anatomic radiology, nuclear medicine enables assessment of physiology. This function-based
approach to medical evaluation has useful applications in most subspecialties, notably oncology, neurology, and
cardiology. Gamma cameras are used in e.g. scintigraphy, SPECT and PET to detect regions of biologic activity that
may be associated with disease. Relatively short lived isotope, such as 123I is administered to the patient. Isotopes
are often preferentially absorbed by biologically active tissue in the body, and can be used to identify tumors or
fracture points in bone. Images are acquired after collimated photons are detected by a crystal that gives off a light
signal, which is in turn amplified and converted into count data.
Scintigraphy ("scint") is a form of diagnostic test wherein radioisotopes are taken internally, for example
intravenously or orally. Then, gamma camera capture and form two-dimensional[2] images from the radiation
emitted by the radiopharmaceuticals.
Medical imaging 9

SPECT is a 3D tomographic technique that uses gamma camera data from many projections and can be
reconstructed in different planes. A dual detector head gamma camera combined with a CT scanner, which
provides localization of functional SPECT data, is termed a SPECT/CT camera, and has shown utility in
advancing the field of molecular imaging.
Positron emission tomography (PET) uses coincidence detection to image functional processes. Short-lived
positron emitting isotope, such as 18F, is incorporated with an organic substance such as glucose, creating
F18-fluorodeoxyglucose, which can be used as a marker of metabolic utilization. Images of activity distribution
throughout the body can show rapidly growing tissue, like tumor, metastasis, or infection. PET images can be
viewed in comparison to computed tomography scans to determine an anatomic correlate. Modern scanners
combine PET with a CT, or even MRI, to optimize the image reconstruction involved with positron imaging. This
is performed on the same equipment without physically moving the patient off of the gantry. The resultant hybrid
of functional and anatomic imaging information is a useful tool in non-invasive diagnosis and patient
management.

Photoacoustic imaging
Photoacoustic imaging is a recently developed hybrid biomedical imaging modality based on the photoacoustic
effect. It combines the advantages of optical absorption contrast with ultrasonic spatial resolution for deep imaging
in (optical) diffusive or quasi-diffusive regime. Recent studies have shown that photoacoustic imaging can be used in
vivo for tumor angiogenesis monitoring, blood oxygenation mapping, functional brain imaging, and skin melanoma
detection, etc.

Breast Thermography
Digital infrared imaging thermography is based on the principle that metabolic activity and vascular circulation in
both pre-cancerous tissue and the area surrounding a developing breast cancer is almost always higher than in normal
breast tissue. Cancerous tumors require an ever-increasing supply of nutrients and therefore increase circulation to
their cells by holding open existing blood vessels, opening dormant vessels, and creating new ones
(neoangiogenesis). This process frequently results in an increase in regional surface temperatures of the breast.
Digital infrared imaging uses extremely sensitive medical infrared cameras and sophisticated computers to detect,
analyze, and produce high-resolution diagnostic images of these temperature variations. Because of DII's sensitivity,
these temperature variations may be among the earliest signs of breast cancer and/or a pre-cancerous state of the
breast[3] .

Tomography
Tomography is the method of imaging a single plane, or slice, of an object resulting in a tomogram. There are
several forms of tomography:
Linear tomography: This is the most basic form of tomography. The X-ray tube moved from point "A" to point
"B" above the patient, while the cassette holder (or "bucky") moves simultaneously under the patient from point
"B" to point "A." The fulcrum, or pivot point, is set to the area of interest. In this manner, the points above and
below the focal plane are blurred out, just as the background is blurred when panning a camera during exposure.
No longer carried out and replaced by computed tomography.
Poly tomography: This was a complex form of tomography. With this technique, a number of geometrical
movements were programmed, such as hypocycloidic, circular, figure 8, and elliptical. Philips Medical Systems
[4] produced one such device called the 'Polytome.' This unit was still in use into the 1990s, as its resulting
images for small or difficult physiology, such as the inner ear, was still difficult to image with CTs at that time.
As the resolution of CTs got better, this procedure was taken over by the CT.
Medical imaging 10

Zonography: This is a variant of linear tomography, where a limited arc of movement is used. It is still used in
some centres for visualising the kidney during an intravenous urogram (IVU).
Orthopantomography (OPT or OPG): The only common tomographic examination in use. This makes use of a
complex movement to allow the radiographic examination of the mandible, as if it were a flat bone. It is often
referred to as a "Panorex", but this is incorrect, as it is a trademark of a specific company.
Computed Tomography (CT), or Computed Axial Tomography (CAT: A CT scan, also known as a CAT scan, is
a helical tomography (latest generation), which traditionally produces a 2D image of the structures in a thin
section of the body. It uses X-rays. It has a greater ionizing radiation dose burden than projection radiography;
repeated scans must be limited to avoid health effects.

Ultrasound
Medical ultrasonography uses high frequency broadband sound waves in the megahertz range that are reflected by
tissue to varying degrees to produce (up to 3D) images. This is commonly associated with imaging the fetus in
pregnant women. Uses of ultrasound are much broader, however. Other important uses include imaging the
abdominal organs, heart, breast, muscles, tendons, arteries and veins. While it may provide less anatomical detail
than techniques such as CT or MRI, it has several advantages which make it ideal in numerous situations, in
particular that it studies the function of moving structures in real-time, emits no ionizing radiation, and contains
speckle that can be used in elastography. It is very safe to use and does not appear to cause any adverse effects,
although information on this is not well documented. It is also relatively inexpensive and quick to perform.
Ultrasound scanners can be taken to critically ill patients in intensive care units, avoiding the danger caused while
moving the patient to the radiology department. The real time moving image obtained can be used to guide drainage
and biopsy procedures. Doppler capabilities on modern scanners allow the blood flow in arteries and veins to be
assessed.

Medical imaging topics

Maximizing imaging procedure use

The amount of data obtained in a single MR or CT scan is very extensive. Some of the data that radiologists discard
could save patients time and money, while reducing their exposure to radiation and risk of complications from
invasive procedures.[5]

Creation of three-dimensional images

Recently, techniques have been developed to enable CT, MRI and ultrasound scanning software to produce 3D
images for the physician.[6] Traditionally CT and MRI scans produced 2D static output on film. To produce 3D
images, many scans are made, then combined by computers to produce a 3D model, which can then be manipulated
by the physician. 3D ultrasounds are produced using a somewhat similar technique. In diagnosing disease of the
viscera of abdomen,ultrasound is particularly sensitive on imaging of biliary tract,urinary tract and female
reproductive organs(ovary,fallopian tubes).As for example,diagnosis of gall stone by dilatation of common bile duct
and stone in common bile duct . With the ability to visualize important structures in great detail, 3D visualization
methods are a valuable resource for the diagnosis and surgical treatment of many pathologies. It was a key resource
for the famous, but ultimately unsuccessful attempt by Singaporean surgeons to separate Iranian twins Ladan and
Laleh Bijani in 2003. The 3D equipment was used previously for similar operations with great success.
Other proposed or developed techniques include:
Diffuse optical tomography
Elastography
Electrical impedance tomography
Medical imaging 11

Optoacoustic imaging
Ophthalmology
A-scan
B-scan
Corneal topography
Optical coherence tomography
Scanning laser ophthalmoscopy
Some of these techniques are still at a research stage and not yet used in clinical routines.

Compression of medical images

Medical imaging techniques produce very large amounts of data, especially from CT, MRI and PET modalities. As a
result, storage and communications of electronic image data are prohibitive without the use of compression. JPEG
2000 is the state-of-the-art image compression DICOM standard for storage and transmission of medical images.
The cost and feasibility of accessing large image data sets over low or various bandwidths are further addressed by
use of another DICOM standard, called JPIP, to enable efficient streaming of the JPEG 2000 compressed image data.

Non-diagnostic imaging
Neuroimaging has also been used in experimental circumstances to allow people (especially disabled persons) to
control outside devices, acting as a brain computer interface.

Archiving and recording

Used primarily in ultrasound imaging, capturing the image a medical imaging device is required for archiving and
telemedicine applications. In most scenarios, a frame grabber is used in order to capture the video signal from the
medical device and relay it to a computer for further processing and operations.[7]

Open source software for medical image analysis

Several open source software packages are available for performing analysis of medical images:
ImageJ
3D Slicer
ITK
OsiriX
GemIdent
MicroDicom
FreeSurfer
FreePlugin

Use in pharmaceutical clinical trials

Medical imaging has become a major tool in clinical trials since it enables rapid diagnosis with visualization and
quantitative assessment.
A typical clinical trial goes through multiple phases and can take up to eight years. Clinical endpoints or outcomes
are used to determine whether the therapy is safe and effective. Once a patient reaches the endpoint, he/she is
generally excluded from further experimental interaction. Trials that rely solely on clinical endpoints are very costly
as they have long durations and tend to need large number of patients.
In contrast to clinical endpoints, surrogate endpoints have been shown to cut down the time required to confirm
whether a drug has clinical benefits. Imaging biomarkers (a characteristic that is objectively measured by an imaging
Medical imaging 12

technique, which is used as an indicator of pharmacological response to a therapy) and surrogate endpoints have
shown to facilitate the use of small group sizes, obtaining quick results with good statistical power.[8]
Imaging is able to reveal subtle change that is indicative of the progression of therapy that may be missed out by
more subjective, traditional approaches. Statistical bias is reduced as the findings are evaluated without any direct
patient contact.
For example, measurement of tumour shrinkage is a commonly used surrogate endpoint in solid tumour response
evaluation. This allows for faster and more objective assessment of the effects of anticancer drugs. In evaluating the
extent of Alzheimers disease, it is still prevalent to use behavioural and cognitive tests. MRI scans on the entire
brain can accurately pinpoint hippocampal atrophy rate while PET scans is able to measure the brains metabolic
activity by measuring regional glucose metabolism.[8]
An imaging-based trial will usually be made up of three components:
1. A realistic imaging protocol. The protocol is an outline that standardizes (as far as practically possible) the way in
which the images are acquired using the various modalities (PET, SPECT, CT, MRI). It covers the specifics in
which images are to be stored, processed and evaluated.
2. An imaging centre that is responsible for collecting the images, perform quality control and provide tools for data
storage, distribution and analysis. It is important for images acquired at different time points are displayed in a
standardised format to maintain the reliability of the evaluation. Certain specialised imaging contract research
organizations provide to end medical imaging services, from protocol design and site management through to data
quality assurance and image analysis.
3. Clinical sites that recruit patients to generate the images to send back to the imaging centre.

See also
Preclinical imaging Magnetic field imaging Pneumoencephalogram
Cardiac PET Medical examination Radiology information system
Biomedical informatics Medical radiography Segmentation (image processing)
Digital Imaging and Communications in Medical test Signal-to-noise ratio
Medicine
Digital Mammography and PACS Neuroimaging Society for Imaging Science and Technology
EMMI European Master in Molecular Imaging Non-invasive (medical) Tomogram
Fotofinder PACS Virtopsy
Full-body scan JPEG 2000
compression
VoluMedic JPIP streaming

Further reading
Burger, Wilhelm; Burge, Mark James, eds (2008). Digital Image Processing: An Algorithmic Introduction using
Java. Texts in Computer Science series. New York: Springer Science+Business Media.
doi:10.1007/978-1-84628-968-2. ISBN978-1-84628-379-6.
Baert, Albert L., ed (2008). Encyclopedia of Diagnostic Imaging. Berlin: Springer-Verlag.
doi:10.1007/978-3-540-35280-8. ISBN978-3-540-35278-5.
Tony F. Chan and Jackie Shen (2005). Image Processing and Analysis - Variational, PDE, Wavelet, and
Stochastic Methods [9]. SIAM
Terry Yoo(Editor) (2004), Insight into Images.
Robb, RA (1999). Biomedical Imaging, Visualization, and Analysis. John Wiley & Sons, Inc. ISBN0471283533.
Journal of Digital Imaging (New York: Springer Science+Business Media). ISSN0897-1889.
Medical imaging 13

Using JPIP for Standard-Compliant Sharing of Medical Image Data [10] a white paper by Aware Inc. [11]

External links
Medical imaging [12] at the Open Directory Project
Medical Image Database [13] Free Indexed Online Images
http://www.aware.com/imaging/accuradjpip.htm What is JPIP?

References
[1] Society of Nuclear Medicine (http:/ / www. snm. org)
[2] thefreedictionary.com > scintigraphy (http:/ / medical-dictionary. thefreedictionary. com/ scintigraphy) Citing: Dorland's Medical Dictionary
for Health Consumers, 2007 by Saunders; Saunders Comprehensive Veterinary Dictionary, 3 ed. 2007; McGraw-Hill Concise Dictionary of
Modern Medicine, 2002 by The McGraw-Hill Companies
[3] http:/ / www. breastthermography. com/ breast_thermography_mf. htm
[4] http:/ / www. medical. philips. com/ main/ index. asp
[5] Freiherr G. Waste not, want not: Getting the most from imaging procedures (http:/ / www. diagnosticimaging. com/ news/ display/ article/
113619/ 1541872). Diagnostic Imaging. March 19, 2010.
[6] Udupa, J.K. and Herman, G. T., 3D Imaging in Medicine, 2nd Edition, CRC Press, 2000
[7] Treating Medical Ailments in Real Time (http:/ / www. epiphan. com/ solutions_new/ ?arid=16)
[8] Hajnal, J. V., Hawkes, D. J., & Hill, D. L. (2001). Medical Image Registration. CRC Press.
[9] http:/ / jackieneoshen. googlepages. com/ ImagingNewEra. html
[10] http:/ / www. aware. com/ imaging/ whitepapers/ wp_jpipwado. htm
[11] http:/ / www. aware. com/ imaging/ whitepapers. htm
[12] http:/ / www. dmoz. org/ Health/ Medicine/ Imaging/ /
[13] http:/ / rad. usuhs. edu/ medpix/ index. html?
 14

Digital images

Quantization (signal processing)

In digital signal processing, quantization is the process of
approximating ("mapping") a continuous range of values (or a very
large set of possible discrete values) by a relatively small ("finite") set
of ("values which can still take on continuous range") discrete symbols
or integer values. For example, rounding a real number in the interval
to an integer

In other words, quantization can be described as a mapping that

represents a finite continuous interval of the range of a Sampled signal (discrete signal): discrete time,
continuous values.
continuous valued signal, with a single number c, which is also on that
interval. For example, rounding to the nearest integer (rounding up)
replaces the interval with the number c, for integer c.
After that quantization we produce a finite set of values which can be
encoded by binary techniques for example.
In signal processing, quantization refers to approximating the output by
one of a discrete and finite set of values, while replacing the input by a
discrete set is called discretization, and is done by sampling: the
resulting sampled signal is called a discrete signal (discrete time), and
need not be quantized (it can have continuous values). To produce a Quantized signal: continuous time, discrete
digital signal (discrete time and discrete values), one both samples values.

(discrete time) and quantizes the resulting sample values (discrete

values).

Applications
In electronics, adaptive quantization is a quantization process that
varies the step size based on the changes of the input signal, as a means
of efficient compression. Two approaches commonly used are forward
adaptive quantization and backward adaptive quantization.
Digital signal (sampled, quantized): discrete time,
In signal processing the quantization process is the necessary and discrete values.
natural follower of the sampling operation. It is necessary because in
practice the digital computer with its general purpose CPU is used to implement DSP algorithms. And since
computers can only process finite word length (finite resolution/precision) quantities, any infinite precision
continuous valued signal should be quantized to fit a finite resolution, so that it can be represented (stored) in CPU
registers and memory.
We shall be aware of the fact that, it is not the continuous values of the analog function that inhibits its binary
encoding, rather it is the existence of infinitely many such values due to the definition of continuity,(which therefore
requires infinitely many bits to represent). For example we can design a quantizer such that it represents a signal
with a single bit (just two levels) such that, one level is "pi=3,14..." (say encoded with a 1) and the other level is
"e=2.7183..." ( say encoded with a 0), as we can see, the quantized values of the signal take on infinite precision,
Quantization (signal processing) 15

irrational numbers. But there are only two levels. And we can represent the output of the quantizer with a binary
symbol. Concluding from this we can see that it is not the discreteness of the quantized values that enable them to be
encoded but the finiteness enabling the encoding with finite number of bits.
In theory there is no relation between quantization values and binary code words used to encode them (rather than a
table that shows the corresponding mapping, just as examplified above). However due to practical reasons we may
tend to use code words such that their binary mathematical values has a relation with the quantization levels that is
encoded. And this last option merges the first two paragrahs in such a way that, if we wish to process the output of a
quantizer within a DSP/CPU system (which is always the case) then we can not allow the representation levels of the
quantizers to take on arbitrary values, but only a restricted range such that they can fit in computer registers.
A quantizer is identified with its number of levels M, the decision boundaries {di} and the corresponding
representation values {ri}.
The output of a quantizer has two important properties: 1) a Distortion resulting from the approximation and 2) a
Bit-Rate resulting from binary encoding of its levels. Therefore the Quantizer design problem is a Rate-Distortion
optimization type.
If we are only allowed to use fixed length code for the output level encoding (the practical case) then the problem
reduces into a distortion minimization one.
The design of a quantizer usually means the process to find the sets {di} and {ri} such that a measure of optimality is
satisfied (such as MMSEQ (Minimum Mean Squared Quantization Error))
Given the number of levels M, the optimal quantizer which minimizes the MSQE with regards to the given signal
statistics is called the Max-Lloyd quantizer, which is a non-uniform type in general.
The most common quantizer type is the uniform one. It is simple to design and implement and for most cases it
suffices to get satisfactory results. Indeed by the very inherent nature of the design process, a given quantizer will
only produce optimal results for the assumed signal statistics. Since it is very difficult to correctly predict that in
advance, any static design will never produce actual optimal performance whenever the input statistics deviates from
that of the design assumption. The only solution is to use an adaptive quantizer.

External links
Quantization threads in Comp.DSP [1]
Signal to quantization noise in quantized sinusoidal [2] - Analysis of quantization error on a sine wave

References
[1] http:/ / www. dsprelated. com/ comp. dsp/ keyword/ Quantization. php
[2] http:/ / www. dsplog. com/ 2007/ 03/ 19/ signal-to-quantization-noise-in-quantized-sinusoidal/
Brightness 16

Brightness
Brightness is an attribute of visual perception in which a source appears to be radiating or reflecting light.[1] In other
words, brightness is the perception elicited by the luminance of a visual target. This is a subjective attribute/property
of an object being observed.

Terminology
The adjective bright derives from an Old English beorht with the same meaning via metathesis giving Middle
English briht. The word is from a Common Germanic *berhtaz, ultimately from a PIE root with a closely related
meaning, *bhereg- "white, bright".
"Brightness" was formerly used as a synonym for the photometric term luminance and (incorrectly) for the
radiometric term radiance. As defined by the US Federal Glossary of Telecommunication Terms (FS-1037C),
"brightness" should now be used only for non-quantitative references to physiological sensations and perceptions of
light.[2]
A given target luminance can elicit different perceptions of brightness in different contexts; see, for example, White's
illusion and Wertheimer-Benary illusion.
In the RGB color space, brightness can be thought of as the arithmetic mean of the red, green, and blue color
coordinates (although some of the three components make the light seem brighter than others, which, again, may be
compensated by some display systems automatically):[3]

Brightness is also a color coordinate in the HSB or HSV color space (hue, saturation, and brightness or value).
With regard to stars, brightness is quantified as apparent magnitude and absolute magnitude

Brightness of sounds
The term "brightness" is also used in discussions of sound timbres, in a rough analogy with visual brightness. Timbre
researchers consider brightness to be one of the perceptually strongest distinctions between sounds[4] , and formalize
it acoustically as an indication of the amount of high-frequency content in a sound, using a measure such as the
spectral centroid.

See also
Luma (video)
Luminance (relative)
Luminosity
Brightness 17

External links
Poynton's Color FAQ [5]

References
[1] Merriam-Webster.com (http:/ / www. merriam-webster. com/ dictionary/ bright) Merriam-Webster Dictionary definition of bright
[2] Brightness (http:/ / www. its. bldrdoc. gov/ fs-1037/ dir-005/ _0719. htm) in Federal Standard 1037C, the Federal Glossary of
Telecommunication Terms (1996)
[3] What are HSB and HLS? (http:/ / www. poynton. com/ notes/ colour_and_gamma/ ColorFAQ. html#RTFToC36), Charles Poynton: "The
usual formulation of HSB and HLS compute so-called "lightness" or "brightness" as (R+G+B)/3. This computation conflicts badly with the
properties of colour vision, as it computes yellow to be about six times more intense than blue with the same "lightness" value (say L=50)."
[4] D. Wessel, Timbre space as a musical control structure (http:/ / mediatheque. ircam. fr/ articles/ textes/ Wessel78a/ ), Computer Music
Journal, 3 (1979), pp. 4552.
[5] http:/ / www. poynton. com/ ColorFAQ. html

Luminance
Luminance is a photometric measure of the luminous intensity per unit area of light travelling in a given direction. It
describes the amount of light that passes through or is emitted from a particular area, and falls within a given solid
angle. The SI unit for luminance is candela per square metre (cd/m2). A non-SI term for the same unit is the "nit".
The CGS unit of luminance is the stilb, which is equal to one candela per square centimetre or 10kcd/m2.
Luminance is often used to characterize emission or reflection from flat, diffuse surfaces. The luminance indicates
how much luminous power will be perceived by an eye looking at the surface from a particular angle of view.
Luminance is thus an indicator of how bright the surface will appear. In this case, the solid angle of interest is the
solid angle subtended by the eye's pupil. Luminance is used in the video industry to characterize the brightness of
displays. A typical computer display emits between 50 and 300cd/m2. The sun has luminance of about
1.6109cd/m2 at noon.[1]
Luminance is invariant in geometric optics. This means that for an ideal optical system, the luminance at the output
is the same as the input luminance. For real, passive, optical systems, the output luminance is at most equal to the
input. As an example, if you form a demagnified image with a lens, the luminous power is concentrated into a
smaller area, meaning that the illuminance is higher at the image. The light at the image plane, however, fills a larger
solid angle so the luminance comes out to be the same assuming there is no loss at the lens. The image can never be
"brighter" than the source.

Definition
Luminance is defined by

where
is the luminance (cd/m2),
is the luminous flux or luminous power (lm),
is the angle between the surface normal and the specified direction,
is the area of the surface (m2), and
is the solid angle (sr).
Luminance 18

See also
Diffuse reflection
Etendue
Exposure value
Illuminance
Lambert
Lambertian reflectance
Lightness, property of a color
Luma, the representation of luminance in a video monitor
Lumen (unit)
Radiance

External links
A Kodak guide to Estimating Luminance and Illuminance [2] using a camera's exposure meter. Also available in
PDF form [3] and Google Docs online version [4]

Quantity Symbol SI unit Abbr. Notes

Luminous energy Qv lumen second lms units are sometimes called talbots

Luminous flux F lumen (=cdsr) lm also called luminous power

Luminous intensity Iv candela (=lm/sr) cd an SI base unit

Luminance Lv candela per square cd/m2 units are sometimes called "nits"
metre

Illuminance Ev lx Used for light incident on a surface

lux (= lm/m2)

Luminous Mv lx Used for light emitted from a

lux (= lm/m2)
emittance surface

Luminous efficacy lumen per watt lm/W ratio of luminous flux to radiant flux

See also SI Photometry

References
[1] "Luminance" (http:/ / www. schorsch. com/ kbase/ glossary/ luminance. html). Lighting Design Glossary. . Retrieved Apr. 13, 2009.
[2] http:/ / www. kodak. com/ cluster/ global/ en/ consumer/ products/ techInfo/ am105/ am105kic. shtml
[3] http:/ / www. kodak. com/ cluster/ global/ en/ consumer/ products/ techInfo/ am105/ am105kic. pdf
[4] http:/ / docs. google. com/ gview?a=v& q=cache:3ygRnO1mEpEJ:www. kodak. com/ cluster/ global/ en/ consumer/ products/ techInfo/
am105/ am105kic. pdf+ estimating+ luminance+ and+ illuminance+ with+ reflection+ type& hl=en& gl=ie&
sig=AFQjCNGlsuOrMAs62ZwlzQilxa6mkx0QaA
Contrast (vision) 19

Contrast (vision)
Contrast is the difference in visual properties that makes an object
(or its representation in an image) distinguishable/darker or
brighter from other objects and the background. In visual
perception of the real world, contrast is determined by the
difference in the color and brightness of the object and other
objects within the same field of view. Because the human visual
system is more sensitive to contrast than absolute luminance, we
can perceive the world similarly regardless of the huge changes in
illumination over the day or from place to place.

The human contrast sensitivity function shows a typical band-pass Left side of the image has low contrast, the right has
shape peaking at around 4cycles per degree with sensitivity higher contrast.

dropping off either side of the peak.[1] This tells us that the human
visual system is most sensitive in detecting contrast differences
occurring at 4cycles per degree, i.e. at this spatial frequency
humans can detect lower contrast differences than at any other
spatial frequency.

The high-frequency cut-off represents the optical limitations of the

visual system's ability to resolve detail and is typically about
60cycles per degree. The high-frequency cut-off is related to the
packing density of the retinal photoreceptor cells: a finer matrix
can resolve finer gratings.
The low frequency drop-off is due to lateral inhibition within the
retinal ganglion cells. A typical retinal ganglion cell presents a
centre region with either excitation or inhibition and a surround
region with the opposite sign. By using coarse gratings, the bright Changes in the amount of contrast in a photo
bands fall on the inhibitory as well as the excitatory region of the
ganglion cell resulting in lateral inhibition and account for the low-frequency drop-off of the human contrast
sensitivity function.

One experimental phenomenon is the inhibition of blue in the periphery if blue light is displayed against white,
leading to a yellow surrounding. The yellow is derived from the inhibition of blue on the surroundings by the center.
Since white minus blue is red and green, this mixes to become yellow.[2]
For example, in the case of graphical computer displays, contrast depends on the properties of the picture source or
file and the properties of the computer display, including its variable settings. For some screens the angle between
the screen surface and the observer's line of sight is also important.
Contrast is also the difference between the color or shading of the printed material on a document and the
background on which it is printed, for example in optical character recognition.
Contrast (vision) 20

Formula
There are many possible definitions of
contrast. Some include color; others do not.
Travnikova laments, "Such a multiplicity of
notions of contrast is extremely
inconvenient. It complicates the solution of
many applied problems and makes it
difficult to compare the results published by
different authors."[3]

Various definitions of contrast are used in

different situations. Here, luminance
contrast is used as an example, but the
formulas can also be applied to other
physical quantities. In many cases, the An image of the Notre Dame cathedral as seen from the Eiffel Tower
definitions of contrast represent a ratio of
the type

The same image, with added global contrast, and local contrast (acutance)
increased through unsharp masking.

The rationale behind this is that a small difference is negligible if the average luminance is high, while the same
small difference matters if the average luminance is low (see WeberFechner law). Below, some common
definitions are given.
Contrast (vision) 21

Weber contrast
The Weber contrast is defined as

with and representing the luminance of the features and the background luminance, respectively. It is
commonly used in cases where small features are present on a large uniform background, i.e. the average luminance
is approximately equal to the background luminance.

Michelson contrast
The Michelson contrast[4] is commonly used for patterns where both bright and dark features are equivalent and take
up similar fractions of the area. The Michelson contrast is defined as

with and representing the highest and lowest luminance. The denominator represents twice the average
of the luminance.

RMS contrast
Root mean square (RMS) contrast does not depend on the spatial frequency content or the spatial distribution of
contrast in the image. RMS contrast is defined as the standard deviation of the pixel intensities:[5]

where intensities are the -th -th element of the two dimensional image of size by . is the average
intensity of all pixel values in the image. The image is assumed to have its pixel intensities normalized in the
range .
Contrast (vision) 22

Contrast sensitivity
Contrast sensitivity is a measure of the ability to discern between luminances of different levels in a static image.
Contrast sensitivity varies between individuals, reaching a maximum at approximately 20 years of age, and at spatial
frequencies of about 25 cycles/degree. In addition it can decline with age and also due to other factors such as
cataracts and diabetic retinopathy.[6]

In this image, the contrast amplitude depends only on the vertical coordinate, while the spatial frequency depends on the horizontal
coordinate. Observe that for medium frequency you need less contrast than for high or low frequency to detect the sinusoidal
fluctuation.

Improving contrast sensitivity

It was once thought that contrast sensitivity was relatively fixed and could only get worse with age. However new
research has shown that playing videogames can slightly improve contrast sensitivity. [7]

See also
Acutance
Radiocontrast
Contrast ratio
Contrast (vision) 23

External links
Details on luminance contrast [8]

References
[1] Campbell, FW and Robson, JG (1968). Application of Fourier analysis to the visibility of gratings. J. Physiol.
[2] "eye, human."Encyclopdia Britannica. 2008. Encyclopdia Britannica 2006 Ultimate Reference Suite DVD
[3] Travnikova, N. P. (1985). Efficiency of Visual Search. p.4. Mashinostroyeniye.
[4] Michelson, A. (1927). Studies in Optics. U. of Chicago Press.
[5] E. Peli (Oct. 1990). "Contrast in Complex Images" (http:/ / www. eri. harvard. edu/ faculty/ peli/ papers/ ContrastJOSA. pdf). Journal of the
Optical Society of America A 7 (10): 20322040. doi:10.1364/JOSAA.7.002032. .
[6] Peter Wenderoth. "The Contrast Sensitivity Function" (http:/ / vision. psy. mq. edu. au/ ~peterw/ csf. html). .
[7] "Contrast sensitivity improvement" (http:/ / www. clinical-ophthalmology. com/ index. php?option=com_content& view=article&
id=327:contrast-sensitivity-improved-by-video-games& catid=14:professional-news& Itemid=13). .
[8] http:/ / colorusage. arc. nasa. gov/ luminance_cont. php

Color space
A color model is an abstract mathematical model describing the way
colors can be represented as tuples of numbers, typically as three or
four values or color components (e.g. RGB and CMYK are color
models). However, a color model with no associated mapping function
to an absolute color space is a more or less arbitrary color system with
no connection to any globally-understood system of color
interpretation.

Adding a certain mapping function between the color model and a

certain reference color space results in a definite "footprint" within the
reference color space. This "footprint" is known as a gamut, and, in
combination with the color model, defines a new color space. For
example, Adobe RGB and sRGB are two different absolute color A comparison of the chromaticities enclosed by
spaces, both based on the RGB model. some color spaces.

In the most generic sense of the definition above, color spaces can be
defined without the use of a color model. These spaces, such as Pantone, are in effect a given set of names or
numbers which are defined by the existence of a corresponding set of physical color swatches. This article focuses
on the mathematical model concept.
Color space 24

Understanding the concept

A wide range of colors can be created by the primary colors of pigment
(cyan (C), magenta (M), yellow (Y), and black (K)). Those colors then
define a specific color space. To create a three-dimensional
representation of a color space, we can assign the amount of magenta
color to the representation's X axis, the amount of cyan to its Y axis,
and the amount of yellow to its Z axis. The resulting 3-D space
provides a unique position for every possible color that can be created
by combining those three pigments.

However, this is not the only possible color space. For instance, when
colors are displayed on a computer monitor, they are usually defined in
the RGB (red, green and blue) color space. This is another way of
making nearly the same colors (limited by the reproduction medium, A comparison of RGB and CMYK color models.
This image demonstrates the difference between
such as the phosphor (CRT) or filters and backlight (LCD)), and red,
how colors will look on a computer monitor
green and blue can be considered as the X, Y and Z axes. Another way (RGB) compared to how they will reproduce in a
of making the same colors is to use their Hue (X axis), their Saturation CMYK print process.
(Y axis), and their brightness Value (Z axis). This is called the HSV
color space. Many color spaces can be represented as three-dimensional (X,Y,Z) values in this manner, but some
have more, or fewer dimensions, and some cannot be represented in this way at all.

Notes
When formally defining a color space, the usual reference standard is the CIELAB or CIEXYZ color spaces, which
were specifically designed to encompass all colors the average human can see.
Since "color space" is a more specific term for a certain combination of a color model plus a mapping function, the
term "color space" tends to be used to also identify color models, since identifying a color space automatically
identifies the associated color model. Informally, the two terms are often used interchangeably, though this is strictly
incorrect. For example, although several specific color spaces are based on the RGB model, there is no such thing as
the RGB color space.
Since any color space defines colors as a function of the absolute reference frame, color spaces, along with device
profiling, allow reproducible representations of color, in both analogue and digital representations.

Conversion
Color space conversion is the translation of the representation of a color from one basis to another. This typically
occurs in the context of converting an image that is represented in one color space to another color space, the goal
being to make the translated image look as similar as possible to the original.

Density
The RGB color model is implemented in different ways, depending on the capabilities of the system used. By far the
most common general-used incarnation as of 2006 is the 24-bit implementation, with 8bits, or 256 discrete levels of
color per channel. Any color space based on such a 24-bit RGB model is thus limited to a range of 256256256
16.7million colors. Some implementations use 16bits per component for 48bits total, resulting in the same gamut
with a larger number of distinct colors. This is especially important when working with wide-gamut color spaces
(where most of the more common colors are located relatively close together), or when a large number of digital
filtering algorithms are used consecutively. The same principle applies for any color space based on the same color
Color space 25

model, but implemented in different bit depths.

Partial list of color spaces

CIE 1931 XYZ color space was one of the first attempts to produce a color space based on measurements of human
color perception (earlier efforts were by James Clerk Maxwell, Knig & Dieterici, and Abney at Imperial College)[1]
and it is the basis for almost all other color spaces. Derivatives of the CIE XYZ space include CIELUV, CIEUVW,
and CIELAB.

Generic color models

RGB uses additive color mixing, because it describes what kind of
light needs to be emitted to produce a given color. Light is added
together to create form from out of the darkness. RGB stores
individual values for red, green and blue. RGBA is RGB with an
additional channel, alpha, to indicate transparency.
Common color spaces based on the RGB model include sRGB,
Adobe RGB and ProPhoto RGB.
CMYK uses subtractive color mixing used in the printing process,
because it describes what kind of inks need to be applied so the
light reflected from the substrate and through the inks produces a
given color. One starts with a white substrate (canvas, page, etc),
and uses ink to subtract color from white to create an image.
CMYK stores ink values for cyan, magenta, yellow and black. Additive color mixing: Three overlapping lightbulbs in
a vacuum, adding together to create white.
There are many CMYK color spaces for different sets of inks,
substrates, and press characteristics (which change the dot gain or
transfer function for each ink and thus change the appearance).

YIQ was formerly used in NTSC (North America, Japan and

elsewhere) television broadcasts for historical reasons. This
system stores a luminance value with two chrominance values,
corresponding approximately to the amounts of blue and red in the
color. It is similar to the YUV scheme used in most video capture
systems[2] and in PAL (Australia, Europe, except France, which
uses SECAM) television, except that the YIQ color space is
rotated 33 with respect to the YUV color space. The YDbDr
scheme used by SECAM television is rotated in another way.

YPbPr is a scaled version of YUV. It is most commonly seen in its

digital form, YCbCr, used widely in video and image compression
schemes such as MPEG and JPEG. Subtractive color mixing: Three splotches of paint on
white paper, subtracting together to turn the paper
xvYCC is a new international digital video color space standard black.
published by the IEC (IEC 61966-2-4). It is based on the ITU
BT.601 and BT.709 standards but extends the gamut beyond the R/G/B primaries specified in those standards.
HSV (hue, saturation, value), also known as HSB (hue, saturation, brightness) is often used by artists because it is
often more natural to think about a color in terms of hue and saturation than in terms of additive or subtractive color
components. HSV is a transformation of an RGB colorspace, and its components and colorimetry are relative to the
RGB colorspace from which it was derived.
Color space 26

HSL (hue, saturation, lightness/luminance), also known as HLS or HSI (hue, saturation, intensity) is quite similar to
HSV, with "lightness" replacing "brightness". The difference is that the brightness of a pure color is equal to the
brightness of white, while the lightness of a pure color is equal to the lightness of a medium gray.

Commercial color spaces

Munsell color system
Natural Color System (NCS)

Special-purpose color spaces

The RG Chromaticity space is used in Computer vision applications. It shows the color of light (red, yellow,
green etc.), but not its intensity (dark, bright).

Obsolete color spaces

Early color spaces had two components. They largely ignored blue light because the added complexity of a
3-component process provided only a marginal increase in fidelity when compared to the jump from monochrome to
2-component color.
RG for early Technicolor film
RGK for early color printing

See also
Color theory
List of colors

External links
Color FAQ [5], Charles Poynton
FAQ about color physics [3], Stephen Westland
Color Science [4], Dan Bruton
Color Spaces [5], Rolf G. Kuehni (October 2003)
Colour spaces - perceptual, historical and applicational background [6], Marko Tkali (2003)
Color formats [7] for image and video processing - Color conversion [8] between RGB, YUV, YCbCr and YPbPr.

References
[1] William David Wright, 50 years of the 1931 CIE Standard Observer. Die Farbe, 29:4/6 (1981).
[2] Dean Anderson. "Color Spaces in Frame Grabbers: RGB vs. YUV" (http:/ / www. sensoray. com/ support/ frame_grabber_capture_modes.
htm). . Retrieved 2008-04-08.
[3] http:/ / www. colourware. co. uk/ cpfaq. htm
[4] http:/ / www. physics. sfasu. edu/ astro/ color. html
[5] http:/ / www4. ncsu. edu/ ~rgkuehni/ PDFs/ ColSp. pdf
[6] http:/ / ldos. fe. uni-lj. si/ docs/ documents/ 20030929092037_markot. pdf
[7] http:/ / www. equasys. de/ colorformat. html
[8] http:/ / www. equasys. de/ colorconversion. html
Color mapping 27

Color mapping
Color mapping example

Source image

Reference image

Source image color mapped using histogram matching

Color mapping is a function that maps (transforms) the colors of one (source) image to the colors of another (target)
image. A color mapping may be referred to as the algorithm that results in the mapping function or the algorithm that
transforms the image colors. Color mapping is also sometimes called "color transfer".

Algorithms
There are two types of color mapping algorithms: Those that employ the statistics of the colors of two images, and
those that rely on a given pixel correspondence between the images.
An example of an algorithm that employs the statistical properties of the images is histogram matching. This is a
classic algorithm for color mapping, suffering from the problem of sensitivity to image content differences. Newer
statistic-based algorithms deal with this problem. An example of such algorithm is adjusting the mean and the
standard deviation of Lab channels of the two images.[1]
A common algorithm for computing the color mapping when the pixel correspondence is given is building the
joint-histogram (see also co-occurrence matrix) of the two images and finding the mapping by using dynamic
programming based on the joint-histogram values.[2]
When the pixel correspondence is not given and the image contents are different (due to different point of view), the
statistics of the image corresponding regions can be used as an input to statistics-based algorithms, such as histogram
matching. The corresponding regions can be found by detecting the corresponding features.[3]
Color mapping 28

Applications
Color mapping can serve two different purposes: One is calibrating the colors of two cameras for further processing
using two or more sample images. The second is adjusting the colors of two images for perceptual visual
compatibility.
Color calibration is an important pre-processing task in computer vision applications. Many applications
simultaneously process two or more images and, therefore, need their colors to be calibrated. Examples of such
applications are: Image differencing, registration, object recognition, multi-camera tracking, co-segmentation and
stereo reconstruction.

References
[1] Color Transfer between Images (http:/ / www. cse. iitd. ac. in/ ~pkalra/ cs783/ assignment2/ ColorTransfer. pdf)
[2] Inter-Camera Color Calibration using Cross-Correlation Model Function (http:/ / www. porikli. com/ mysite/ pdfs/ icip2003-porikli. pdf)
[3] Piecewise-consistent Color Mappings of Images Acquired Under Various Conditions (http:/ / www. faculty. idc. ac. il/ toky/ Publications/
Conference/ colorMap. pdf)

Color management
In digital imaging systems, color management is the controlled conversion between the color representations of
various devices, such as image scanners, digital cameras, monitors, TV screens, film printers, computer printers,
offset presses, and corresponding media.
The primary goal of color management is to obtain a good match across color devices; for example, a video which
should appear the same color on a computer LCD monitor, a plasma TV screen, and on a printed frame of video.
Color management helps to achieve the same appearance on all of these devices, provided the devices are capable of
delivering the needed color intensities.
Parts of this technology are implemented in the operating system (OS), helper libraries, the application, and devices.
A cross-platform view of color management is the use of an ICC-compatible color management system. The
International Color Consortium (ICC) is an industry consortium which has defined an open standard for a Color
Matching Module (CMM) at the OS level, and color profiles for the devices and working space (color space the user
edits in).
There are other approaches to color management besides using ICC profiles. This is partly due to history and partly
because of other needs than the ICC standard covers. The film and broadcasting industries make use of many of the
same concepts, but they more frequently rely on boutique solutions. The film industry, for instance, uses 3D LUTs
(lookup table) to characterize color. At the consumer level, color management currently applies more to still images
than video, in which color management is still in its infancy.[1]

Hardware

Characterization
In order to describe the behavior of the various output devices, they must be compared (measured) in relation to a
standard color space. Often a step called linearization is performed first, in order to undo the effect of gamma
correction that was done to get the most out of limited 8-bit color paths. Instruments used for measuring device
colors include colorimeters and spectrophotometers. As an intermediate result, the device gamut is described in the
form of scattered measurement data. The transformation of the scattered measurement data into a more regular form,
usable by the application, is called profiling. Profiling is a complex process involving mathematics, intense
computation, judgment, testing, and iteration. After the profiling is finished, an idealized color description of the
Color management 29

device is created. This description is called a profile.

Calibration
Calibration is like characterization, except that it can include the adjustment of the device, as opposed to just the
measurement of the device. Color management is sometimes sidestepped by calibrating devices to a common
standard color space such as sRGB; when such calibration is done well enough, no color translations are needed to
get all devices to handle colors consistently. This avoidance of the complexity of color management was one of the
goals in the development of sRGB.

Color profiles

Embedding
Image formats themselves (such as TIFF, JPEG, PNG, EPS, PDF, and SVG) may contain embedded color profiles
but are not required to do so by the image format. The International Color Consortium standard was created to bring
various developers and manufacturers together. The ICC standard permits the exchange of output device
characteristics and color spaces in the form of metadata. This allows the embedding of color profiles into images as
well as storing them in a database or a profile directory.

Working spaces
Working spaces, such as sRGB, Adobe RGB or ProPhoto are color spaces that facilitate good results while editing.
For instance, pixels with equal values of R,G,B should appear neutral. Using a large (gamut) working space will lead
to posterization, while using a small working space will lead to clipping.[2] This trade-off is a consideration for the
critical image editor.

Color translation
Color translation, or color space conversion, is the translation of the representation of a color from one color space to
another. This calculation is required whenever data is exchanged inside a color-managed chain. Transforming
profiled color information to different output devices is achieved by referencing the profile data into a standard color
space. It is easy to convert colors from one device to a selected standard and from that color space to the colors of
another device. By ensuring that the reference color space covers the many possible colors that humans can see, this
concept allows one to exchange colors between many different color output devices.

Profile connection space

In the terminology of the International Color Consortium, a translation between two color spaces can go through a
profile connection space (PCS): Color Space 1 PCS (CIELAB or CIEXYZ) Color space 2; conversions into
and out of the PCS are each specified by a profile.[3]

Gamut mapping
Since different devices don't have the same gamut, they need some rearrangement near the borders of the gamut.
Some colors need to be shifted to the inside of the gamut as they otherwise cannot be represented on the output
device and would simply be clipped. For instance to print a mostly saturated blue from a monitor to paper with a
typical CMYK printer will surely fail. The paper blue will not be that saturated. Conversely, the bright cyan of an
inkjet printer cannot be easily presented on an average computer monitor. The color management system can utilize
various methods to achieve desired results and give experienced users control of the gamut mapping behavior.
Color management 30

Rendering intent
When the gamut of source color space exceeds that of the destination, saturated colors are liable to become clipped
(inaccurately represented). The color management module can deal with this problem in several ways. The ICC
specification includes four different rendering intents: absolute colorimetric, relative colorimetric, perceptual, and
saturation.[3]
Absolute colorimetric
Absolute colorimetry and relative colorimetry actually use the same table but differ in the adjustment for the white
point media. If the output device has a much larger gamut than the source profile, i.e., all the colors in the source can
be represented in the output, using the absolute colorimetry rendering intent would "ideally" (ignoring noise,
precision, etc) give an exact output of the specified CIELAB values. Perceptually, the colors may appear incorrect,
but instrument measurements of the resulting output would match the source. Colors outside of the proof print
system's possible color are mapped to the boundary of the color gamut. Absolute colorimetry is useful to get an exact
specified color (e.g., IBM blue), or to quantify the accuracy of mapping methods.
Relative colorimetric
The goal in relative colorimetry is to be truthful to the specified color, with only a correction for the media. Relative
colorimetry is useful in proofing applications, since you are using it to get an idea of how a print on one device will
appear on a different device. Media differences are the only thing you really would like to adjust for. Obviously there
has to be some gamut mapping going on also. Usually this is done in a way where hue and lightness are maintained
at the cost of reduced saturation. Relative colorimetric is the default rendering intent on most systems.
Perceptual and Saturation
The perceptual and saturation intents are where the results really depend upon the profile maker. This is even how
some of the competitors in this market differentiate themselves. These intents should be created by the profile maker
so that pleasing images occur with the perceptual intent while eye-catching business graphics occur with the
saturation intent. This is achieved through the use of different perceptual remaps of the data as well as different
gamut mapping methods. Perceptual rendering is recommended for color separation

Implementation

Color management module

Color matching module (also -method or -system) is a software algorithm that adjusts the numerical values that get
sent to or received from different devices so that the perceived color they produce remains consistent. The key issue
here is how to deal with a color that cannot be reproduced on a certain device in order to show it through a different
device as if it were visually the same color, just as when the reproducible color range between color transparencies
and printed matters are different. There is no common method for this process, and the performance depends on the
capability of each color matching method.
Some well known CMMs are ColorSync, Adobe CMM [4], LittleCMS, and ArgyllCMS [5].
Color management 31

Operating system level

Beginning with Windows Vista, color management in Windows is handled at the OS level through an ICC color
management standard and API known as Windows Color System.[6] WCS supplements the Image Color
Management (ICM) system in Windows 2000 and Windows XP, originally written by Heidelberg.[7] [8]
Apple's Mac operating systems have provided OS-level color management since 1993, through ColorSync.
Operating systems which use the X Window System for graphics use ICC profiles, but support for color
management on Linux is still quite nascent, with only a handful of applications supporting, some through LittleCMS.

Application level
Most web browsers ignore color profiles.[9] Notable exceptions are Safari, starting with version 2.0, and Firefox
starting with version 3.0. Although disabled by default in Firefox 3.0, users can enable ICC v2 and ICC v4 color
management by using an add-on[10] or setting the value "gfx.color_management.enabled" to "true" in Firefox 3's
"about:config" file.[11] . Starting from Firefox 3.5 color management is enabled by default only for tagged images,
although support is limited to ICC v2 profiles owing to a change in color management systems from 3.0[12] .
FastPictureViewer, a commercial image viewer for Windows, features full color management support (monitor
profile and image profiles).

See also
Color chart
International Color Consortium
IT8

Further reading
Fraser, Bruce; Bunting, Fred; Murphy, Chris (2004). Real World Color Management. Berkeley, CA, USA:
Peachpit Press. ISBN0-201-77340-6.
Schwartz, Charles S. (2004). Understanding Digital Cinema: A Professional Handbook. Focal Press.
ISBN978-0240806174.
Morovic, Jan (2008). Color Gamut Mapping. Wiley. ISBN978-0470030325.

External links
Color management and color science: Introduction [13] by Norman Koren.
ColorWiki [14] by Steve Upton.

References
[1] Fairchild, Mark. "A Color Scientist Looks at Video" (http:/ / www. cis. rit. edu/ fairchild/ PDFs/ PRO29. pdf). . Retrieved 2008-05-09.
[2] Rodney, Andrew. "The role of working spaces in Adobe applications" (http:/ / www. adobe. com/ digitalimag/ pdfs/ phscs2ip_colspace. pdf).
Adobe. . Retrieved 2008-05-09.
[3] Rodney, Andrew (2005). Color Management for Photographers. Focal Press. pp.3233. ISBN0240806492.
[4] http:/ / www. adobe. com/ support/ downloads/ detail. jsp?ftpID=3618
[5] http:/ / www. argyllcms. com/
[6] Upton, Steve (February 2008). Vista's New Color Management System: WCS (http:/ / www. colorwiki. com/ wiki/
Vista's_New_Color_Management_System_-_WCS).
[7] Microsoft (1997-04-23). "Microsoft Licenses LinoColorCMM Technology To Improve Color Management in Windows" (http:/ / www.
microsoft. com/ presspass/ press/ 1997/ apr97/ linopr. mspx). . Retrieved 2008-05-08.
[8] The reader may verify this by examining the Properties of any ICM profile. The Profile Information tab should contain the entry
"LinoColorCMM by Heidelberger Druckmaschinen AG".
Color management 32

[9] Smith, Colin; Kabili, Jan (2005). How to Wow: Photoshop CS2 for the Web (http:/ / books. google. com/ books?id=I1H0fNcAPHcC&
pg=PT17& dq=web+ browsers+ ignore+ "icc+ profiles"& ei=FXsjSIKHBae-ygSmmJjGDQ& client=firefox-a&
sig=xr_yp7U_lzTv6M3KeChpRgAkTX8). Berkeley, CA: Peachpit Press. pp.13. ISBN0-321-39394-5. . Retrieved 2008-05-08. "Many
designers choose not to include ICC Profiles with Web files because most Web browsers can't read them and they increase the size of a file."
[10] Color Management add-on (https:/ / addons. mozilla. org/ en-US/ firefox/ addon/ 6891) by Sean Hayes.
[11] Firefox 3: Color profile support (oh the pretty, pretty colors) (http:/ / www. dria. org/ wordpress/ archives/ 2008/ 04/ 29/ 633/ ), Deb
Richardson, Mozilla Corporation.
[12] (https:/ / bugzilla. mozilla. org/ show_bug. cgi?id=463221)
[13] http:/ / www. normankoren. com/ color_management. html
[14] http:/ / www. colorwiki. com

Digital Imaging and Communications in

Medicine
Digital Imaging and Communications in Medicine (DICOM) is a
standard for handling, storing, printing, and transmitting information in
medical imaging. It includes a file format definition and a network
communications protocol. The communication protocol is an
application protocol that uses TCP/IP to communicate between
systems. DICOM files can be exchanged between two entities that are
capable of receiving image and patient data in DICOM format. The
National Electrical Manufacturers Association (NEMA) holds the A nuclear medicine image in DICOM format
copyright to this standard.[1] It was developed by the DICOM
Standards Committee, whose members[2] are also partly members of NEMA.[3]

DICOM enables the integration of scanners, servers, workstations, printers, and network hardware from multiple
manufacturers into a picture archiving and communication system (PACS). The different devices come with DICOM
conformance statements which clearly state the DICOM classes they support. DICOM has been widely adopted by
hospitals and is making inroads in smaller applications like dentists' and doctors' offices.
DICOM is known as NEMA Standard PS3, and as ISO Standard 12052.

Parts of the DICOM Standard

The DICOM standard is divided into related but independent parts:[4]
The links below are to the 2009 version published in January 2010. Additions to the standard (Supplements and
Change Proposals) since that publication are available through the DICOM Web site [5].
PS 3.1: Introduction and Overview [6]PDF(241KB)
PS 3.2: Conformance [7]PDF(6.46MB)
PS 3.3: Information Object Definitions [8]PDF(6.96MB)
PS 3.4: Service Class Specifications [9]PDF(1.07MB)
PS 3.5: Data Structure and Encoding [10]PDF(1.43MB)
PS 3.6: Data Dictionary [11]PDF(7.32MB)
PS 3.7: Message Exchange [12]PDF(1.97MB)
PS 3.8: Network Communication Support for Message Exchange [13]PDF(901KB)
PS 3.9: Retired (formerly Point-to-Point Communication Support for Message Exchange)
PS 3.10: Media Storage and File Format for Data Interchange [14]PDF(406KB)
PS 3.11: Media Storage Application Profiles [15]PDF(1.14MB)
PS 3.12: Storage Functions and Media Formats for Data Interchange [16]PDF(593KB)
Digital Imaging and Communications in Medicine 33

PS 3.13: Retired (formerly Print Management Point-to-Point Communication Support)

PS 3.14: Grayscale Standard Display Function [17]PDF(2.88MB)
PS 3.15: Security and System Management Profiles [18]PDF(1.00MB)
PS 3.16: Content Mapping Resource [19]PDF(3.08MB)
PS 3.17: Explanatory Information [20]PDF(3.28MB)
PS 3.18: Web Access to DICOM Persistent Objects (WADO) [21]PDF(291KB)

History
DICOM is the third version of a standard developed by
American College of Radiology (ACR) and National
Electrical Manufacturers Association (NEMA).
In the beginning of the 1980s it was almost impossible
for anyone other than manufacturers of computed
tomography or magnetic resonance imaging devices to
decode the images that the machines generated.
Radiologists wanted to use the images for dose-planning
for radiation therapy. ACR and NEMA joined forces and
formed a standard committee in 1983. Their first
standard, ACR/NEMA 300, was released in 1985. Very
soon after its release, it became clear that improvements
were needed. The text was vague and had internal
contradictions.

In 1988 the second version was released. This version

gained more acceptance among vendors. The image
transmission was specified as over a dedicated 25
differential (EIA-485) pair cable. The first demonstration
Front page of ACR/NEMA 300, version 1.0, which was released
of ACR/NEMA V2.0 interconnectivity technology was
in 1985
held at Georgetown University, May 21-23, 1990. Six
companies participated in this event, DeJarnette Research
Systems, General Electric Medical Systems, Merge Technologies, Siemens Medical Systems, Vortech (acquired by
Kodak that same year) and 3M. Commercial equipment supporting ACR/NEMA 2.0 was presented at the annual
meeting of the Radiological Society of North America (RSNA) in 1990 by these same vendors. Many soon realized
that the second version also needed improvement. Several extensions to ACR/NEMA 2.0 were created, like Papyrus
(developed by the University Hospital of Geneva, Switzerland) and SPI, (Standard Product Interconnect, driven by
Siemens Medical Systems and Philips Medical Systems).
The first large scale deployment of ACR/NEMA technology was made in 1992 by the US Army and Air Force as
part of the MDIS (Medical Diagnostic Imaging Support) [22] program run out of Ft. Detrick, Maryland. Loral
Aerospace and Siemens Medical Systems led a consortium of companies in deploying the first US military PACS
(Picture Archiving and Communications System) at all major Army and Air Force medical treatment facilities and
teleradiology nodes at a large number of US military clinics. DeJarnette Research Systems and Merge Technologies
provided the modality gateway interfaces from third party imaging modalities to the Siemens SPI network. The
Veterans Administration and the Navy also purchased systems off this contract.
In 1993 the third version of the standard was released. Its name was then changed to DICOM so as to improve the
possibility of international acceptance as a standard. New service classes were defined, network support added and
the Conformance Statement was introduced. Officially, the latest version of the standard is still 3.0, however, it has
Digital Imaging and Communications in Medicine 34

been constantly updated and extended since 1993. Instead of using the version number the standard is often
version-numbered using the release year, like "the 2007 version of DICOM".
While the DICOM standard has achieved a near universal level of acceptance amongst medical imaging equipment
vendors and healthcare IT organizations, the standard has its limitations. DICOM is a standard directed at addressing
technical interoperability issues in medical imaging. It is not a framework or architecture for achieving a useful
clinical workflow. RSNA's Integrating the Healthcare Enterprise (IHE) initiative layered on top of DICOM (and
HL-7) provides this final piece of the medical imaging interoperability puzzle.

Derivations
There are some derivations from the DICOM standard into other application areas. This includes
DICONDE - Digital Imaging and Communication in Nondestructive Evaluation, was established in 2004 as a
way for nondestructive testing manufacturers and users to share image data.[23]
DICOS - Digital Imaging and Communication in Security was established in 2009 to be used for image sharing in
airport security.[24]

DICOM Data Format

DICOM differs from other data formats in that it groups
information into data sets. That means that a file of a chest X-Ray
image, for example, actually contains the patient ID within the file,
so that the image can never be separated from this information by
mistake.
A DICOM data object consists of a number of attributes, including
items such as name, ID, etc., and also one special attribute
containing the image pixel data (i.e. logically, the main object has
no "header" as such - merely a list of attributes, including the pixel
data). A single DICOM object can only contain one attribute
containing pixel data. For many modalities, this corresponds to a
single image. But note that the attribute may contain multiple
"frames", allowing storage of cine loops or other multi-frame data.
Another example is NM data, where an NM image by definition is
a multi-dimensional multi-frame image. In these cases three- or four-dimensional data can be encapsulated in a
single DICOM object. Pixel data can be compressed using a variety of standards, including JPEG, JPEG Lossless,
JPEG 2000, and Run-length encoding (RLE). LZW (zip) compression can be used for the whole data set (not just the
pixel data) but this is rarely implemented.

DICOM uses three different Data Element encoding schemes. With Explicit Value Representation (VR) Data
Elements, for VRs that are not OB, OW, OF, SQ, UT, or UN, the format for each Data Element is: GROUP (2 bytes)
ELEMENT (2 bytes) VR (2 bytes) LengthInByte (2 bytes) Data (variable length). For the other Explicit Data
Elements or Implicit Data Elements, see section 7.1 of Part 5 of the DICOM Standard.
The same basic format is used for all applications, including network and file usage, but when written to a file,
usually a true "header" (containing copies of a few key attributes and details of the application which wrote it) is
added.
Digital Imaging and Communications in Medicine 35

DICOM Value Representations

Extracted from Chapter 6.2 of
PS 3.5: Data Structure and Encoding [25]PDF(1.43MiB)

Value Description
Representation

AE Application Entity

AS Age String

AT Attribute Tag

CS Code String

DA Date

DS Decimal String

DT Date/Time

FL Floating Point Single (4 bytes)

FD Floating Point Double (8 bytes)

IS Integer String

LO Long String

LT Long Text

OB Other Byte

OF Other Float

OW Other Word

PN Person Name

SH Short String

SL Signed Long

SQ Sequence of Items

SS Signed Short

ST Short Text

TM Time

UI Unique Identifier

UL Unsigned Long

UN Unknown

US Unsigned Short

UT Unlimited Text

In addition to a Value Representation, each attribute also has a Value Multiplicity to indicate the number of data
elements contained in the attribute. For character string value representations, if more than one data element is being
encoded, the successive data elements are separated by the backslash character "\".
Digital Imaging and Communications in Medicine 36

DICOM Services
DICOM consists of many different services, most of which involve transmission of data over a network, and the file
format below is a later and relatively minor addition to the standard.

Store
The DICOM Store service is used to send images or other persistent objects (structured reports, etc.) to a PACS or
workstation.

Storage Commitment
The DICOM storage commitment service is used to confirm that an image has been permanently stored by a device
(either on redundant disks or on backup media, e.g. burnt to a CD). The Service Class User (SCU - similar to a
client), a modality or workstation, etc., uses the confirmation from the Service Class Provider (SCP - similar to a
server), an archive station for instance, to make sure that it is safe to delete the images locally.

Query/Retrieve
This enables a workstation to find lists of images or other such objects and then retrieve them from a PACS.

Modality Worklist
This enables a piece of imaging equipment (a modality) to obtain details of patients and scheduled examinations
electronically, avoiding the need to type such information multiple times (and the mistakes caused by retyping).

Modality Performed Procedure Step

A complementary service to Modality Worklist, this enables the modality to send a report about a performed
examination including data about the images acquired, beginning time, end time, and duration of a study, dose
delivered, etc. It helps give the radiology department a more precise handle on resource (acquisition station) use.
Also known as MPPS, this service allows a modality to better coordinate with image storage servers by giving the
server a list of objects to send before or while actually sending such objects.

Printing
The DICOM Printing service is used to send images to a DICOM Printer, normally to print an "X-Ray" film. There
is a standard calibration (defined in DICOM Part 14) to help ensure consistency between various display devices,
including hard copy printout.

Off-line Media (DICOM Files)

The off-line media files correspond to Part 10 of the DICOM standard. It describes how to store medical imaging
information on removable media. Except for the data set containing, for example, an image and demography, it's also
mandatory to include the File Meta Information.
DICOM restricts the filenames on DICOM media to 8 characters (some systems wrongly use 8.3, but this is not
legal). No information must be extracted from these names (PS3.10 Section 6.2.3.2). This is a common source of
problems with media created by developers who did not read the specifications carefully. This is a historical
requirement to maintain compatibility with older existing systems. It also mandates the presence of a media
directory, the DICOMDIR file, which provides index and summary information for all the DICOM files on the
media. The DICOMDIR information provides substantially greater information about each file than any filename
could, so there is less need for meaningful file names.
DICOM files typically have a .dcm file extension.
Digital Imaging and Communications in Medicine 37

The MIME type for DICOM files is defined by RFC 3240 as application/dicom.
There is also an ongoing media exchange test and "connectathon" process for CD media and network operation that
is organized by the IHE organization.

Application areas
Modality Description

BI Modality of type Biomagnetic Imaging

CD Modality of type Color Flow Doppler-Retired 2008

CR Modality of type Computed Radiography

CT Modality of type Computed Tomography

DD Modality of type Duplex Doppler-Retired 2008

DG Modality of type Diaphanography

DX Modality of type Digital Radiography

EC Modality of type Echo cardiography - Retired

EM Modality of type Electron Microscope

ES Modality of type Endoscopy

GM Modality of type General Microscopy

LS Modality of type Laser Surface Scan

MA Modality of type Magnetic Resonance Angiography - Retired

MG Modality of type Mammography

MR Modality of type Magnetic Resonance

MS Modality of type Magnetic Resonance Spectroscopy - Retired

NM Modality of type Nuclear Medicine

OT Modality of type Other

PT Modality of type Positron Emission Tomography

RF Modality of type Radio Fluoroscopy

RG Modality of type Radiographic Imaging (conventional film screen)

RT Modality of type Radiation Therapy

SC Modality of type Secondary Capture

SM Modality of type Slide Microscopy

ST Modality of type Single-Photon Emission Computed Tomography - Retired 2008

TG Modality of type Thermography

US Modality of type Ultra Sound

VL Modality of type Visible Light

XA Modality of type X-Ray Angiography

XC Modality of type External Camera (Photography)

Examples of Modalities supported in DICOM are:

AS = Angioscopy-Retired
BI = Biomagnetic Imaging
Digital Imaging and Communications in Medicine 38

CD = Color Flow Doppler-Retired

CF = Cinefluorography (retired)
CP = Colposcopy Retired
CR = Computed Radiography
CS = CystoscopyRetired
CT = Computed Tomography
DD = Duplex Doppler Retired
DF = Digital Fluoroscopy (retired)
DG = Diaphanography
DM = Digital Microscopy
DS = Digital Subtraction Angiography Retired
DX = Digital radiography
EC = Echocardiography Retired
ES = Endoscopy
FA = Fluorescein Angiography Retired
FS = Fundoscopy Retired
HC = Hard Copy
LP = Laparoscopy Retired
LS = Laser Surface Scan
MA = Magnetic resonance angiography Retired
MG = Mammography
MR = Magnetic Resonance
MS = Magnetic Resonance Spectroscopy Retired
NM = Nuclear Medicine
OT = Other
PT = Positron Emission Tomography (PET)
RF = Radio Fluoroscopy
RG = Radiographic Imaging (conventional film screen)
RTDOSE (a.k.a. RD) = Radiotherapy Dose
RTIMAGE = Radiotherapy Image
RTPLAN (a.k.a. RP) = Radiotherapy Plan
RTSTRUCT (a.k.a. RS) = Radiotherapy Structure Set
SR = Structured Reporting
ST = Single-photon Emission Computed Tomography Retired
TG = Thermography
US = Ultrasound
VF = Videofluorography (retired)
XA = X-Ray Angiography
XC = eXternal Camera
ECG = Electrocardiograms
Digital Imaging and Communications in Medicine 39

DICOM port numbers

DICOM have reserved the following TCP and UDP port numbers by the Internet Assigned Numbers Authority
(IANA):
104 well-known port for DICOM over TCP or UDP. Since 104 is in the reserved subset, many operating systems
require special privileges to use it.
2761 registered port for DICOM using Integrated Secure Communication Layer (ISCL) over TCP or UDP
2762 registered port for DICOM using Transport Layer Security (TLS) over TCP or UDP
11112 registered port for DICOM using standard, open communication over TCP or UDP
The standard recommends but does not require the use of these port numbers.

External links
The latest DICOM specification [26]
DICOM Standard Status (approved and proposed changes) [27]
Brief introduction to DICOM [28]
Introduction to DICOM using OsiriX [29]
Medical Image FAQ part 2 [30] - Standard formats including DICOM.
Medical Image FAQ part 8 [31] - Contains a long list DICOM software.
Collection of DICOM images (clinical images and technical testpatterns) [32]

References
[1] DICOM brochure (http:/ / medical. nema. org/ dicom/ geninfo/ Brochure. pdf)
[2] MEMBERS of the DICOM STANDARDS COMMITTEE (http:/ / medical. nema. org/ members. pdf)
[3] NEMA Members (http:/ / www. nema. org/ about/ members/ )
[4] "6.1 DIMSE Services" (ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 08_01pu. pdf). Digital Imaging and Communications in Medicine
(DICOM) Part 1: Introduction and Overview. National Electrical Manufacturers Association. 2006. pp.11. .
[5] http:/ / dicom. nema. org
[6] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_01pu. pdf
[7] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_02pu. pdf
[8] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_03pu3. pdf
[9] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_04pu. pdf
[10] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_05pu3. pdf
[11] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_06pu3. pdf
[12] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_07pu2. pdf
[13] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_08pu. pdf
[14] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_10pu. pdf
[15] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_11pu. pdf
[16] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_12pu. pdf
[17] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_14pu. pdf
[18] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_15pu. pdf
[19] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_16pu. pdf
[20] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_17pu. pdf
[21] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/ 09_18pu. pdf
[22] http:/ / www. ncbi. nlm. nih. gov/ pubmed/ 7612705?dopt=Abstract
[23] http:/ / www. astm. org: If a Picture Is Worth 1,000 Words, then Pervasive, Ubiquitous Imaging Is Priceless (http:/ / www. astm. org/
SNEWS/ OCTOBER_2003/ voelker_oct03. html)
[24] http:/ / www. nema. org: Industrial Imaging and Communications Section (http:/ / www. nema. org/ prod/ security/ indust-Img. cfm)
[25] http:/ / medical. nema. org/ dicom/ 2007/ 07_05pu. pdf
[26] ftp:/ / medical. nema. org/ medical/ dicom/ 2009/
[27] http:/ / www. dclunie. com/ dicom-status/ status. html
[28] http:/ / www. cabiatl. com/ mricro/ dicom/ index. html
[29] http:/ / www. saravanansubramanian. com/ Saravanan/ Articles_On_Software/ Entries/ 2010/ 2/ 10_Introduction_to_the_DICOM_Standard.
html
Digital Imaging and Communications in Medicine 40

[30] http:/ / www. dclunie. com/ medical-image-faq/ html/ part2. html

[31] http:/ / www. dclunie. com/ medical-image-faq/ html/ part8. html
[32] http:/ / www. aycan. de/ main/ lp/ sample_dicom_images. html

JPEG 2000

Comparison of JPEG 2000 with the original JPEG format.

Filename extension .jp2, .j2k, .jpf, .jpx, .jpm, .mj2

Internet media type image/jp2, image/jpx, image/jpm, video/mj2

Uniform Type Identifier public.jpeg-2000

Developed by Joint Photographic Experts Group

Type of format graphics file format

Standard(s) ISO/IEC 15444

JPEG 2000 is a wavelet-based image compression standard and coding system. It was created by the Joint
Photographic Experts Group committee in 2000 with the intention of superseding their original discrete cosine
transform-based JPEG standard (created in 1992). The standardized filename extension is .jp2 for ISO/IEC 15444-1
conforming files and .jpx for the extended part-2 specifications, published as ISO/IEC 15444-2. The registered
MIME types are defined in RFC 3745. For ISO/IEC 15444-1 it is image/jp2.
While there is a modest increase in compression performance of JPEG 2000 compared to JPEG, the main advantage
offered by JPEG 2000 is the significant flexibility of the codestream. The codestream obtained after compression of
an image with JPEG 2000 is scalable in nature, meaning that it can be decoded in a number of ways; for instance, by
truncating the codestream at any point, one may obtain a representation of the image at a lower resolution, or
signal-to-noise ratio. By ordering the codestream in various ways, applications can achieve significant performance
increases. However, as a consequence of this flexibility, JPEG 2000 requires encoders/decoders that are complex and
computationally demanding. Another difference, in comparison with JPEG, is in terms of visual artifacts: JPEG 2000
produces ringing artifacts, manifested as blur and rings near edges in the image, while JPEG produces ringing
artifacts and 'blocking' artifacts, due to its 88 blocks.
JPEG 2000 has been published as an ISO standard, ISO/IEC 15444. As of 2010, JPEG 2000 is not widely supported
in web browsers, and hence is not generally used on the World Wide Web.
JPEG 2000 41

Features
Superior compression performance: at high bit rates, where artifacts become nearly imperceptible, JPEG 2000 has
a small machine-measured fidelity advantage over JPEG. At lower bit rates (e.g., less than 0.25 bits/pixel for
gray-scale images), JPEG 2000 has a much more significant advantage over certain modes of JPEG: artifacts are
less visible and there is almost no blocking. The compression gains over JPEG are attributed to the use of DWT
and a more sophisticated entropy encoding scheme.
Multiple resolution representation: JPEG 2000 decomposes the image into a multiple
resolution representation in the course of its compression process. This representation can
be put to use for other image presentation purposes beyond compression as such.
Progressive transmission by pixel and resolution accuracy, commonly referred to as
progressive decoding and signal-to-noise ratio (SNR) scalability: JPEG 2000 provides
efficient code-stream organizations which are progressive by pixel accuracy and by image
resolution (or by image size). This way, after a smaller part of the whole file has been
received, the viewer can see a lower quality version of the final picture. The quality then
improves progressively through downloading more data bits from the source. The 1992
JPEG standard also has a progressive transmission feature but it's rarely used.

Lossless and lossy compression: Like JPEG 1992,[1] the JPEG 2000 standard provides
both lossless and lossy compression in a single compression architecture. Lossless
compression is provided by the use of a reversible integer wavelet transform in JPEG
2000.
Random code-stream access and processing, also referred as Region Of Interest (ROI):
JPEG 2000 code streams offer several mechanisms to support spatial random access or
region of interest access at varying degrees of granularity. This way it is possible to store
different parts of the same picture using different quality.
Error resilience: Like JPEG 1992, JPEG 2000 is robust to bit errors introduced by noisy
communication channels, due to the coding of data in relatively small independent blocks.
Flexible file format: The JP2 and JPX file formats allow for handling of color-space
information, metadata, and for interactivity in networked applications as developed in the Top-to-bottom
JPEG Part 9 JPIP protocol. demonstration of the
artifacts of JPEG 2000
Side channel spatial information: it fully supports transparency and alpha planes. compression. The
numbers indicate the
More advantages associated with JPEG 2000 can be referred to from the Official JPEG 2000
compression ratio
page [2]. used. See the unscaled
image for an accurate
view.
JPEG 2000 42

JPEG 2000 image coding system - Parts

The JPEG 2000 image coding system (ISO/IEC 15444) consists of following parts:

JPEG 2000 image coding system - Parts[3] [4]

Part Number First public Latest public Latest Identical Title Description
release date release date amendment ITU-T
(First (edition) standard
edition)

Part 2000 2004 [6] [7] Core coding system the basic characteristics of
ISO/IEC 15444-1 2006 T.800
1 [5] JPEG 2000 compression
(.jp2)

Part 2004 2004 [9] [10] Extensions (.jpx, .jpf)

ISO/IEC 15444-2 2006 T.801
2 [8]

Part 2002 [12] [13] Motion JPEG 2000 (.mj2)

ISO/IEC 15444-3 2007 T.802
3 [11]

Part 2002 [15] [16] Conformance testing

ISO/IEC 15444-4 2004 T.803
4 [14]

Part 2003 2003 [18] [19] Reference software Java and C

ISO/IEC 15444-5 2003 T.804
5 [17] implementations

Part 2003 2003 [21] Compound image file (.jpm) e.g. document
ISO/IEC 15444-6 2007
6 [20] format imaging, for pre-press and
fax-like applications

Part [3]
abandoned Guideline of minimum (Technical Report on
7 support function of Minimum Support
[22] [23]
ISO/IEC 15444-1 Functions )

Part 2007 2007 [25] [26] Secure JPEG 2000 JPSEC (security aspects)
ISO/IEC 15444-8 2008 T.807
8 [24]

Part 2005 2005 [28] [29] Interactivity tools, APIs JPIP (interactive protocols
ISO/IEC 15444-9 2008 T.808
9 [27] and protocols and API)

Part [31] [32]

ISO/IEC 15444-10 2008 2008 Extensions for JP3D (volumetric
2008 T.809
10 [30] three-dimensional data imaging)

Part [34] [35]

ISO/IEC 15444-11 2007 Wireless JPWL (wireless
2007 T.810
11 [33] applications)

Part [37]
ISO/IEC 15444-12 2004 ISO base media file format
2008
12 [36]

Part [39] [40]

ISO/IEC 15444-13 2008 An entry level JPEG 2000
2008 T.812
13 [38] encoder

Part XML structural [43]

under JPXML
14 [41] representation and
development
[42] reference
JPEG 2000 43

Technical discussion
The aim of JPEG 2000 is not only improving compression performance over JPEG but also adding (or improving)
features such as scalability and editability. In fact, JPEG 2000's improvement in compression performance relative to
the original JPEG standard is actually rather modest and should not ordinarily be the primary consideration for
evaluating the design. Very low and very high compression rates are supported in JPEG 2000. In fact, the graceful
ability of the design to handle a very large range of effective bit rates is one of the strengths of JPEG 2000. For
example, to reduce the number of bits for a picture below a certain amount, the advisable thing to do with the first
JPEG standard is to reduce the resolution of the input image before encoding it. That's unnecessary when using JPEG
2000, because JPEG 2000 already does this automatically through its multiresolution decomposition structure. The
following sections describe the algorithm of JPEG 2000.

Color components transformation

Initially, images have to be transformed from the RGB color space to another color space, leading to three
components that are handled separately. There are two possible choices:
1. Irreversible Color Transform (ICT) uses the well known YCBCR color space. It is called "irreversible" because it
has to be implemented in floating or fix-point and causes round-off errors.
2. Reversible Color Transform (RCT) uses a modified YUV color space that does not introduce quantization errors,
so it is fully reversible. Proper implementation of the RCT requires that numbers are rounded as specified that
cannot be expressed exactly in matrix form. The transformation is:

and

The chrominance components can be, but do not necessarily have to be, down-scaled in resolution; in fact, since the
wavelet transformation already separates images into scales, downsampling is more effectively handled by dropping
the finest wavelet scale. This step is called multiple component transformation in the JPEG 2000 language since its
usage is not restricted to the RGB color model.

Tiling
After color transformation, the image is split into so-called tiles, rectangular regions of the image that are
transformed and encoded separately. Tiles can be any size, and it is also possible to consider the whole image as one
single tile. Once the size is chosen, all the tiles will have the same size (except optionally those on the right and
bottom borders). Dividing the image into tiles is advantageous in that the decoder will need less memory to decode
the image and it can opt to decode only selected tiles to achieve a partial decoding of the image. The disadvantage of
this approach is that the quality of the picture decreases due to a lower peak signal-to-noise ratio. Using many tiles
can create a blocking effect similar to the older JPEG 1992 standard.
JPEG 2000 44

Wavelet transform
These tiles are then wavelet transformed to an
arbitrary depth, in contrast to JPEG 1992 which uses
an 88 block-size discrete cosine transform. JPEG
2000 uses two different wavelet transforms:
1. irreversible: the CDF 9/7 wavelet transform. It is
said to be "irreversible" because it introduces
quantization noise that depends on the precision of
the decoder.
2. reversible: a rounded version of the biorthogonal
CDF 5/3 wavelet transform. It uses only integer
CDF 5/3 wavelet used for lossless compression.
coefficients, so the output does not require
rounding (quantization) and so it does not
introduce any quantization noise. It is used in
lossless coding.

The wavelet transforms are implemented by the lifting

scheme or by convolution.

Quantization
After the wavelet transform, the coefficients are
scalar-quantized to reduce the amount of bits to
represent them, at the expense of a loss of quality. The
output is a set of integer numbers which have to be
encoded bit-by-bit. The parameter that can be changed
to set the final quality is the quantization step: the
greater the step, the greater is the compression and the
loss of quality. With a quantization step that equals 1,
no quantization is performed (it is used in lossless
compression). An example of the wavelet transform that is used in JPEG 2000. This
is a 2nd-level CDF 9/7 wavelet transform.

Coding
The result of the previous process is a collection of sub-bands which represent several approximation scales. A
sub-band is a set of coefficients real numbers which represent aspects of the image associated with a certain
frequency range as well as a spatial area of the image.
The quantized sub-bands are split further into precincts, rectangular regions in the wavelet domain. They are
typically selected in a way that the coefficients within them across the sub-bands form approximately spatial blocks
in the (reconstructed) image domain, though this is not a requirement.
Precincts are split further into code blocks. Code blocks are located in a single sub-band and have equal sizes
except those located at the edges of the image. The encoder has to encode the bits of all quantized coefficients of a
code block, starting with the most significant bits and progressing to less significant bits by a process called the
EBCOT scheme. EBCOT here stands for Embedded Block Coding with Optimal Truncation. In this encoding
process, each bit plane of the code block gets encoded in three so-called coding passes, first encoding bits (and
signs) of insignificant coefficients with significant neighbors (i.e., with 1-bits in higher bit planes), then refinement
bits of significant coefficients and finally coefficients without significant neighbors. The three passes are called
JPEG 2000 45

Significance Propagation, Magnitude Refinement and Cleanup pass, respectively.

Clearly, in lossless mode all bit planes have to be encoded by the EBCOT, and no bit planes can be dropped.
The bits selected by these coding passes then get encoded by a context-driven binary arithmetic coder, namely the
binary MQ-coder. The context of a coefficient is formed by the state of its nine neighbors in the code block.
The result is a bit-stream that is split into packets where a packet groups selected passes of all code blocks from a
precinct into one indivisible unit. Packets are the key to quality scalability (i.e., packets containing less significant
bits can be discarded to achieve lower bit rates and higher distortion).
Packets from all sub-bands are then collected in so-called layers. The way the packets are built up from the
code-block coding passes, and thus which packets a layer will contain, is not defined by the JPEG 2000 standard, but
in general a codec will try to build layers in such a way that the image quality will increase monotonically with each
layer, and the image distortion will shrink from layer to layer. Thus, layers define the progression by image quality
within the code stream.
The problem is now to find the optimal packet length for all code blocks which minimizes the overall distortion in a
way that the generated target bitrate equals the demanded bit rate.
While the standard does not define a procedure as to how to perform this form of ratedistortion optimization, the
general outline is given in one of its many appendices: For each bit encoded by the EBCOT coder, the improvement
in image quality, defined as mean square error, gets measured; this can be implemented by an easy table-lookup
algorithm. Furthermore, the length of the resulting code stream gets measured. This forms for each code block a
graph in the ratedistortion plane, giving image quality over bitstream length. The optimal selection for the
truncation points, thus for the packet-build-up points is then given by defining critical slopes of these curves, and
picking all those coding passes whose curve in the ratedistortion graph is steeper than the given critical slope. This
method can be seen as a special application of the method of Lagrange multiplier which is used for optimization
problems under constraints. The Lagrange multiplier, typically denoted by ?, turns out to be the critical slope, the
constraint is the demanded target bitrate, and the value to optimize is the overall distortion.
Packets can be reordered almost arbitrarily in the JPEG 2000 bit-stream; this gives the encoder as well as image
servers a high degree of freedom.
Already encoded images can be sent over networks with arbitrary bit rates by using a layer-progressive encoding
order. On the other hand, color components can be moved back in the bit-stream; lower resolutions (corresponding to
low-frequency sub-bands) could be sent first for image previewing. Finally, spatial browsing of large images is
possible through appropriate tile and/or partition selection. All these operations do not require any re-encoding but
only byte-wise copy operations.
JPEG 2000 46

Performance
Compared to the previous JPEG standard, JPEG 2000
delivers a typical compression gain in the range of
20%, depending on the image characteristics.
Higher-resolution images tend to benefit more, where
JPEG-2000's spatial-redundancy prediction can
contribute more to the compression process. In very
low-bitrate applications, studies have shown JPEG
2000 to be outperformed[44] by the intra-frame coding
mode of H.264. Good applications for JPEG 2000 are
large images, images with low-contrast edges e.g.,
medical images.

File format and code stream

Similar to JPEG-1, JPEG 2000 defines both a file
format and a code stream. Whereas the latter entirely
describes the image samples, the former includes This image shows the (accentuated) difference between an image
saved as JPEG 2000 (quality 50%) and the original.
additional meta-information such as the resolution of
the image or the color space that has been used to
encode the image. JPEG 2000 images should if stored as files be boxed in the JPEG 2000 file format, where
they get the .jp2 extender. The part-2 extension to JPEG 2000, i.e., ISO/IEC 15444-2, also enriches this file format
by including mechanisms for animation or composition of several code streams into one single image. Images in this
extended file-format use the .jpx extension.

There is no standardized extension for code-stream data because code-stream data is not to be considered to be stored
in files in the first place, though when done for testing purposes, the extension .jpc or .j2k appear frequently.

Metadata
For traditional JPEG, additional metadata, e.g. lighting and exposure conditions, is kept in an application marker in
the Exif format specified by the JEITA. JPEG 2000 chooses a different route, encoding the same metadata in XML
form. The reference between the Exif tags and the XML elements is standardized by the ISO TC42 committee in the
standard 12234-1.4.
Extensible Metadata Platform can also be embedded in JPEG 2000.

Applications of JPEG 2000

Some markets and applications intended to be served by this standard are listed below:
Consumer applications such as multimedia devices (e.g., digital cameras, personal digital assistants, 3G mobile
phones, color facsimile, printers, scanners, etc.)
Client/server communication (e.g., the Internet, Image database, Video streaming, video server, etc.)
Military/surveillance (e.g., HD satellite images, Motion detection, network distribution and storage, etc.)
Medical imagery, esp. the DICOM specifications for medical data interchange.
Remote sensing
High-quality frame-based video recording, editing and storage.
Digital cinema
JPEG 2000 47

JPEG 2000 has many design commonalities with the ICER image compression format that is used to send images
back from the Mars rovers.
World Meteorological Organization has built JPEG 2000 Compression into the new GRIB2 file format. The
GRIB file structure is designed for global distribution of meteorological data. The implementation of JPEG 2000
compression in GRIB2 has reduced file sizes up to 80%.[45]

Comparison with PNG

Although JPEG 2000 format supports lossless encoding, it is not intended to completely supersede today's dominant
lossless image file formats.
The PNG (Portable Network Graphics) format is still more space-efficient in the case of images with many pixels of
the same color, such as diagrams, and supports special compression features that JPEG 2000 does not.

Legal issues
JPEG 2000 is by itself licensed, but the contributing companies and organizations agreed that licenses for its first
part the core coding system can be obtained free of charge from all contributors.
The JPEG committee has stated:
It has always been a strong goal of the JPEG committee that its standards should be implementable in their
baseline form without payment of royalty and license fees... The up and coming JPEG 2000 standard has been
prepared along these lines, and agreement reached with over 20 large organizations holding many patents in
this area to allow use of their intellectual property in connection with the standard without payment of license
fees or royalties.[46]
However, the JPEG committee has also noted that undeclared and obscure submarine patents may still present a
hazard:
It is of course still possible that other organizations or individuals may claim intellectual property rights that
affect implementation of the standard, and any implementers are urged to carry out their own searches and
investigations in this area.[47]
Because of this statement, controversy remains in the software community concerning the legal status of the JPEG
2000 standard.
JPEG 2000 is included in most Linux distributions.

Related standards
Several additional parts of the JPEG 2000 standard exist; Amongst them are ISO/IEC 15444-2:2000, JPEG 2000
extensions defining the .jpx file format, featuring for example Trellis quantization, an extended file format and
additional color spaces[48] , ISO/IEC 15444-4:2000, the reference testing and ISO/IEC 15444-6:2000, the compound
image file format (.jpm), allowing compression of compound text/image graphics.[49]
Extensions for secure image transfer, JPSEC (ISO/IEC 15444-8), enhanced error-correction schemes for wireless
applications, JPWL (ISO/IEC 15444-11) and extensions for encoding of volumetric images, JP3D (ISO/IEC
15444-10) are also already available from the ISO.
JPEG 2000 48

JPIP protocol for streaming JPEG 2000 images

In 2005, a JPEG 2000 based image browsing protocol, called JPIP has been published as ISO/IEC 15444-9.[50]
Within this framework, only selected regions of potentially huge images have to be transmitted from an image server
on the request of a client, thus reducing the required bandwidth.

Motion JPEG 2000

Motion JPEG 2000 is defined in ISO/IEC 15444-3 and in ITU-T T.802[51] . It specifies the use of the JPEG 2000
codec for timed sequences of images (motion sequences), possibly combined with audio, and composed into an
overall presentation.[52] [53] It also defines a file format, based on ISO base media file format (ISO 15444-12).
Filename extensions for Motion JPEG 2000 video files are .mj2 and .mjp2 according to RFC 3745.
Motion JPEG 2000 (often referenced as MJ2 or MJP2) also is under consideration as a digital archival format by the
Library of Congress. It is an open ISO standard and an advanced update to MJPEG (or MJ), which was based on the
legacy JPEG format. Unlike common video codecs, such as MPEG-4, WMV, and DivX, MJ2 does not employ
temporal or inter-frame compression. Instead, each frame is an independent entity encoded by either a lossy or
lossless variant of JPEG 2000. Its physical structure does not depend on time ordering, but it does employ a separate
profile to complement the data. For audio, it supports LPCM encoding, as well as various MPEG-4 variants, as "raw"
or complement data.[54] [55]

ISO base media file format

ISO/IEC 15444-12 is identical with ISO/IEC 14496-12 (MPEG-4 Part 12) and it defines ISO base media file format.
For example, Motion JPEG 2000 file format, MP4 file format or 3GP file format are also based on this ISO base
media file format.[56] [57] [58] [59] [60]

GML JP2 Georeferencing

The Open Geospatial Consortium (OGC) has defined a metadata standard for georeferencing JPEG 2000 images
with embedded XML using the Geography Markup Language (GML) format: GML in JPEG 2000 for Geographic
Imagery Encoding (GMLJP2), version 1.0.0, dated 2006-01-18 [61] .
JP2 and JPX files containing GMLJP2 markup can be located and displayed in the correct position on the Earth's
surface by a suitable Geographic Information System (GIS), in a similar way to GeoTIFF images.

Application support

Applications

Application support for JPEG 2000

Program [62] [62] License
Basic Advanced

Read Write Read Write

ACDSee Yes Yes ? ? Proprietary

Yes Yes ? ? Proprietary

Accusoft Pegasus [63]

[64] Yes Yes Yes Yes Proprietary

Adobe Photoshop

Apple iPhoto Yes No Yes No Proprietary

Apple Preview Yes Yes Yes ? Proprietary

JPEG 2000 49

[65] Yes Yes ? ? GPL

Blender

Corel Photo-Paint Yes Yes Yes Yes Proprietary

[66] Yes No ? No Proprietary

DBGallery

[67] [68] [69] GPL

digiKam (KDE Yes Yes ? ?
)

FastStone Image Viewer Yes Yes Yes Yes Proprietary

FastStone MaxView Yes No Yes No Proprietary

- - LGPL
via extension
Mozilla Firefox 3 [70] ?

GIMP [71] No ? No GPL

Yes

[69] Yes Yes ? ? GPL

Gwenview (KDE )

ImageMagick License
ImageMagick Yes Yes No No [72]

IrfanView [73] [73] No No Proprietary

Yes Partial

[69] Yes Yes ? ? GPL

KolourPaint (KDE )

Mathematica Yes Yes ? ? Proprietary

via toolbox via via via Proprietary

Matlab
toolbox toolbox toolbox

Opera via Quicktime - ? - Proprietary

Paint Shop Pro Yes Yes Yes Yes Proprietary

Pixel image editor Yes Yes ? ? Proprietary

Safari Yes - ? - Proprietary

SilverFast Yes Yes Yes Yes Proprietary

SView5 Yes Yes ? ? Proprietary

[74] Yes Yes Yes Yes Proprietary

XnView

Ziproxy Yes Yes No No GPL

[1] The JPEG Still Picture Compression Standard (http:/ / www. cis. temple. edu/ ~vasilis/ Courses/ CIS750/ Papers/ doc_jpeg_c_5. pdf) pp.6-7
[2] http:/ / www. jpeg. org/ JPEG2000. htm
[3] JPEG. "Joint Photographic Experts Group, JPEG2000" (http:/ / www. jpeg. org/ jpeg2000/ index. html). . Retrieved 2009-11-01.
[4] IGN Standardization Team. "JPEG2000 (ISO 15444)" (http:/ / eden. ign. fr/ std/ JPEG2000/ index_html). . Retrieved 2009-11-01.
[5] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_ics/ catalogue_detail_ics. htm?csnumber=27687
[6] ISO. "ISO/IEC 15444-1:2004 - Information technology -- JPEG 2000 image coding system: Core coding system" (http:/ / www. iso. org/ iso/
iso_catalogue/ catalogue_ics/ catalogue_detail_ics. htm?csnumber=27687). . Retrieved 2009-11-01.
[7] http:/ / www. itu. int/ rec/ T-REC-T. 800
[8] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=33160
[9] ISO. "ISO/IEC 15444-2:2004 - Information technology -- JPEG 2000 image coding system: Extensions" (http:/ / www. iso. org/ iso/
iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=33160). . Retrieved 2009-11-01.
[10] http:/ / www. itu. int/ rec/ T-REC-T. 801
[11] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_ics/ catalogue_detail_ics. htm?csnumber=33875
[12] ISO. "ISO/IEC 15444-3:2007 - Information technology -- JPEG 2000 image coding system: Motion JPEG 2000" (http:/ / www. iso. org/ iso/
iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=41570). . Retrieved 2009-11-01.
[13] http:/ / www. itu. int/ rec/ T-REC-T. 802
[14] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_ics/ catalogue_detail_ics. htm?csnumber=33876
JPEG 2000 50

[15] ISO. "ISO/IEC 15444-4:2004 - Information technology -- JPEG 2000 image coding system: Conformance testing" (http:/ / www. iso. org/
iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=39079). . Retrieved 2009-11-01.
[16] http:/ / www. itu. int/ rec/ T-REC-T. 803
[17] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=33877
[18] ISO. "ISO/IEC 15444-5:2003 - Information technology -- JPEG 2000 image coding system: Reference software" (http:/ / www. iso. org/ iso/
iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=33877). . Retrieved 2009-11-01.
[19] http:/ / www. itu. int/ rec/ T-REC-T. 804
[20] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=35458
[21] ISO. "ISO/IEC 15444-6:2003 - Information technology -- JPEG 2000 image coding system -- Part 6: Compound image file format" (http:/ /
www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=35458). . Retrieved 2009-11-01.
[22] ISO/IEC JTC 1/SC 29/WG 1 (2000-12-08). "JPEG, JBIG - Resolutions of 22nd WG1 New Orleans Meeting" (http:/ / www. itscj. ipsj. or. jp/
sc29/ open/ 29view/ 29n39731. doc) (DOC). . Retrieved 2009-11-01.
[23] "22nd WG1 New Orleans Meeting, Draft Meeting Report" (http:/ / www. itscj. ipsj. or. jp/ sc29/ open/ 29view/ 29n39741. doc) (DOC).
2000-12-08. . Retrieved 2009-11-01.
[24] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=37382
[25] ISO. "ISO/IEC 15444-8:2007 - Information technology -- JPEG 2000 image coding system: Secure JPEG 2000" (http:/ / www. iso. org/ iso/
iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=37382). . Retrieved 2009-11-01.
[26] http:/ / www. itu. int/ rec/ T-REC-T. 807
[27] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=39413
[28] ISO. "ISO/IEC 15444-9:2005 - Information technology -- JPEG 2000 image coding system: Interactivity tools, APIs and protocols" (http:/ /
www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=39413). . Retrieved 2009-11-01.
[29] http:/ / www. itu. int/ rec/ T-REC-T. 808
[30] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=40024
[31] ISO. "ISO/IEC 15444-10:2008 - Information technology -- JPEG 2000 image coding system: Extensions for three-dimensional data" (http:/ /
www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=40024). . Retrieved 2009-11-01.
[32] http:/ / www. itu. int/ rec/ T-REC-T. 809
[33] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=40025
[34] ISO. "ISO/IEC 15444-11:2007 - Information technology -- JPEG 2000 image coding system: Wireless" (http:/ / www. iso. org/ iso/
iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=40025). . Retrieved 2009-11-01.
[35] http:/ / www. itu. int/ rec/ T-REC-T. 810
[36] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_ics/ catalogue_detail_ics. htm?csnumber=38612
[37] ISO. "ISO/IEC 15444-12:2008 - Information technology -- JPEG 2000 image coding system -- Part 12: ISO base media file format" (http:/ /
www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=51537). . Retrieved 2009-11-01.
[38] http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=42271
[39] ISO. "ISO/IEC 15444-13:2008 - Information technology -- JPEG 2000 image coding system: An entry level JPEG 2000 encoder" (http:/ /
www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=42271). . Retrieved 2009-11-01.
[40] http:/ / www. itu. int/ rec/ T-REC-T. 812
[41] ISO (2007-07-01). "ISO/IEC AWI 15444-14 - Information technology -- JPEG 2000 image coding system -- Part 14: XML structural
representation and reference" (http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=50410). . Retrieved
2009-11-01.
[42] ISO/IEC JTC 1/SC 29/WG 1 (2009-01-23). "Resolutions of 47th WG1 San Francisco Meeting" (http:/ / www. itscj. ipsj. or. jp/ sc29/ open/
29view/ 29n100261. doc) (DOC). . Retrieved 2009-11-01.
[43] "Resolutions of 41st WG1 San Jose Meeting" (http:/ / www. itscj. ipsj. or. jp/ sc29/ open/ 29view/ 29n83811. doc) (DOC). 2007-04-27. .
Retrieved 2009-11-01.
[44] Halbach, Till (July 2002). "Performance comparison: H.26L intra coding vs. JPEG2000" (http:/ / etill. net/ papers/ jvt-d039. pdf). . Retrieved
2008-04-22.
[45] wgrib2 home page (http:/ / www. cpc. ncep. noaa. gov/ products/ wesley/ wgrib2/ )
[46] JPEG 2000 Concerning recent patent claims (http:/ / www. jpeg. org/ newsrel1. html)
[47] JPEG 2000 Committee Drafts (http:/ / www. jpeg. org/ jpeg2000/ CDs15444. html)
[48] International Organization for Standardization (2004). "ISO/IEC 15444-2:2004, Information technology -- JPEG 2000 image coding system:
Extensions" (http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=33160). . Retrieved 2009-06-11.
[49] International Organization for Standardization (2003). "ISO/IEC 15444-6:2003, Information technology -- JPEG 2000 image coding system
-- Part 6: Compound image file format" (http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=35458). .
Retrieved 2009-06-11.
[50] International Organization for Standardization (2005). "ISO/IEC 15444-9:2005, Information technology -- JPEG 2000 image coding system:
Interactivity tools, APIs and protocols" (http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=39413). .
Retrieved 2009-06-11.
[51] "T.802 : Information technology - JPEG 2000 image coding system: Motion JPEG 2000" (http:/ / www. itu. int/ rec/ T-REC-T. 802/ en).
2005-01. . Retrieved 2009-11-01.
JPEG 2000 51

[52] International Organization for Standardization (2007). "ISO/IEC 15444-3:2007, Information technology -- JPEG 2000 image coding system:
Motion JPEG 2000" (http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=41570). . Retrieved
2009-06-11.
[53] JPEG (2007). "Motion JPEG 2000 (Part 3)" (http:/ / www. jpeg. org/ jpeg2000/ j2kpart3. html). . Retrieved 2009-11-01.
[54] Motion JPEG 2000 (Part 3) (http:/ / www. jpeg. org/ jpeg2000/ j2kpart3. html)
[55] Motion JPEG 2000 mj2 File Format (http:/ / www. digitalpreservation. gov/ formats/ fdd/ fdd000127. shtml). Sustainability of Digital
Formats Planning for Library of Congress Collections.
[56] ISO (2006-04). ISO Base Media File Format white paper - Proposal (http:/ / web. archive. org/ web/ 20080714101745/ http:/ / www.
chiariglione. org/ mpeg/ technologies/ mp04-ff/ index. htm). archive.org. . Retrieved 2009-12-26.
[57] ISO (2005-10). MPEG-4 File Formats white paper - Proposal (http:/ / web. archive. org/ web/ 20080115035235/ http:/ / www. chiariglione.
org/ mpeg/ technologies/ mp04-ff/ index. htm). archive.org. . Retrieved 2009-12-26.
[58] ISO (2009-10). ISO Base Media File Format white paper - Proposal (http:/ / mpeg. chiariglione. org/ technologies/ mpeg-4/ mp04-ff/ index.
htm). chiariglione.org. . Retrieved 2009-12-26.
[59] International Organization for Standardization (2004). "ISO/IEC 14496-12:2004, Information technology -- Coding of audio-visual objects
-- Part 12: ISO base media file format" (http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail. htm?csnumber=38539). .
Retrieved 2009-06-11.
[60] International Organization for Standardization (2008). "ISO/IEC 15444-12:2008, Information technology -- JPEG 2000 image coding
system -- Part 12: ISO base media file format" (http:/ / www. iso. org/ iso/ iso_catalogue/ catalogue_tc/ catalogue_detail.
htm?csnumber=51537). . Retrieved 2009-06-11.
[61] Open Geospatial Consortium GMLJP2 Home Page (http:/ / www. opengeospatial. org/ standards/ gmljp2)
[62] basic and advanced support refer to conformance with, respectively, Part1 and Part2 of the JPEG 2000 Standard.
[63] http:/ / www. accusoft. com/ index. htm
[64] Adobe Photoshop CS2 and CS3's official JPEG 2000 plug-in package is not installed by default and must be manually copied from the
install disk/folder to the Plug-Ins > File Formats folder.
[65] "Blender 2.49" (http:/ / www. blender. org/ development/ release-logs/ blender-249/ ). 2009-05-30. . Retrieved 2010-01-20.
[66] http:/ / www. grrsystems. com/ DBGallery
[67] "The digiKam Handbook - Supported File Formats" (http:/ / docs. kde. org/ development/ en/ extragear-graphics/ digikam/
using-fileformatsupport. html). docs.kde.org. . Retrieved 2010-01-20.
[68] "The Showfoto Handbook - Supported File Formats" (http:/ / docs. kde. org/ development/ en/ extragear-graphics/ showfoto/
using-fileformatsupport. html). . Retrieved 2010-01-20.
[69] "Development/Architecture/KDE3/Imaging and Animation" (http:/ / techbase. kde. org/ Development/ Architecture/ KDE3/
Imaging_and_Animation). . Retrieved 2010-01-20.
[70] Mozilla support for JPEG 2000 was requested in April 2000, but the report was closed as WONTFIX in August 2009. (https:/ / bugzilla.
mozilla. org/ show_bug. cgi?id=36351)
[71] The GIMP Team (2009-08-16). "GIMP 2.7 RELEASE NOTES" (http:/ / www. gimp. org/ release-notes/ gimp-2. 7. html). . Retrieved
2009-11-17.
[72] http:/ / www. imagemagick. org/ script/ license. php
[73] IrfanView's official plug-in package supports reading of .jp2 files but writing is quite limited until plug-in is purchased separately.
[74] XnView supports JPEG 2000 compression only in MS Windows version (http:/ / www. xnview. com/ en/ formats. html)

Libraries

Library support for JPEG 2000

Program Basic Advanced Language License

Read Write Read Write

C++ Proprietary
AccuRad J2KSuite (http:/ / www. aware. com/ imaging/ Yes Yes Yes Yes
accuradj2k. htm)

C LGPL
in progress in progress
FFmpeg [1] [1] ? ?

GTK (from 2.14) Yes No ? No C/GTK LGPL

JasPer Yes Yes No No C MIT License-style

Java Special (similar to

JJ2000 (http:/ / code. google. com/ p/ jj2000/ ) Yes Yes No No
BSD-style)
JPEG 2000 52

Yes Yes Yes Yes C++ Proprietary

Kakadu (http:/ / www. kakadusoftware. com/ )

C, C++, Proprietary
LEADTOOLS (http:/ / www. leadtools. com/ sdk/ Yes Yes Yes Yes .NET
compression/ jpeg2000. htm)

OpenJPEG Yes Yes ? ? C BSD

[1] not merged to main repository, only in soc repository (http:/ / svn. mplayerhq. hu/ soc/ jpeg2000/ )

See also
Comparison of graphics file formats
DjVu a compression format that also uses wavelets and that is designed for use on the web.
ECW a wavelet compression format that compares well to JPEG 2000.
QuickTime a multimedia framework, application and web browser plugin developed by Apple, capable of
encoding, decoding and playing various multimedia files (including JPEG 2000 images by default).
MrSID a wavelet compression format that compares well to JPEG 2000
PGF a fast wavelet compression format that compares well to JPEG 2000
JPIP JPEG 2000 Interactive Protocol

References
Official JPEG 2000 page (http://www.jpeg.org/JPEG2000.htm)
Final Committee Drafts of JPEG 2000 standard (http://www.jpeg.org/jpeg2000/CDs15444.html) (as the
official JPEG 2000 standard is not freely available, the final drafts are the most accurate freely available
documentation about this standard)
Gormish Notes on JPEG 2000 (http://www.crc.ricoh.com/~gormish/jpeg2000.html)
Technical overview of JPEG 2000 (http://www.rii.ricoh.com/~gormish/pdf/dcc2000_jpeg2000_note.pdf)
(PDF)
Everything you always wanted to know about JPEG 2000 - published by intoPIX in 2008 (http://www.intopix.
com/pdf/JPEG 2000 Handbook.pdf) (PDF)

External links
RFC 3745, MIME Type Registrations for JPEG 2000 (ISO/IEC 15444)
Official JPEG 2000 website (http://www.jpeg.org/JPEG2000.htm)
All published books about JPEG 2000 (http://www.watermarkingworld.com/books_list?k=Jpeg2000)
JPEG2000 Solutions: LuraTech (http://www.luratech.com)
JPEG2000 Compression for Medical Imaging: Aware, Inc. (http://www.aware.com/imaging/accuradj2k.htm)
JPEG 2000 comparisons
Side-by side comparison of appearance of 16k JPEG and JPEG 2000 files (http://www.fnordware.com/j2k/
jp2samples.html)
JPEG and JPEG 2000 Artifacts (http://kt.ijs.si/aleks/jpeg/artifacts.htm)
 53

Operations on images

Linear filter
A linear filter applies a linear operator to a time-varying input signal. Linear filters are very common in electronics
and digital signal processing (see the article on electronic filters), but they can also be found in mechanical
engineering and other technologies.
They are often used to eliminate unwanted frequencies from an input signal or to select a desired frequency among
many others. There are a wide range of types of filters and filter technologies, of which this article will present an
overview.
Regardless of whether they are electronic, electrical, or mechanical, or what frequency ranges or timescales they
work on, the mathematical theory of linear filters is universal.

Classification by transfer function

Impulse response
Linear filters can be divided into two classes: infinite impulse response (IIR) and finite impulse response (FIR)
filters.
An FIR filter (which may only be implemented in discrete time) may be described as a weighted sum of delayed
inputs. For such a filter, if the input becomes zero at any time, then the output will eventually become zero as
well, as soon as enough time has passed so that all the delayed inputs are zero, too. Therefore, the impulse
response lasts only a finite time, and this is the reason for the name finite impulse response. A discrete-time
transfer function of such a filter contains only poles at the origin (i.e., delays) and zeros; it cannot have off-origin
poles.
For an IIR filter, by contrast, if the input is set to 0 and the initial conditions are non-zero, then the set of time
where the output is non-zero will be unbounded; the filter's energy will decay but will be ever present. Therefore,
the impulse response extends to infinity, and the filter is said to have an infinite impulse response. There are no
special restrictions on the transfer function of an IIR filter; it can have arbitrary poles and zeros, and it need not be
expressible as a rational transfer function (for example, a sinc filter).
Until about the 1970s, only analog IIR filters were practical to construct. The distinction between FIR and IIR filters
is generally applied only in the discrete-time domain. Because digital systems necessarily have discrete-time
domains, both FIR and IIR filters are straightforward to implement digitally. Analog FIR filters can be built with
analog delay lines.
Linear filter 54

Frequency response
Here is an image comparing the Fourier transform (i.e., "frequency response") of several popular continuous-time
IIR filters: Butterworth, Chebyshev, and elliptic filters. The filters in this illustration are all fifth-order low-pass
filters.

As is clear from the image, elliptic filters are sharper than the others, but they show ripples in their passband.
There are several common kinds of linear filters:
A low-pass filter passes low frequencies.
A high-pass filter passes high frequencies.
A band-pass filter passes a limited range of frequencies.
A band-stop filter passes all frequencies except a limited range.
An all-pass filter passes all frequencies, but alters the phase relationship among them.
A notch filter is a specific type of band-stop filter that acts on a particularly narrow range of frequencies.
Some filters are not designed to stop any frequencies, but instead to gently vary the amplitude response at
different frequencies: filters used as pre-emphasis filters, equalizers, or tone controls are good examples of this
Band-stop and bandpass filters can be constructed by combining low-pass and high-pass filters. A popular form of 2
pole filter is the Sallen-Key type. This is able to provide low-pass, band-pass, and high pass versions. A particular
bandform of filter can be obtained by transformation of a prototype filter of that class.
Linear filter 55

Mathematics of filter design

Linear analog electronic filters

[1]

LTI system theory describes linear time-invariant (LTI) filters of all types. LTI filters can be completely described
by their frequency response and phase response, the specification of which uniquely defines their impulse response,
and vice versa. From a mathematical viewpoint, continuous-time IIR LTI filters may be described in terms of linear
differential equations, and their impulse responses considered as Green's functions of the equation. Continuous-time
LTI filters may also be described in terms of the Laplace transform of their impulse response, which allows all of the
characteristics of the filter to be analyzed by considering the pattern of poles and zeros of their Laplace transform in
the complex plane. Similarly, discrete-time LTI filters may be analyzed via the Z-transform of their impulse
response.
Before the advent of computer filter synthesis tools, graphical tools such as Bode plots and Nyquist plots were
extensively used as design tools. Even today, they are invaluable tools to understanding filter behavior. Reference
books[2] had extensive plots of frequency response, phase response, group delay, and impulse response for various
types of filters, of various orders. They also contained tables of values showing how to implement such filters as
RLC ladders - very useful when amplifying elements were expensive compared to passive components. Such a
ladder can also be designed to have minimal sensitivity to component variation[3] a property hard to evaluate without
computer tools.
Many different analog filter designs have been developed, each trying to optimise some feature of the system
response. For practical filters, a custom design is sometimes desirable, that can offer the best tradeoff between
different design criteria, which may include component count and cost, as well as filter response characteristics.
These descriptions refer to the mathematical properties of the filter (that is, the frequency and phase response). These
can be implemented as analog circuits (for instance, using a Sallen Key filter topology, a type of active filter), or as
algorithms in digital signal processing systems.
Digital filters are much more flexible to synthesize and use than analog filters, where the constraints of the design
permits their use. Notably, there is no need to consider component tolerances, and very high Q levels may be
obtained.
FIR digital filters may be implemented by the direct convolution of the desired impulse response with the input
signal. They can easily be designed to give a matched filter for any arbitrary pulse shape.
IIR digital filters are often more difficult to design, due to problems including dynamic range issues, quantization
noise and instability. Typically digital IIR filters are designed as a series of digital biquad filters.
All low-pass second-order continuous-time filters have a transfer function given by

All band-pass second-order continuous-time have a transfer function given by

where
K is the gain (low-pass DC gain, or band-pass mid-band gain) (K is 1 for passive filters)
Q is the Q factor
is the center frequency
is the complex frequency
Linear filter 56

See also
Filter design
Laplace transform
Green's function
Prototype filter
Z-transform
System theory
LTI system theory
Nonlinear filter
Wiener filter
Gabor filter

Further reading
Williams, Arthur B & Taylor, Fred J (1995). Electronic Filter Design Handbook. McGraw-Hill.
ISBN0-07-070441-4.
National Semiconductor AN-779 [4] application note describing analog filter theory
Lattice AN6017 [5] application note comparing and contrasting filters (in order of damping coefficient, from
lower to higher values): Gaussian, Bessel, linear phase, Butterworth, Chebyshev, Legendre, elliptic. (with
graphs).
"Design and Analysis of Analog Filters: A Signal Processing Perspective" [6] by L. D. Paarmann

References
[1] http:/ / en. wikipedia. org/ wiki/ Template:Linear
[2] A. Zverev, Handbook of Filter Synthesis, John Wiley and Sons, 1967, ISBN 0-471-98680-1
[3] Normally, computing sensitivities is a very laborious operation. But in the special case of an LC ladder driven by an impedance and
terminated by a resistor, there is a neat argument showing the sensitivities are small. In such as case, the transmission at the maximum
frequency(s) transfers the maximal possible energy to the output load, as determined by the physics of the source and load impedances. Since
this point is a maximum, all derivatives with respect to all component values must be zero, since the result of changing any component value
in any direction can only result in a reduction. This result only strictly holds true at the peaks of the response, but is roughly true at nearby
points as well.
[4] http:/ / www. national. com/ an/ AN/ AN-779. pdf
[5] http:/ / www. latticesemi. com/ lit/ docs/ appnotes/ pac/ an6017. pdf
[6] http:/ / books. google. com/ books?id=l7oC-LJwyegC& pg=PA267& lpg=PA267& dq=%22legendre+ filter%22& source=web&
ots=xRLtCLfslz& sig=0Nw2zhb8Y7FSrylN3wDaoIMkekQ#PPA238,M1
Histogram 57

Histogram
Histogram

One of the Seven Basic Tools of Quality

First described by Karl Pearson

Purpose To roughly assess the probability distribution of a given variable by depicting the frequencies of observations
occurring in certain ranges of values

In statistics, a histogram is a graphical display of tabular frequencies, shown as adjacent rectangles. Each rectangle
is erected over an interval, with an area equal to the frequency of the interval. The height of a rectangle is also equal
to the frequency density of the interval, i.e. the frequency divided by the width of the interval. The total area of the
histogram is equal to the number of data. A histogram may also be based on the relative frequencies instead. It then
shows what proportion of cases fall into each of several categories (a form of data binning), and the total area then
equals 1. The categories are usually specified as consecutive, non-overlapping intervals of some variable. The
categories (intervals) must be adjacent, and often are chosen to be of the same size,[1] but not necessarily so.
Histograms are used to plot density of data, and often for density estimation: estimating the probability density
function of the underlying variable. The total area of a histogram used for probability density is always normalized to
1. If the length of the intervals on the x-axis are all 1, then a histogram is identical to a relative frequency plot.
An alternative to the histogram is kernel density estimation, which uses a kernel to smooth samples. This will
construct a smooth probability density function, which will in general more accurately reflect the underlying
variable.
The histogram is one of the seven basic tools of quality control.[2]
Histogram 58

Etymology
The etymology of the word histogram is uncertain. Sometimes it is
said to be derived from the Greek histos 'anything set upright' (as the
masts of a ship, the bar of a loom, or the vertical bars of a histogram);
and gramma 'drawing, record, writing'. It is also said that Karl Pearson,
who introduced the term in 1895, derived the name from "historical
diagram".
[3]

Examples
As an example we consider data collected by the U.S. Census Bureau
An example histogram of the heights of 31 Black on time to travel to work (2000 census, [4], Table 2). The census found
Cherry trees. that there were 124 million people who work outside of their homes.
An interesting feature of this graph is that the number recorded for "at
least 30 but less than 35 minutes" is higher than for the bands on either side. This is likely to have arisen from people
rounding their reported journey time. This rounding is a common phenomenon when collecting data from people.

Histogram of travel time, US 2000 census. Area under the curve equals the total number
of cases. This diagram uses Q/width from the table.

Data by absolute numbers

Interval Width Quantity Quantity/width

0 5 4180 836

5 5 13687 2737

10 5 18618 3723

15 5 19634 3926

20 5 17981 3596

25 5 7190 1438

30 5 16369 3273

35 5 3212 642
Histogram 59

40 5 4122 824

45 15 9200 613

60 30 6461 215

90 60 3435 57

This histogram shows the number of cases per unit interval so that the height of each bar is equal to the proportion of
total people in the survey who fall into that category. The area under the curve represents the total number of cases
(124 million). This type of histogram shows absolute numbers.

Histogram of travel time, US 2000 census. Area under the curve equals 1. This diagram
uses Q/total/width from the table.

Data by proportion
Interval Width Quantity Q/total/width
(Q)

0 5 4180 0.0067

5 5 13687 0.0221

10 5 18618 0.0300

15 5 19634 0.0316

20 5 17981 0.0290

25 5 7190 0.0116

30 5 16369 0.0264

35 5 3212 0.0052

40 5 4122 0.0066

45 15 9200 0.0049

60 30 6461 0.0017

90 60 3435 0.0005

This histogram differs from the first only in the vertical scale. The height of each bar is the decimal percentage of the
total that each category represents, and the total area of all the bars is equal to 1, the decimal equivalent of 100%.
The curve displayed is a simple density estimate. This version shows proportions, and is also known as a unit area
histogram.
Histogram 60

In other words a histogram represents a frequency distribution by means of rectangles whose widths represent class
intervals and whose areas are proportional to the corresponding frequencies. The intervals are placed together in
order to show that the data represented by the histogram while being exclusive is also continuous. (E.g., in a
histogram it is possible to have two connnecting intervals of 10.5-20.5 and 20.5-33.5, but not two connecting
intervals of 10.5-20.5 and 22.5-32.5. Empty intervals are represented as empty and not skipped.)[5]

Activities and demonstrations

The SOCR resource pages contain a number of hands-on interactive activities demonstrating the concept of a
histogram, histogram construction [6] and manipulation [7] using Java applets and charts [8].

Mathematical definition
In a more general mathematical sense,
a histogram is a mapping mi that
counts the number of observations that
fall into various disjoint categories
(known as bins), whereas the graph of
a histogram is merely one way to
represent a histogram. Thus, if we let n
be the total number of observations
and k be the total number of bins, the
histogram mi meets the following
conditions: An ordinary and a cumulative histogram of the same data. The data shown is a random
sample of 10,000 points from a normal distribution with a mean of 0 and a standard
deviation of 1.

Cumulative histogram
A cumulative histogram is a mapping that counts the cumulative number of observations in all of the bins up to the
specified bin. That is, the cumulative histogram Mi of a histogram mj is defined as:

Number of bins and width

There is no "best" number of bins, and different bin sizes can reveal different features of the data. Some theoreticians
have attempted to determine an optimal number of bins, but these methods generally make strong assumptions about
the shape of the distribution. You should always experiment with bin widths before choosing one (or more) that
illustrate the salient features in your data.
The number of bins k can be assigned directly or can be calculated from a suggested bin widthh as:

The braces indicate the ceiling function.

Sturges' formula[9]
Histogram 61

which implicitly bases the bin sizes on the range of the data, and can perform poorly ifn<30.
Scott's choice[10]

where is the sample standard deviation.

Square-Root Choice

which takes the square root of the number of data points in the sample (used by Excel histograms and many others)
FreedmanDiaconis' choice[11]

which is based on the interquartile range. A good discussion of this and other rules for choice of bin widths is in
Modern Applied Statistics with S, 5.6: Density Estimation.[12]

See also
Data binning
FreedmanDiaconis rule
Image histogram
Density estimation
Kernel density estimation, a smoother but more complex method of density estimation

Further reading
Lancaster, H.O. An Introduction to Medical Statistics. John Wiley and Sons. 1974. ISBN 0 471 51250-8

External links
Journey To Work and Place Of Work [13] (location of census document cited in example)
Understanding histograms in digital photography [14]
Histograms: Construction, Analysis and Understanding with external links and an application to particle Physics.
[15]

A Method for Selecting the Bin Size of a Histogram [16]

Interactive histogram generator [17]
Matlab function to plot nice histograms [18]
Histogram 62

References
[1] Howitt, D. and Cramer, D. (2008) "Statistics in Psychology". Prentice Hall
[2] Nancy R. Tague (2004). "Seven Basic Quality Tools" (http:/ / www. asq. org/ learn-about-quality/ seven-basic-quality-tools/ overview/
overview. html). The Quality Toolbox. Milwaukee, Wisconsin: American Society for Quality. p. 15. . Retrieved 2010-02-05.
[3] M. Eileen Magnello (December 1856). "Karl Pearson and the Origins of Modern Statistics: An Elastician becomes a Statistician" (http:/ /
www. rutherfordjournal. org/ article010107. html). The New Zealand Journal for the History and Philosophy of Science and Technology 1
volume. ISSN11771380. .
[4] http:/ / www. census. gov/ prod/ 2004pubs/ c2kbr-33. pdf
[5] Dean, S., & Illowsky, B. (2009, February 19). Descriptive Statistics: Histogram. Retrieved from the Connexions Web site: http:/ / cnx. org/
content/ m16298/ 1. 11/
[6] http:/ / wiki. stat. ucla. edu/ socr/ index. php/ SOCR_EduMaterials_ModelerActivities_MixtureModel_1
[7] http:/ / wiki. stat. ucla. edu/ socr/ index. php/ SOCR_EduMaterials_Activities_PowerTransformFamily_Graphs
[8] http:/ / www. socr. ucla. edu/ htmls/ SOCR_Charts. html
[9] Sturges, H. A. (1926). "The choice of a class interval". J. American Statistical Association: 6566.
[10] Scott, David W. (1979). "On optimal and data-based histograms". Biometrika 66 (3): 605610. doi:10.1093/biomet/66.3.605.
[11] Freedman, David; Diaconis, P. (1981). "On the histogram as a density estimator: L2 theory". Zeitschrift fr Wahrscheinlichkeitstheorie und
verwandte Gebiete 57 (4): 453476. doi:10.1007/BF01025868.
[12] W. N. Venables and B. D. Ripley: "Modern Applied Statistics with S", Springer, in (4thedition) section 5.6: Density Estimation
[13] http:/ / www. census. gov/ population/ www/ socdemo/ journey. html
[14] http:/ / www. luminous-landscape. com/ tutorials/ understanding-series/ understanding-histograms. shtml
[15] http:/ / quarknet. fnal. gov/ toolkits/ ati/ histograms. html
[16] http:/ / 2000. jukuin. keio. ac. jp/ shimazaki/ res/ histogram. html
[17] http:/ / www. shodor. org/ interactivate/ activities/ histogram/
[18] http:/ / www. mathworks. com/ matlabcentral/ fileexchange/ 27388-plot-and-compare-nice-histograms-by-default

Image histogram

Sunflower image

Histogram of Sunflower image

An image histogram is a type of histogram which acts as a graphical representation of the tonal distribution in a
digital image.[1] It plots the number of pixels for each tonal value. By looking at the histogram for a specific image, a
viewer will be able to judge the entire tonal distribution at a glance.
Image histograms are present on many modern digital cameras. Photographers can use them as an aid to show the
distribution of tones captured, and whether image detail has been lost to blown-out highlights or blacked-out
shadows.[2]
Image histogram 63

The horizontal axis of the graph represents the tonal variations, while the vertical axis represents the number of
pixels in that particular tone.[1] The left side of the horizontal axis represents the black and dark areas, the middle
represents medium grey, and the right hand side represents light and pure white areas. The vertical axis represents
the size of the area that is captured in each one of these zones.

Image manipulation and histograms

Image editors typically have provisions to create a histogram of the image being edited. The histogram plots the
number of pixels in the image (vertical axis) with a particular brightness value (horizontal axis). Algorithms in the
digital editor allow the user to visually adjust the brightness value of each pixel and to dynamically display the
results as adjustments are made.[3] Improvements in picture brightness and contrast can thus be obtained.

See also
Histogram
Image editing
Color histogram, a multidimensional histogram of the distribution of color in an image
Histogram equalization
Histogram matching

References
[1] Ed Sutton. "Histograms and the Zone System" (http:/ / www. illustratedphotography. com/ photography-tips/ basic/ contrast). Illustrated
Photography. .
[2] Michael Freeman (2005). The Digital SLR Handbook. Ilex. ISBN1904705-36-7.
[3] Martin Evening (2007). Adobe Photoshop CS3 for Photographers: A Professional Image Editor's Guide... (http:/ / books. google. com/
books?id=0XZO93ZcbqkC& pg=PA136& dq="image+ histogram"& lr=& as_brr=0& ei=dvL2R--yFI2AsgPP7Z2FCg&
sig=UoCi37U--m3IUzm6fUG-YB3za_Y#PPA136,M1). Focal Press. ISBN0240520289. .
Color histogram 64

Color histogram
In image processing and photography, a color histogram is a representation of the distribution of colors in an image.
For digital images, it is basically the number of pixels that have colors in each of a fixed list of color ranges, that
span the image's color space, the set of all possible colors.
The color histogram can be built for any kind of color space, although the term is more often used for
three-dimensional spaces like RGB or HSV. For monochromatic images, the term intensity histogram may be used
instead. For multi-spectral images, where each pixel is represented by a Nof measurements, each within its own
wavelength range of the light spectrum, some of which may be outside the visible spectrum, the colour histogram is
N-dimensional.
If the set of possible color values is sufficiently small, each of those colors may be placed on a range by itself; then
the histogram is merely the count of pixels that have each possible color. Most often, the space is divided into an
appropriate number of ranges, often arranged as a regular grid, each containing many similar color values. The color
histogram may also be represented and/or displayed as a smooth function defined over the color space, that
approximates the pixel counts.
Like other kinds of histograms, the color histogram is a statistic that can be viewed as an approximation of an
underlying continuous distribution of colors values.

Overview
Color histograms are flexible constructs that can be built from images in various color spaces, whether RGB, rg
chromaticity or any other color space of any dimension. A histogram of an image is produced first by discretization
of the colors in the image into a number of bins, and counting the number of image pixels in each bin. For example,
a RedBlue chromaticity histogram can be formed by first normalizing color pixel values by dividing RGB values
by R+G+B, then quantizing the normalized R and B coordinates into N bins each; say N = 4, which might yield a 2D
histogram that looks like this table:

red

0-63 64-127 128-191 192-255

blue 0-63 43 78 18 0

64-127 45 67 33 2

128-191 127 58 25 8

192-255 140 47 47 13

Similarly a histogram can be made three-dimensional, though it is harder to display.[1]

The histogram provides a compact summarization of the distribution of data in an image. The color histogram of an
image is relatively invariant with translation and rotation about the viewing axis, and varies only slowly with the
angle of view.[2] By comparing histograms signatures of two images and matching the color content of one image
with the other, the color histogram is particularly well suited for the problem of recognizing an object of unknown
position and rotation within a scene. Importantly, translation of an RGB image into the illumination invariant
rg-chromaticity space allows the histogram to operate well in varying light levels.
The main drawback of histograms for classification is that the representation is dependent of the color of the object
being studied, ignoring its shape and texture. Color histograms can potentially be identical for two images with
different object content which happens to share color information. Conversely, without spatial or shape information,
similar objects of different color may be indistinguishable based solely on color histogram comparisons. There is no
way to distinguish a red and white cup from a red and white plate. Put another way, histogram-based algorithms have
Color histogram 65

no concept of a generic 'cup', and a model of a red and white cup is no use when given an otherwise identical blue
and white cup. Another problem is that color histograms have high sensitivity to noisy interference such as lighting
intensity changes and quantization errors. High dimensionality(bins) of color histograms are also another issue.
Some color histogram feature spaces often occupy more than one hundred dimensions[8].
Some of the proposed solutions have been color histogram intersection, color constant indexing, cumulative color
histogram, quadratic distance, and last but not least color correlograms [8]. Check out the external link to Standford
for in depth look at the equations.
Although there are drawbacks of using histograms for indexing/classifications, using color in a real-time system has
several relative advantages. One is that color information is faster to compute, compared to other "invariants." It has
been shown in some cases that color can a be an efficient method for identifying objects of known location and
appearances (refer to external link for findings in study)[8].
Further research into the relationship between color histograms data to the physical properties of the objects in an
image has shown they can represent not only object color and illumination but relate to surface roughness and image
geometry and provide improved estimate of illumination and object color.[3]
Usually Euclidean distance, histogram intersection, or cosine or quadratic distances are used for the calculation
of the images similarity rating.[4] . Any of these values does not reflect the similarity rate of two images in itself. It
is useful only with comparison to other similar values. This is the reason that all the practical implementations of
content-based image retrieval must complete computation of all images from the database. It is the main
disadvantage of these implementations.
Other approach to representative color image content is 2D-color histogram. 2D-color histogram considers the
relation between the pixel pair colors (not only the lighting component).[5] 2D-color histogram is a two-dimensional
array, Cmax*Cmax, where Cmax is the number of colors that was used in the phase of color quantization. These
arrays are treated as matrices, each element of which stores a normalized count of pixel pairs, with each color
corresponding to the index of an element in each pixel neighbourhood. For comparison of 2D-color histograms it is
suggested calculating their correlation, because a 2D-color histogram, constructed as described above, is a random
vector (in other words, a multidimensional random value). While creating a set of final images, the images should be
arranged in decreasing order of the correlation coefficient. Correlation coefficient may be used also for color
histograms comparison. Retrieval results with correlation coefficient are better than with other metrics.[6]
==

Intensity histogram of continuous data

The idea of an intensity histogram can be generalized to continuous data, say audio signals represented by real
functions or images represented by functions with two-dimensional domain.
Let (see Lebesgue space), then the cumulative histogram operator can be defined by:
.
is the Lebesgue measure of sets. in turn is a real function. The (non-cumulative) histogram is defined as
its derivative.
.
Color histogram 66

External links
3D Color Inspector/Color Histogram [7], by Kai Uwe Barthel [8]. (Free Java applet.)
QBIC Image Retrieval [9], by State Hermitage Museum
Standford Student Project on Image Based Retrieval [10] - more in depth look at equations/application
MATLAB/Octave code for plotting Color Histograms and Color Clouds [11] - The source code can be ported to
other languages

References
[1] Nello Zuech and Richard K. Miller (1989). Machine Vision. Springer. ISBN0442237375.
[2] Shapiro, Linda G. and Stockman, George C. "Computer Vision" Prentice Hall, 2003 ISBN 0130307963
[3] Anatomy of a color histogram; Novak, C.L.; Shafer, S.A.; Computer Vision and Pattern Recognition, 1992. Proceedings CVPR '92., 1992
IEEE Computer Society Conference on 15-18 June 1992 Page(s):599 - 605 doi:10.1109/CVPR.1992.223129
[4] Integrated Spatial and Feature Image Systems: Retrieval, Analysis and Compression; Smith, J.R.; Graduate School of Arts and Sciences,
Columbia University, 1997
[5] Effectiveness estimation of image retrieval by 2D color histogram; Bashkov, E.A.; Kostyukova, N.S.; Jornal of Automation and Information
Sciences, 2006 (6) Page(s): 84-89
[6] Content-Based Image Retrieval Using Color Histogram Correlation; Bashkov, E.A.; Shozda, N.S.; Graphicon proceedings, 2002 Page(s):
458-461 (http:/ / www. graphicon. ru/ 2002/ pdf/ Bashkov_Zhozda_Re. pdf)
[7] http:/ / rsb. info. nih. gov/ ij/ plugins/ color-inspector. html
[8] http:/ / www. f4. fhtw-berlin. de/ ~barthel/
[9] http:/ / www. hermitagemuseum. org/ fcgi-bin/ db2www/ qbicSearch. mac/ qbic?selLang=English
[10] http:/ / scien. stanford. edu/ class/ psych221/ projects/ 02/ sojeong/
[11] http:/ / www. mathworks. com/ matlabcentral/ fileexchange/ 20757

Affine transformation
In geometry, an affine transformation or affine map or an
affinity (from the Latin, affinis, "connected with") between two
vector spaces (strictly speaking, two affine spaces) consists of a
linear transformation followed by a translation:

An image of a fern which exhibits affine self-similarity

Affine transformation 67

In the finite-dimensional case each affine transformation is given by a matrix A and a vector b, satisfying certain
properties described below.
Geometrically, an affine transformation in Euclidean space is one that preserves
1. The collinearity relation between points; i.e., three points which lie on a line continue to be collinear after the
transformation
2. Ratios of distances along a line; i.e., for distinct collinear points , , , the ratio
is preserved
In general, an affine transformation is composed of linear transformations (rotation, scaling or shear) and a
translation (or "shift"). Several linear transformations can be combined into a single one, so that the general formula
given above is still applicable.
In the one-dimensional case, A and b are called, respectively, slope and intercept.

Representation
Ordinary vector algebra uses matrix multiplication to represent linear transformations, and vector addition to
represent translations. Using an augmented matrix, it is possible to represent both using matrix multiplication. The
technique requires that all vectors are augmented with a "1" at the end, and all matrices are augmented with an extra
row of zeros at the bottom, an extra columnthe translation vectorto the right, and a "1" in the lower right corner.
If A is a matrix,

is equivalent to the following

This representation exhibits the set of all invertible affine transformations as the semidirect product of Kn and GL(n,
k). This is a group under the operation of composition of functions, called the affine group.
Ordinary matrix-vector multiplication always maps the origin to the origin, and could therefore never represent a
translation, in which the origin must necessarily be mapped to some other point. By appending a "1" to every vector,
one essentially considers the space to be mapped as a subset of a space with an additional dimension. In that space,
the original space occupies the subset in which the final index is 1. Thus the origin of the original space can be found
at (0,0, ... 0, 1). A translation within the original space by means of a linear transformation of the higher-dimensional
space is then possible (specifically, a shear transformation). This is an example of homogeneous coordinates.
The advantage of using homogeneous coordinates is that one can combine any number of affine transformations into
one by multiplying the matrices. This device is used extensively by graphics software.

Properties
An affine transformation is invertible if and only if A is invertible. In the matrix representation, the inverse is:

The invertible affine transformations form the affine group, which has the general linear group of degree n as
subgroup and is itself a subgroup of the general linear group of degree n + 1.
The similarity transformations form the subgroup where A is a scalar times an orthogonal matrix. If and only if the
determinant of A is 1 or 1 then the transformation preserves area; these also form a subgroup. Combining both
conditions we have the isometries, the subgroup of both where A is an orthogonal matrix.
Affine transformation 68

Each of these groups has a subgroup of transformations which preserve orientation: those where the determinant of A
is positive. In the last case this is in 3D the group of rigid body motions (proper rotations and pure translations).
For any matrix A the following propositions are equivalent:
A I is invertible
A does not have an eigenvalue equal to 1
for all b the transformation has exactly one fixed point
there is a b for which the transformation has exactly one fixed point
affine transformations with matrix A can be written as a linear transformation with some point as origin
If there is a fixed point, we can take that as the origin, and the affine transformation reduces to a linear
transformation. This may make it easier to classify and understand the transformation. For example, describing a
transformation as a rotation by a certain angle with respect to a certain axis is easier to get an idea of the overall
behavior of the transformation than describing it as a combination of a translation and a rotation. However, this
depends on application and context. Describing such a transformation for an object tends to make more sense in
terms of rotation about an axis through the center of that object, combined with a translation, rather than by just a
rotation with respect to some distant point. As an example: "move 200 m north and rotate 90 anti-clockwise", rather
than the equivalent "with respect to the point 141 m to the northwest, rotate 90 anti-clockwise".
Affine transformations in 2D without fixed point (so where A has eigenvalue 1) are:
pure translations
scaling in a given direction, with respect to a line in another direction (not necessarily perpendicular), combined
with translation that is not purely in the direction of scaling; the scale factor is the other eigenvalue; taking
"scaling" in a generalized sense it includes the cases that the scale factor is zero (projection) and negative; the
latter includes reflection, and combined with translation it includes glide reflection.
shear combined with translation that is not purely in the direction of the shear (there is no other eigenvalue than 1;
it has algebraic multiplicity 2, but geometric multiplicity 1)

Affine transformation of the plane

To visualise the general affine transformation of the Euclidean plane, take labelled parallelograms ABCD and
ABCD. Whatever the choices of points, there is an affine transformation T of the plane taking A to A, and each
vertex similarly. Supposing we exclude the degenerate case where ABCD has zero area, there is a unique such affine
transformation T. Drawing out a whole grid of parallelograms based on ABCD, the image T(P) of any point P is
determined by noting that T(A) = A, T applied to the line segment AB is AB, T applied to the line segment AC is
AC, and T respects scalar multiples of vectors based at A. [If A, E, F are collinear then the ratio
length(AF)/length(AE) is equal to length(AF)/length(AE).] Geometrically T transforms the grid based on ABCD to
that based in ABCD.
Affine transformations don't respect lengths or angles; they multiply area by a constant factor
area of A B C D / area of ABCD.
A given T may either be direct (respect orientation), or indirect (reverse orientation), and this may be determined by
its effect on signed areas (as defined, for example, by the cross product of vectors).
Affine transformation 69

Example of an affine transformation

The following equation expresses an affine transformation in GF(2) (with "+" representing XOR):

where [M] is the matrix

and {v} is the vector

For instance, the affine transformation of the element {a} = x7 + x6 + x3 + x = {11001010} in big-endian binary
notation = {CA} in big-endian hexadecimal notation, is calculated as follows:

Thus, {a} = x7 + x6 + x5 + x3 + x2 + 1 = {11101101} = {ED}.

See also
The transformation matrix for an affine transformation
Affine geometry
Homothetic transformation
Similarity transformation
Linear transformation (the second meaning is affine transformation in 1D)
3D projection
Flat (geometry)
Affine transformation 70

External links
Geometric Operations: Affine Transform [1], R. Fisher, S. Perkins, A. Walker and E. Wolfart.
Weisstein, Eric W., "Affine Transformation [2]" from MathWorld.
Affine Transform [3] by Bernard Vuilleumier, Wolfram Demonstrations Project.
Affine Transformation on PlanetMath [4]

References
[1] http:/ / homepages. inf. ed. ac. uk/ rbf/ HIPR2/ affine. htm
[2] http:/ / mathworld. wolfram. com/ AffineTransformation. html
[3] http:/ / demonstrations. wolfram. com/ AffineTransform/
[4] http:/ / planetmath. org/ encyclopedia/ AffineTransformation. html

Scaling (geometry)
In Euclidean geometry, uniform scaling or isotropic scaling[1] is a linear transformation that enlarges or increases
or diminishes objects; the scale factor is the same in all directions; it is also called a homothety. The result of
uniform scaling is similar (in the geometric sense) to the original. A scale factor of 1 is normally allowed, so that
congruent shapes are also classed as similar, but some school text books specifically exclude this possibility.
More general is scaling with a separate scale factor for each axis direction. Non-uniform or anisotropic scaling is
obtained when at least one of the scaling factors is different from the others; a special case is directional scaling or
stretching (in one direction). Non-uniform scaling changes the shape of the object; e.g. a square may change into a
rectangle, or into a parallelogram if the sides of the square are not parallel to the scaling axes (the angles between
lines parallel to the axes are preserved, but not all angles).

Matrix representation
A scaling can be represented by a scaling matrix. To scale an object by a vector v = (vx, vy, vz), each point p = (px, py,
pz) would need to be multiplied with this scaling matrix:

As shown below, the multiplication will give the expected result:

Such a scaling changes the diameter of an object by a factor between the scale factors, the area by a factor between
the smallest and the largest product of two scale factors, and the volume by the product of all three.
A scaling in the most general sense is any affine transformation with a diagonalizable matrix. It includes the case
that the three directions of scaling are not perpendicular. It includes also the case that one or more scale factors are
equal to zero (projection), and the case of one or more negative scale factors. The latter corresponds to a combination
of scaling proper and a kind of reflection: along lines in a particular direction we take the reflection in the point of
intersection with a plane that need not be perpendicular; therefore it is more general than ordinary reflection in the
plane.
Scaling (geometry) 71

Using homogeneous coordinates

Often, it is more useful to use homogeneous coordinates, since translation cannot be accomplished with a 3-by-3
matrix. To scale an object by a vector v = (vx, vy, vz), each homogeneous vector p = (px, py, pz, 1) would need to be
multiplied with this scaling matrix:

As shown below, the multiplication will give the expected result:

The scaling is uniform if and only if the scaling factors are equal. If all scale factors except one are 1 we have
directional scaling.
Since the last component of a homogeneous coordinate can be viewed as the denominator of the other three
components, a scaling by a common factor s can be accomplished by using this scaling matrix:

For each homogeneous vector p = (px, py, pz, 1) we would have

which would be homogenized to

See also
Scale (ratio)
Scale (map)
Scales of scale models
Scale (disambiguation)
Scaling in gravity
Transformation matrix
Scaling (geometry) 72

External links
Understanding 2D Scaling [2] and Understanding 3D Scaling [3] by Roger Germundsson, The Wolfram
Demonstrations Project.

References
[1] Durand and Cutler (n.d.). Transformations (http:/ / groups. csail. mit. edu/ graphics/ classes/ 6. 837/ F03/ lectures/ 04_transformations. ppt).
Massachusetts Institute of Technology. Retrieved 12 September 2008.
[2] http:/ / demonstrations. wolfram. com/ Understanding2DScaling/
[3] http:/ / demonstrations. wolfram. com/ Understanding3DScaling/

Rotation (mathematics)
In geometry and linear algebra, a rotation is a transformation in a
plane or in space that describes the motion of a rigid body around a
fixed point. A rotation is different from a translation, which has no
fixed points, and from a reflection, which "flips" the bodies it is
transforming. A rotation and the above-mentioned transformations are
isometries; they leave the distance between any two points unchanged
after the transformation.

It is important to know the frame of reference when considering

rotations, as all rotations are described relative to a particular frame of
reference. In general for any orthogonal transformation on a body in a
Rotation of an object in two dimensions around a
coordinate system there is an inverse transformation which if applied point
to the frame of reference results in the body being at the same
coordinates. For example in two dimensions rotating a body clockwise about a point keeping the axes fixed is
equivalent to rotating the axes counterclockwise about the same point while the body is kept fixed.

Two dimensions
Only a single angle is needed to specify a rotation in two dimensions
the angle of rotation. To calculate the rotation two methods can be
used, either matrix algebra or complex numbers. In each the rotation is
acting to rotate an object counterclockwise through an angle about
the origin.

Matrix algebra
To carry out a rotation using matrices the point (x, y) to be rotated is
written as a vector, then multiplied by a matrix calculated from the
angle, , like so:
A plane rotation around a point followed by
another rotation around a different point results in
a total motion which is either a rotation (as in this
picture), or a translation.
Rotation (mathematics) 73

A reflection against an axis followed by a

reflection against a second axis not parallel to the
first one results in a total motion that is a rotation
around the point of intersection of the axes.

where (x, y) are the co-ordinates of the point after rotation, and the formulae for x and y can be seen to be

The vectors and have the same magnitude and are separated by an angle as expected.

Complex numbers
Points can also be rotated using complex numbers, as the set of all such numbers, the complex plane, is
geometrically a two dimensional plane. the point (x, y) on the plane is represented by the complex number

This can be rotated through an angle by multiplying it by ei, then expanding the product using Euler's formula as
follows:

which gives the same result as before,

Like complex numbers rotations in two dimensions are commutative, unlike in higher dimensions. They have only
one degree of freedom, as such rotations are entirely determined by the angle of rotation.[1]
Rotation (mathematics) 74

Three dimensions
Rotations in ordinary three-dimensional space differ than those in two dimensions in a number of important ways.
Rotations in three dimensions are generally not commutative, so the order in which rotations are applied is
important. They have three degrees of freedom, the same as the number of dimensions
A three dimensional rotation can be specified in a number of ways. The most usual methods are as follows.

Matrix algebra
As in two dimensions a matrix can be used to rotate a point (x, y, z) to a point (x, y, z). The matrix used is a 33
matrix,

This is multiplied by a vector representing the point to give the result

The matrix A is a member of the three dimensional special orthogonal group, SO(3), that is it is an orthogonal matrix
with determinant 1. That it is an orthogonal matrix means that its rows are a set of orthogonal unit vectors (so they
are an orthonormal basis) as are its columns, making it easy to spot and check if a matrix is a valid rotation matrix.
The determinant must be 1 as if it is -1 (the only other possibility for an orthogonal matrix) then the transformation
given by it is a reflection, improper rotation or inversion in a point, i.e. not a rotation.
Matrices are often used for doing transformations, especially when a large number of points are being transformed,
as they are a direct representation of the linear operator. Rotations represented in other ways are often converted to
matrices before being used. They can be extended to represent rotations and transformations at the same time using
Homogeneous coordinates. Transformations in this space are represented by 44 matrices, which are not rotation
matrices but which have a 33 rotation matrix in the upper left corner.
The main disadvantage of matrices is that they are more expensive to calculate and do calculations with. Also in
calculations where numerical stability is a concern matrices can be more prone to it, so calculations to restore
orthonormality, which are expensive to do for matrices, need to be done more often.

Euler angles
One way of generalising the two dimensional angle of rotation is to
specify three rotation angles, carried out in turn about the three
principal axes. They individually can be labelled yaw, pitch, and roll,
but in mathematics are more often known by their mathematical name,
Euler angles. They have the advantage of modelling a number of
physical systems such as gimbals, and joysticks, so are easily
visualised, and are a very compact way of storing a rotation. But they
are difficult to use in calculations as even simple operations like
combining rotations are expensive to do, and suffer from a form of
The principal axes of rotation in space
gimbal lock where the angles cannot be uniquely calculated for certain
rotations.
Rotation (mathematics) 75

Axis angle
A second way of generalising the two dimensional angle of rotation is to specify an angle with the axis about which
the rotation takes place. It can be used to model motion constrained by a hinges and Axles, and so is easily
visualised, perhaps even more so than Euler angles. There are two ways to represent it;
as a pair consisting of the angle and a unit vector for the axis, or
as a vector obtained by multiplying the angle with this unit vector, called the rotation vector.
Usually the angle and axis pair is easier to work with, while the rotation vector is more compact, requiring only three
numbers like Euler angles. But like Euler angles it is usually converted to another representation before being used.

Quaternions
Quaternions are in some ways the least intuitive representation of three dimensional rotations. They are not the three
dimensional instance of a general approach, like matrices; nor are they easily related to real world models, like Euler
angles or axis angles. But they are more compact than matrices and easier to work with than all other methods, so are
often preferred in real world applications.
A rotation quaternion consists of four real numbers, constrained so the length of the quaternion considered as a
vector is 1. This constraint limits the degree of freedom of the quaternion to three, as required. It can be thought of as
a generalisation of the complex numbers, by e.g. the CayleyDickson construction, and generates rotations in a
similar way by multiplication. But unlike matrices and complex numbers two multiplications are needed:

where q is the quaternion q-1 is its inverse and x is the vector treated as a quaternion. The quaternion can be related
to the rotation vector form of the axis angle rotation by the exponential map over the quaternions,

Where v is the rotation vector treated as a quaternion.

Four dimensions
A general rotation in four dimensions has only one fixed point, the centre of rotation, and no axis of rotation. Instead
the rotation has two mutually orthogonal planes of rotation, each of which is fixed in the sense that points in each
plane stay within the planes. The rotation has two angles of rotation, one for each plane of rotation, through which
points in the planes rotate. If these are 1 and 2 then all points not in the planes rotate through an angle between 1
and 2.
If 1 = 2 the rotation is a double rotation and all points rotate through the same angle so any two orthogonal planes
can be taken as the planes of rotation. If one of 1 and 2 is zero, one plane is fixed and the rotation is simple. If both
1 and 2 are zero the rotation is the identity rotation.[2]
Rotations in four dimensions can be represented by 4th order orthogonal matrices, as a generalisation of the rotation
matrix. Quaternions can also be generalised into four dimensions, as even Multivectors of the four dimensional
Geometric algebra. A third approach, which only works in four dimensions, is to use a pair of unit quaternions.
Rotations in four dimensions have six degrees of freedom, most easily seen when two unit quaternions are used, as
each has three degrees of freedom (they lie on the surface of a 3-sphere) and 2 3 = 6.
Rotation (mathematics) 76

Relativity
One application of this is special relativity, as it can be considered to operate in a four dimensional space, spacetime,
spanned by three space dimensions and one of time. In special relativity this space is linear and the four dimensional
rotations, called Lorentz transformations, have practical physical interpretations.
If a simple rotation is only in the three space dimensions, i.e. about a plane that is entirely in space, then this rotation
is the same as a spatial rotation in three dimensions. But a simple rotation about a plane spanned by a space
dimension and a time dimension is a "boost", a transformation between two different reference frames, which
together with other properties of spacetime determines the relativistic relationship between the frames. The set of
these rotations forms the Lorentz group.[3]

Generalizations

Orthogonal matrices
The set of all matrices M(v,) described above together with the operation of matrix multiplication is called rotation
group: SO(3).
More generally, coordinate rotations in any dimension are represented by orthogonal matrices. The set of all
orthogonal matrices of the n-th dimension which describe proper rotations (determinant = +1), together with the
operation of matrix multiplication, forms the special orthogonal group: SO(n).
Orthogonal matrices have real elements. The analogous complex-valued matrices are the unitary matrices. The set of
all unitary matrices in a given dimension n forms a unitary group of degree n, U(n); and the subgroup of U(n)
representing proper rotations forms a special unitary group of degree n, SU(n). The elements of SU(2) are used in
quantum mechanics to rotate spin.

See also
Rotation representation
Spinor
Charts on SO(3)
Euler angles
Vortical
Rotation group
Coordinate rotations and reflections
Rodrigues' rotation formula
Rotation matrix
Orientation (geometry)
Rotation (mathematics) 77

References
Hestenes, David (1999). New Foundations for Classical Mechanics. Dordrecht: Kluwer Academic Publishers.
ISBN0-7923-5514-8.
Lounesto, Pertti (2001). Clifford algebras and spinors. Cambridge: Cambridge University Press.
ISBN978-0-521-00551-7.

External links
Rotation about any point [4]

References
[1] Lounesto 2001, p.30.
[2] Lounesto 2001, pp. 85, 89.
[3] Hestenes 1999, pp. 580 - 588.
[4] http:/ / www. euclideanspace. com/ maths/ geometry/ affine/ aroundPoint/

Color balance
In photography and image processing, color balance is the global
adjustment of the intensities of the colors (typically red, green, and
blue primary colors). An important goal of this adjustment is to render
specific colors particularly neutral colors correctly; hence, the
general method is sometimes called gray balance, neutral balance, or
white balance. Color balance changes the overall mixture of colors in
an image and is used for color correction; generalized versions of color
balance are used to get colors other than neutrals to also appear correct
or pleasing.
The left half shows the photo as it came from the
Image data acquired by sensors either film or electronic image digital camera. The right half shows the photo
sensors must be transformed from the acquired values to new values adjusted to make a gray surface neutral in the
same light.
that are appropriate for color reproduction or display. Several aspects
of the acquisition and display process make such color correction
essential including the fact that the acquisition sensors do not match
the sensors in the human eye, that the properties of the display medium
must be accounted for, and that the ambient viewing conditions of the
acquisition differ from the display viewing conditions.

The color balance operations in popular image editing applications

usually operate directly on the red, green, and blue channel pixel
values,[1] [2] without respect to any color sensing or reproduction
model. In shooting film, color balance is typically achieved by using A seascape photograph at Clifton Beach, South
color correction filters over the lights or on the camera lens.[3] Arm, Tasmania, Australia. The white balance has
been adjusted towards the warm side for creative
effect.

Generalized color balance

Color balance 78

Sometimes the adjustment to keep neutrals neutral is called white balance, and the phrase color balance refers to the
adjustment that in addition makes other colors in a displayed image appear to have the same general appearance as
the colors in an original scene.[4] It is particularly important that neutral (gray, achromatic, white) colors in a scene
appear neutral in the reproduction. Hence, the special case of balancing the neutral colors (sometimes gray balance,
neutral balance, or white balance) is a particularly important perhaps dominant element of color balancing.
Normally, one would not use the phrase color balance to describe the adjustments needed to account for differences
between the sensors and the human eye, or the details of the display primaries. Color balance is normally reserved to
refer to correction for differences in the ambient illumination conditions. However, the algorithms for transforming
the data do not always clearly separate out the different elements of the correction. Hence, it can be difficult to
assign color balance to a specific step in the color correction process. Moreover, there can be significant differences
in the color balancing goal. Some applications are created to produce an accurate rendering as suggested above. In
other applications, the goal of color balancing is to produce a pleasing rendering. This difference also creates
difficulty in defining the color balancing processing operations.

Illuminant estimation and adaptation

Most digital cameras have a means to select a color correction based on the type of scene illumination, using either
manual illuminant selection, or automatic white balance (AWB), or custom white balance. The algorithm that
performs this analysis performs generalized color balancing, known as illuminant adaptation or chromatic
adaptation.
Many methods are used to achieve color balancing. Setting a button on a camera is a way for the user to indicate to
the processor the nature of the scene lighting. Another option on some cameras is a button which one may press
when the camera is pointed at a gray card or other neutral object. This "custom white balance" step captures an
image of the ambient light, and this information is helpful in controlling color balance.
There is a large literature on how one might estimate the ambient illumination from the camera data and then use this
information to transform the image data. A variety of algorithms have been proposed, and the quality of these have
been debated. A few examples and examination of the references therein will lead the reader to many others.
Examples are Retinex, an artificial neural network[5] or a Bayesian method.[6]

Color balance and chromatic colors

Color balancing an image affects not only the neutrals, but other colors as well. An image that is not color balanced
is said to have a color cast, as everything in the image appears to have been shifted towards one color or another.[7]
Color balancing may be thought in terms of removing this color cast.
Color balance is also related to color constancy. Algorithms and techniques used to attain color constancy are
frequently used for color balancing, as well. Color constancy is, in turn, related to chromatic adaptation.
Conceptually, color balancing consists of two steps: first, determining the illuminant under which an image was
captured; and second, scaling the components (e.g., R, G, and B) of the image or otherwise transforming the
components so they conform to the viewing illuminant.
Viggiano found that white balancing in the camera's native RGB tended to produce less color inconstancy (i.e., less
distortion of the colors) than in monitor RGB for over 4000 hypothetical sets of camera sensitivities.[8] This
difference typically amounted to a factor of more than two in favor of camera RGB. This means that it is
advantageous to get color balance right at the time an image is captured, rather than edit later on a monitor. If one
must color balance later, balancing the raw image data will tend to produce less distortion of chromatic colors than
balancing in monitor RGB.
Color balance 79

Mathematics of color balance

Color balancing is sometimes performed on a three-component image (e.g., RGB) using a 3x3 matrix. This type of
transformation is appropriate if the image were captured using the wrong white balance setting on a digital camera,
or through a color filter.

Scaling monitor R, G, and B

In principle, one wants to scale all relative luminances in an image so that objects which are believed to be neutral
appear so. If, say, a surface with R=240 was believed to be a white object, and if 255 is the count which corresponds
to white, one could multiply all red values by 255/240. Doing analogously for green and blue would result, at least in
theory, in a color balanced image. In this type of transformation the 3x3 matrix is a diagonal matrix.

where , , and are the color balanced red, green, and blue components of a pixel in the image; , ,
and are the red, green, and blue components of the image before color balancing, and , , and are
the red, green, and blue components of a pixel which is believed to be a white surface in the image before color
balancing. This is a simple scaling of the red, green, and blue channels, and is why color balance tools in Photoshop
and the GIMP have a white eyedropper tool. It has been demonstrated that performing the white balancing in the
phosphor set assumed by sRGB tends to produce large errors in chromatic colors, even though it can render the
neutral surfaces perfectly neutral.[8]

Scaling X, Y, Z
If the image may be transformed into CIE XYZ tristimulus values, the color balancing may be performed there. This
has been termed a wrong von Kries transformation.[9] [10] Although it has been demonstrated to offer usually
poorer results than balancing in monitor RGB, it is mentioned here as a bridge to other things. Mathematically, one
computes:

where , , and are the color-balanced tristimulus values; , , and are the tristimulus values of
the viewing illuminant (the white point to which the image is being transformed to conform to); , , and
are the tristimulus values of an object believed to be white in the un-color-balanced image, and , , and
are the tristimulus values of a pixel in the un-color-balanced image. If the tristimulus values of the monitor primaries
are in a matrix so that:

where , , and are the un-gamma corrected monitor RGB, one may use:
Color balance 80

Von Kries's method

Johannes von Kries, whose theory of rods and three different color-sensitive cone types in the retina has survived as
the dominant explanation of color sensation for over 100 years, motivated the method of converting color to the
LMS color space, representing the effective stimuli for the Long-, Medium-, and Short-wavelength cone types that
are modeled as adapting independently. A 3x3 matrix converts RGB or XYZ to LMS, and then the three LMS
primary values are scaled to balance the neutral; the color can then be converted back to the desired final color
space:[11]

where , , and are the color-balanced LMS cone tristimulus values; , , and are the
tristimulus values of an object believed to be white in the un-color-balanced image, and , , and are the
tristimulus values of a pixel in the un-color-balanced image.
Matrices to convert to LMS space were not specified by von Kries, but can be derived from CIE color matching
functions and LMS color matching functions when the latter are specified; matrices can also be found in reference
books.[11]

Scaling camera RGB

By Viggiano's measure, and using his model of gaussian camera spectral sensitivities, most camera RGB spaces
performed better than either monitor RGB or XYZ.[8] If the camera's raw RGB values are known, one may use the
3x3 diagonal matrix:

and then convert to a working RGB space such as sRGB or Adobe RGB after balancing.

Preferred chromatic adaptation spaces

Comparisons of images balanced by diagonal transforms in a number of different RGB spaces have identified
several such spaces that work better than others, and better than camera or monitor spaces, for chromatic adaptation,
as measured by several color appearance models; the systems that performed statistically as well as the best on the
majority of the image test sets used were the "Sharp", "Bradford", "CMCCAT", and "ROMM" spaces.[12]

General illuminant adaptation

The best color matrix for adapting to a change in illuminant is not necessarily a diagonal matrix in a fixed color
space. It has long been known that if the space of illuminants can be described as a linear model with N basis terms,
the proper color transformation will be the weighted sum of N fixed linear transformations, not necessarily
consistently diagonalizable.[13]
Color balance 81

See also
Color temperature
Gamma correction
White point

External links
White Balance [14] - Intro at nikondigital.org
Understanding White Balance [15] - Tutorial
Affine color balance with saturation, with code and on-line demonstration [16]

References
[1] Phyllis Davis (2000). The Gimp for Linux and Unix (http:/ / books. google. com/ books?id=0sEnoWrMw-gC& pg=PA135& dq="color+
balance"+ channels& lr=& as_brr=3& ei=mYt5SJGMPJGAsgOuzZWzAQ&
sig=ACfU3U00P3v5iABD0tZhkM5bSlR1fsPMsQ#PPA134,M1). Peachpit Press. p.134. ISBN0201702533. .
[2] Adobe Creative Team (2000). Adobe Photoshop 6.0 (http:/ / books. google. com/ books?id=MRtx2-0GZc4C& pg=PA277& dq="color+
balance"+ channels& lr=& as_brr=3& ei=mYt5SJGMPJGAsgOuzZWzAQ&
sig=ACfU3U1aB0i0BUJG3nbN1dXJUvBou6Gajg#PPA278,M1). Adobe Press. p.278. ISBN0201710161. .
[3] Blain Brown (2002). Cinematography: Theory and Practice : Imagemaking for Cinematographers, Directors, and Videographers (http:/ /
books. google. com/ books?id=1JL2jFbNPNAC& pg=PA170& dq="color+ balance"& lr=& as_brr=0& ei=Top5SOy7E4y4tAPzsMzGDQ&
sig=ACfU3U2wBQs78QpTuogruMY8iYrfUixNfw). Focal Press. p.170. ISBN0240805003. .
[4] Hsien-Che Lee (2005). Introduction to Color Imaging Science (http:/ / books. google. com/ books?id=CzAbJrLin_AC& pg=PA450&
vq=color+ balance& dq="color+ balance"+ wandell& lr=& as_brr=3& source=gbs_search_s&
sig=ACfU3U05D2xQtbMhci7AzwGsjgKiwCmaFQ#PPA450,M1). Cambridge University Press. p.450. ISBN052184388X. .
[5] Brian Funt, Vlad Cardei, and Kobus Barnard, " Learning color constancy (http:/ / www. cs. sfu. ca/ ~colour/ publications/ ARIZONA/
arizona_abs. html)," in Proceedings of the Fourth IS&T/SID Color Imaging Conference, p 58-60 (1996).
[6] Graham Finlayson, Paul M. Hubel, and Steven Hordley (November 2001). "Color by correlation: a simple, unifying framework for color
constancy" (http:/ / www2. cmp. uea. ac. uk/ Research/ compvis/ Papers/ FinHorHub_PAMI01. pdf) (PDF). IEEE Transactions on Pattern
Analysis and Machine Intelligence 23 (11): 12091221. doi:10.1109/34.969113. .
[7] John A C Yule, Principles of Color Reproduction. New York: Wiley, 1967.
[8] J A Stephen Viggiano, " Comparison of the accuracy of different white balancing options as quantified by their color constancy (http:/ / www.
acolyte-color. com/ papers/ EI_2004. pdf)." Sensors and Camera Systems for Scientific, Industrial, and Digital Photography Applications V:
Proceedings of the SPIE, volume 5301. Bellingham, WA: SPIE: the International Society for Optical Engineering, p 323-333 (2004), retrieved
online 2008-07-28.
[9] Heinz Terstiege (1972). "Chromatic adaptation: a state-of-the-art report". Journal of Color Appearance 1 (4): 1923 (cont. 40).
[10] Mark D Fairchild, Color Appearance Models. Reading, MA: Addison-Wesley, 1998.
[11] Gaurav Sharma (2003). Digital Color Imaging Handbook (http:/ / books. google. com/ books?id=AkByHKRGTsQC& pg=PA153&
ots=5cP5rMeBIR& dq="von+ Kries"& sig=jxdadE-fxwClP1p1b7OnMwF_VWk). CRC Press. p.153. ISBN084930900X. .
[12] Sabine Ssstrunk, Jack Holm, and Graham D. Finlayson (January 2001). "Chromatic Adaptation Performance of Different RGB Sensors"
(http:/ / infoscience. epfl. ch/ getfile. py?recid=34049& mode=best). IS&T/SPIE Electronic Imaging 4300. .
[13] Laurence T. Maloney and Brain A. Wandell (1987). "Color constancy: a method for recovering surface spectral reflectance" (http:/ / books.
google. com/ books?hl=en& lr=& id=W5hLHUI8U-kC& oi=fnd& pg=PA293& dq=maloney+ wandell& ots=-iyJSpg6SZ&
sig=HsjhVTy3RYSbQzJ3WvVp_hzqtjA#PPA293,M1). in Martin A. Fischler and Oscar Firschein. Readings in Computer Vision.
Morgan-Kaufmann. ISBN0934613338. .
[14] http:/ / www. nikondigital. org/ articles/ white_balance. htm
[15] http:/ / www. photoxels. com/ tutorial_white-balance. html
[16] http:/ / www. ipol. im/ pub/ algo/ lmps_simplest_color_balance/
Image registration 82

Image registration
Image registration is the process of transforming different sets of data
into one coordinate system. Data may be multiple photographs, data
from different sensors, from different times, or from different
viewpoints.[1] It is used in, computer vision, medical imaging, military
automatic target recognition, and compiling and analyzing images and
data from satellites. Registration is necessary in order to be able to
compare or integrate the data obtained from these different
measurements.

Algorithm classification

Intensity-based vs feature-based
Image registration or image alignment algorithms can be classified into
intensity-based and feature-based[2] . One of the images is referred to
as the reference or source and the second image is referred to as the
target or sensed. Image registration involves spatially transforming the
Registering and summing multiple exposures of
target image to align with the reference image[2] .Intensity-based the same scene improves signal to noise ratio,
methods compare intensity patterns in images via correlation metrics, allowing to see things previously impossible to
while feature-based methods find correspondence between image see.
[2]
features such as points, lines, and contours . Intensity-based methods
register entire images or subimages. If subimages are registered, centers of corresponding subimages are treated as
corresponding feature points. Feature-based method established correspondence between a number of points in
images. Knowing the correspondence between a number of points in images, a transformation is then determined to
map the target image to the reference images, thereby establishing point-by-point correspondence between the
reference and target images[2] .

Transformation models
Image registration algorithms can also be classified according to the transformation models they use to relate the
target image space to the reference image space. The first broad category of transformation models includes linear
transformations, which include translation, rotation, scaling, and other affine transforms. Linear transformations are
global in nature, thus, they cannot model local geometric differences between images[2] .
The second category of transformations allow 'elastic' or 'nonrigid' transformations. These transformations are
capable of locally warping the target image to align with the reference image. Nonrigid transformations include
radial basis functions (thin-plate or surface splines, multiquadrics, and compactly-supported transformations[2] ),
physical continuum models (viscous fluids), and large deformation models (diffeomorphisms).

Spatial vs. frequency domain methods

Spatial methods operate in the image domain, matching intensity patterns or features in images. Some of the feature
matching algorithms are outgrowths of traditional techniques for performing manual image registration, in which an
operator chooses corresponding control points (CPs) in images. When the number of control points exceeds the
minimum required to define the appropriate transformation model, iterative algorithms like RANSAC can be used to
robustly estimate the parameters of a particular transformation type (e.g. affine) for registration of the images.
Image registration 83

Frequency-domain methods find the transformation parameters for registration of the images while working in the
transform domain. Such methods work for simple transformations, such as translation, rotation, and scaling.
Applying the Phase correlation method to a pair of images produces a third image which contains a single peak. The
location of this peak corresponds to the relative translation between the images. Unlike many spatial-domain
algorithms, the phase correlation method is resilient to noise, occlusions, and other defects typical of medical or
satellite images. Additionally, the phase correlation uses the fast Fourier transform to compute the cross-correlation
between the two images, generally resulting in large performance gains. The method can be extended to determine
rotation and scaling differences between two images by first converting the images to log-polar coordinates. Due to
properties of the Fourier transform, the rotation and scaling parameters can be determined in a manner invariant to
translation.

Single- vs. multi-modality methods

Another classification can be made between single-modality and multi-modality methods. Single-modality methods
tend to register images in the same modality acquired by the same scanner/sensor type, while multi-modality
registration methods tended to register images acquired by different scanner/sensor types.
Multi-modality registration methods are often used in medical imaging as images of a subject are frequently obtained
from different scanners. Examples include registration of brain CT/MRI images or whole body PET/CT images for
tumor localization, registration of contrast-enhanced CT images against non-contrast-enhanced CT images for
segmentation of specific parts of the anatomy, and registration of ultrasound and CT images for prostate localization
in radiotherapy.

Automatic vs. interactive methods

Registration methods may be classified based on the level of automation they provide. Manual, interactive,
semi-automatic, and automatic methods [3] have been developed. Manual methods provide tools to align the images
manually. Interactive methods reduce user bias by performing certain key operations automatically while still relying
on the user to guide the registration. Semi-automatic methods perform more of the registration steps automatically
but depend on the user to verify the correctness of a registration. Automatic methods do not allow any user
interaction and perform all registration steps automatically.

Similarity measures for image registration

Image similarities are broadly used in medical imaging. An image similarity measure quantifies the degree of
similarity between intensity patterns in two images[2] . The choice of an image similarity measure depends on the
modality of the images to be registered. Common examples of image similarity measures include cross-correlation,
mutual information, sum of squared intensity differences, and ratio image uniformity. Mutual information and
normalized mutual information are the most popular image similarity measures for registration of multimodality
images. Cross-correlation, sum of squared intensity differences and ratio image uniformity are commonly used for
registration of images in the same modality.
Image registration 84

Uncertainty
There is a level of uncertainty associated with registering images that have any spatio-temporal differences. A
confident registration with a measure of uncertainty is critical for many change detection applications such as
medical diagnostics.
In remote sensing applications where a digital image pixel may represent several kilometers of spatial distance (such
as NASA's LANDSAT imagery), an uncertain image registration can mean that a solution could be several
kilometers from ground truth. Several notable papers have attempted to quantify uncertainty in image registration in
order to compare results.[4] [5] However, many approaches to quantifying uncertainty or estimating deformations are
computational intensive or are only applicable to limited sets of spatial transformations.

Applications
Image registration has applications in remote sensing (cartography updating), and computer vision. Due to the vast
applications to which image registration can be applied, it is impossible to develop a general method that is
optimized for all uses.
Medical image registration (for data of the same patient taken at different points in time such as change detection or
tumor monitoring)) often additionally involves elastic (also known as nonrigid) registration to cope with
deformation of the subject (due to breathing, anatomical changes, and so forth). Nonrigid registration of medical
images can also be used to register a patient's data to an anatomical atlas, such as the Talairach atlas for
neuroimaging.
It is also used in astrophotography to align images taken of space. Using control points (automatically or manually
entered), the computer performs transformations on one image to make major features align with a second image.
Image registration is essential part of Panoramic image creation. There are many different techniques that can be
implemented in real time and run on embedded devices like cameras and camera-phones.

See also
Spatial normalization

External links
Richard Szeliski, Image Alignment and Stitching: A Tutorial [6]. Foundations and Trends in Computer Graphics
and Computer Vision, 2:1-104, 2006.
B. Fischer, J. Modersitzki: Ill-posed medicine an introduction to image registration [7]. Inverse Problems,
24:119, 2008
Barbara Zitov, Jan Flusser: Image registration methods: a survey [8]. Image Vision Comput. 21(11): 977-1000
(2003).
Image registration 85

References
[1] Lisa Gottesfeld Brown, A survey of image registration techniques (abstract), ACM Computing Surveys (CSUR) archive, Volume 24 , Issue
4, December 1992), Pages: 325 - 376 (http:/ / portal. acm. org/ citation. cfm?id=146374)
[2] A. Ardeshir Goshtasby: 2-D and 3-D Image Registration for Medical, Remote Sensing, and Industrial Applications (http:/ / www. wiley. com/
WileyCDA/ WileyTitle/ productCd-0471649546. html), Wiley Press, 2005.
[3] http:/ / www. imgfsr. com/ ifsr_ir. html
[4] Simonson, K., Drescher, S., Tanner, F., A Statistics Based Approach to Binary Image Registration with Uncertainty Analysis. IEEE Pattern
Analysis and Machine Intelligence, Vol. 29, No. 1, January 2007
[5] Domokos, C., Kato, Z., Francos, J., Parametric estimation of affine deformations of binary images. Proceedings of IEEE International
Conference on Acoustics, Speech, and Signal Processing, 2008
[6] http:/ / research. microsoft. com/ apps/ pubs/ default. aspx?id=75695
[7] http:/ / www. iop. org/ EJ/ article/ 0266-5611/ 24/ 3/ 034008/ ip8_3_034008. pdf
[8] http:/ / library. utia. cas. cz/ prace/ 20030125. pdf

Segmentation (image processing)

In computer vision, segmentation refers to the process of partitioning a digital image into multiple segments (sets of
pixels, also known as superpixels). The goal of segmentation is to simplify and/or change the representation of an
image into something that is more meaningful and easier to analyze.[1] Image segmentation is typically used to locate
objects and boundaries (lines, curves, etc.) in images. More precisely, image segmentation is the process of assigning
a label to every pixel in an image such that pixels with the same label share certain visual characteristics.
The result of image segmentation is a set of segments that collectively cover the entire image, or a set of contours
extracted from the image (see edge detection). Each of the pixels in a region are similar with respect to some
characteristic or computed property, such as color, intensity, or texture. Adjacent regions are significantly different
with respect to the same characteristic(s).[1]

Applications
Some of the practical applications of image segmentation are:
Medical Imaging[2]
Locate tumors and other pathologies
Measure tissue volumes
Computer-guided surgery
Diagnosis
Treatment planning
Study of anatomical structure
Locate objects in satellite images (roads, forests, etc.)
Face recognition
Fingerprint recognition
Traffic control systems
Brake light detection
Machine vision
Several general-purpose algorithms and techniques have been developed for image segmentation. Since there is no
general solution to the image segmentation problem, these techniques often have to be combined with domain
knowledge in order to effectively solve an image segmentation problem for a problem domain.
Segmentation (image processing) 86

Clustering methods
The K-means algorithm is an iterative technique that is used to partition an image into K clusters. The basic
algorithm is:
1. Pick K cluster centers, either randomly or based on some heuristic
2. Assign each pixel in the image to the cluster that minimizes the variance between the pixel and the cluster center
3. Re-compute the cluster centers by averaging all of the pixels in the cluster
4. Repeat steps 2 and 3 until convergence is attained (e.g. no pixels change clusters)
In this case, variance is the squared or absolute difference between a pixel and a cluster center. The difference is
typically based on pixel color, intensity, texture, and location, or a weighted combination of these factors. K can be
selected manually, randomly, or by a heuristic.
This algorithm is guaranteed to converge, but it may not return the optimal solution. The quality of the solution
depends on the initial set of clusters and the value of K.
In statistics and machine learning, the k-means algorithm is clustering algorithm to partition n objects into k clusters,
where k < n. It is similar to the expectation-maximization algorithm for mixtures of Gaussians in that they both
attempt to find the centers of natural clusters in the data. The model requires that the object attributes correspond to
elements of a vector space. The objective it tries to achieve is to minimize total intra-cluster variance, or, the squared
error function. The k-means clustering was invented in 1956. The most common form of the algorithm uses an
iterative refinement heuristic known as Lloyd's algorithm. Lloyd's algorithm starts by partitioning the input points
into k initial sets, either at random or using some heuristic data. It then calculates the mean point, or centroid, of each
set. It constructs a new partition by associating each point with the closest centroid. Then the centroids are
recalculated for the new clusters, and algorithm repeated by alternate application of these two steps until
convergence, which is obtained when the points no longer switch clusters (or alternatively centroids are no longer
changed). Lloyd's algorithm and k-means are often used synonymously, but in reality Lloyd's algorithm is a heuristic
for solving the k-means problem, as with certain combinations of starting points and centroids, Lloyd's algorithm can
in fact converge to the wrong answer. Other variations exist, but Lloyd's algorithm has remained popular, because it
converges extremely quickly in practice. In terms of performance the algorithm is not guaranteed to return a global
optimum. The quality of the final solution depends largely on the initial set of clusters, and may, in practice, be much
poorer than the global optimum. Since the algorithm is extremely fast, a common method is to run the algorithm
several times and return the best clustering found. A drawback of the k-means algorithm is that the number of
clusters k is an input parameter. An inappropriate choice of k may yield poor results. The algorithm also assumes
that the variance is an appropriate measure of cluster scatter.

Histogram-based methods
Histogram-based methods are very efficient when compared to other image segmentation methods because they
typically require only one pass through the pixels. In this technique, a histogram is computed from all of the pixels in
the image, and the peaks and valleys in the histogram are used to locate the clusters in the image.[1] Color or
intensity can be used as the measure.
A refinement of this technique is to recursively apply the histogram-seeking method to clusters in the image in order
to divide them into smaller clusters. This is repeated with smaller and smaller clusters until no more clusters are
formed.[1] [3]
One disadvantage of the histogram-seeking method is that it may be difficult to identify significant peaks and valleys
in the image. In this technique of image classification distance metric and integrated region matching are familiar.
Histogram-based approaches can also be quickly adapted to occur over multiple frames, while maintaining their
single pass efficiency. The histogram can be done in multiple fashions when multiple frames are considered. The
same approach that is taken with one frame can be applied to multiple, and after the results are merged, peaks and
Segmentation (image processing) 87

valleys that were previously difficult to identify are more likely to be distinguishable. The histogram can also be
applied on a per pixel basis where the information result are used to determine the most frequent color for the pixel
location. This approach segments based on active objects and a static environment, resulting in a different type of
segmentation useful in Video tracking.

Edge detection
Edge detection is a well-developed field on its own within image processing. Region boundaries and edges are
closely related, since there is often a sharp adjustment in intensity at the region boundaries. Edge detection
techniques have therefore been used as the base of another segmentation technique.
The edges identified by edge detection are often disconnected. To segment an object from an image however, one
needs closed region boundaries.

Region growing methods

The first region growing method was the seeded region growing method. This method takes a set of seeds as input
along with the image. The seeds mark each of the objects to be segmented. The regions are iteratively grown by
comparing all unallocated neighbouring pixels to the regions. The difference between a pixel's intensity value and
the region's mean, , is used as a measure of similarity. The pixel with the smallest difference measured this way is
allocated to the respective region. This process continues until all pixels are allocated to a region.
Seeded region growing requires seeds as additional input. The segmentation results are dependent on the choice of
seeds. Noise in the image can cause the seeds to be poorly placed. Unseeded region growing is a modified algorithm
that doesn't require explicit seeds. It starts off with a single region the pixel chosen here does not significantly
influence final segmentation. At each iteration it considers the neighbouring pixels in the same way as seeded region
growing. It differs from seeded region growing in that if the minimum is less than a predefined threshold then
it is added to the respective region . If not, then the pixel is considered significantly different from all current
regions and a new region is created with this pixel.
One variant of this technique, proposed by Haralick and Shapiro (1985),[1] is based on pixel intensities. The mean
and scatter of the region and the intensity of the candidate pixel is used to compute a test statistic. If the test statistic
is sufficiently small, the pixel is added to the region, and the regions mean and scatter are recomputed. Otherwise,
the pixel is rejected, and is used to form a new region.

Level set methods

Curve propagation is a popular technique in image analysis for object extraction, object tracking, stereo
reconstruction, etc. The central idea behind such an approach is to evolve a curve towards the lowest potential of a
cost function, where its definition reflects the task to be addressed and imposes certain smoothness constraints.
Lagrangian techniques are based on parameterizing the contour according to some sampling strategy and then evolve
each element according to image and internal terms. While such a technique can be very efficient, it suffers from
various limitations like deciding on the sampling strategy, estimating the internal geometric properties of the curve,
changing its topology, addressing problems in higher dimensions, etc.
The level set method was initially proposed to track moving interfaces by Osher and Sethian in 1988 and has spread
across various imaging domains in the late nineties. It can be used to efficiently address the problem of
curve/surface/etc. propagation in an implicit manner. The central idea is to represent the evolving contour using a
signed function, where its zero level corresponds to the actual contour. Then, according to the motion equation of the
contour, one can easily derive a similar flow for the implicit surface that when applied to the zero-level will reflect
the propagation of the contour. The level set method encodes numerous advantages: it is implicit, parameter free,
provides a direct way to estimate the geometric properties of the evolving structure, can change the topology and is
Segmentation (image processing) 88

intrinsic. Furthermore, they can be used to define an optimization framework as proposed by Zhao, Merriman and
Osher in 1996. Therefore, one can conclude that it is a very convenient framework to address numerous applications
of computer vision and medical image analysis.[4] Furthermore, research into various level set data structures has led
to very efficient implementations of this method.

Graph partitioning methods

Graph partitioning methods can effectively be used for image segmentation. In these methods, the image is modeled
as a weighted, undirected graph. Usually a pixel or a group of pixels are associated with nodes and edge weights
define the (dis)similarity between the neighborhood pixels. The graph (image) is then partitioned according to a
criterion designed to model "good" clusters. Each partition of the nodes (pixels) output from these algorithms are
considered an object segment in the image. Some popular algorithms of this category are normalized cuts [5] ,
random walker [6] , minimum cut [7] , isoperimetric partitioning [8] and minimum spanning tree-based segmentation
[9]
.

Watershed transformation
The watershed transformation considers the gradient magnitude of an image as a topographic surface. Pixels having
the highest gradient magnitude intensities (GMIs) correspond to watershed lines, which represent the region
boundaries. Water placed on any pixel enclosed by a common watershed line flows downhill to a common local
intensity minimum (LIM). Pixels draining to a common minimum form a catch basin, which represents a segment.

Model based segmentation

The central assumption of such an approach is that structures of interest/organs have a repetitive form of geometry.
Therefore, one can seek for a probabilistic model towards explaining the variation of the shape of the organ and then
when segmenting an image impose constraints using this model as prior. Such a task involves (i) registration of the
training examples to a common pose, (ii) probabilistic representation of the variation of the registered samples, and
(iii) statistical inference between the model and the image. State of the art methods in the literature for
knowledge-based segmentation involve active shape and appearance models, active contours and deformable
templates and level-set based methods.

Multi-scale segmentation
Image segmentations are computed at multiple scales in scale-space and sometimes propagated from coarse to fine
scales; see scale-space segmentation.
Segmentation criteria can be arbitrarily complex and may take into account global as well as local criteria. A
common requirement is that each region must be connected in some sense.

One-dimensional hierarchical signal segmentation

Witkin's seminal work[10] [11] in scale space included the notion that a one-dimensional signal could be
unambiguously segmented into regions, with one scale parameter controlling the scale of segmentation.
A key observation is that the zero-crossings of the second derivatives (minima and maxima of the first derivative or
slope) of multi-scale-smoothed versions of a signal form a nesting tree, which defines hierarchical relations between
segments at different scales. Specifically, slope extrema at coarse scales can be traced back to corresponding features
at fine scales. When a slope maximum and slope minimum annihilate each other at a larger scale, the three segments
that they separated merge into one segment, thus defining the hierarchy of segments.
Segmentation (image processing) 89

Image segmentation and primal sketch

There have been numerous research works in this area, out of which a few have now reached a state where they can
be applied either with interactive manual intervention (usually with application to medical imaging) or fully
automatically. The following is a brief overview of some of the main research ideas that current approaches are
based upon.
The nesting structure that Witkin described is, however, specific for one-dimensional signals and does not trivially
transfer to higher-dimensional images. Nevertheless, this general idea has inspired several other authors to
investigate coarse-to-fine schemes for image segmentation. Koenderink[12] proposed to study how iso-intensity
contours evolve over scales and this approach was investigated in more detail by Lifshitz and Pizer[13] .
Unfortunately, however, the intensity of image features changes over scales, which implies that it is hard to trace
coarse-scale image features to finer scales using iso-intensity information.
Lindeberg[14] [15] studied the problem of linking local extrema and saddle points over scales, and proposed an image
representation called the scale-space primal sketch which makes explicit the relations between structures at different
scales, and also makes explicit which image features are stable over large ranges of scale including locally
appropriate scales for those. Bergholm proposed to detect edges at coarse scales in scale-space and then trace them
back to finer scales with manual choice of both the coarse detection scale and the fine localization scale.
Gauch and Pizer[16] studied the complementary problem of ridges and valleys at multiple scales and developed a tool
for interactive image segmentation based on multi-scale watersheds. The use of multi-scale watershed with
application to the gradient map has also been investigated by Olsen and Nielsen[17] and been carried over to clinical
use by Dam[18] Vincken et al.[19] proposed a hyperstack for defining probabilistic relations between image structures
at different scales. The use of stable image structures over scales has been furthered by Ahuja[20] and his co-workers
into a fully automated system.
More recently, these ideas for multi-scale image segmentation by linking image structures over scales have been
picked up by Florack and Kuijper[21] . Bijaoui and Ru[22] associate structures detected in scale-space above a
minimum noise threshold into an object tree which spans multiple scales and corresponds to a kind of feature in the
original signal. Extracted features are accurately reconstructed using an iterative conjugate gradient matrix method.

Semi-automatic segmentation
In this kind of segmentation, the user outlines the region of interest with the mouse clicks and algorithms are applied
so that the path that best fits the edge of the image is shown.
Techniques like Siox, Livewire, or Intelligent Scissors are used in this kind of segmentation.

Neural networks segmentation

Neural Network segmentation relies on processing small areas of an image using an artificial neural network [23] or a
set of neural networks. After such processing the decision-making mechanism marks the areas of an image
accordingly to the category recognized by the neural network. A type of network designed especially for this is the
Kohonen map.
Pulse-Coupled Neural Networks (PCNNs) are neural models proposed by modeling a cats visual cortex and
developed for high-performance biomimetic image processing. In 1989, Eckhorn introduced a neural model to
emulate the mechanism of cats visual cortex. The Eckhorn model provided a simple and effective tool for studying
small mammals visual cortex, and was soon recognized as having significant application potential in image
processing. In 1994, the Eckhorn model was adapted to be an image processing algorithm by Johnson, who termed
this algorithm Pulse-Coupled Neural Network. Over the past decade, PCNNs have been utilized for a variety of
image processing applications, including: image segmentation, feature generation, face extraction, motion detection,
region growing, noise reduction, and so on. A PCNN is a two-dimensional neural network. Each neuron in the
Segmentation (image processing) 90

network corresponds to one pixel in an input image, receiving its corresponding pixels color information (e.g.
intensity) as an external stimulus. Each neuron also connects with its neighboring neurons, receiving local stimuli
from them. The external and local stimuli are combined in an internal activation system, which accumulates the
stimuli until it exceeds a dynamic threshold, resulting in a pulse output. Through iterative computation, PCNN
neurons produce temporal series of pulse outputs. The temporal series of pulse outputs contain information of input
images and can be utilized for various image processing applications, such as image segmentation and feature
generation. Compared with conventional image processing means, PCNNs have several significant merits, including
robustness against noise, independence of geometric variations in input patterns, capability of bridging minor
intensity variations in input patterns, etc.

Open source software

Several open source software packages are available for performing image segmentation
ITK - Insight Segmentation and Registration Toolkit (Open Source)
ITK-SNAP is a GUI tool that combines manual and semi-automatic segmentation with level sets.
GIMP which includes among other tools SIOX (see Simple Interactive Object Extraction)
VXL
ImageMagick
3DSlicer
MITK [24] has a program module for manual segmentation
OpenCV is a computer vision library originally developed by Intel.
GRASS GIS has the program module i.smap [25] for image segmentation
Fiji - Fiji is just ImageJ, an image processing package which includes different segmentation plug-ins.
JavaScript Graph-based Segmentation [26] is a JavaScript implementation of the clustering technique described
here [27].
There are also software packages available free of charge for academic purposes:
GemIdent
CVIPtools [28]
MegaWave [29]

See also
Computer Vision
Data clustering
Graph Theory
Histograms
Image-based meshing
K-means algorithm
Pulse-coupled networks
Range image segmentation
Region growing
Segmentation (image processing) 91

External links
Some sample code that performs basic segmentation [30], by Syed Zainudeen. University Technology of
Malaysia.

References
[1] Linda G. Shapiro and George C. Stockman (2001): Computer Vision, pp 279-325, New Jersey, Prentice-Hall, ISBN 0-13-030796-3
[2] Dzung L. Pham, Chenyang Xu, and Jerry L. Prince (2000): Current Methods in Medical Image Segmentation, Annual Review of Biomedical
Engineering, volume 2, pp 315-337
[3] Ron Ohlander, Keith Price, and D. Raj Reddy (1978): Picture Segmentation Using a Recursive Region Splitting Method, Computer
Graphics and Image Processing, volume 8, pp 313-333
[4] S. Osher and N. Paragios. Geometric Level Set Methods in Imaging Vision and Graphics (http:/ / www. mas. ecp. fr/ vision/ Personnel/ nikos/
osher-paragios/ ), Springer Verlag, ISBN 0387954880, 2003.
[5] Jianbo Shi and Jitendra Malik (2000): "Normalized Cuts and Image Segmentation" (http:/ / www. cs. dartmouth. edu/ ~farid/ teaching/ cs88/
pami00. pdf), IEEE Transactions on pattern analysis and machine intelligence, pp 888-905, Vol. 22, No. 8
[6] Leo Grady (2006): "Random Walks for Image Segmentation" (http:/ / www. cns. bu. edu/ ~lgrady/ grady2006random. pdf), IEEE
Transactions on Pattern Analysis and Machine Intelligence, pp. 1768-1783, Vol. 28, No. 11
[7] Z. Wu and R. Leahy (1993): "An optimal graph theoretic approach to data clustering: Theory and its application to image segmentation" (ftp:/
/ sipi. usc. edu/ pub/ leahy/ pdfs/ MAP93. pdf), IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 1101-1113, Vol. 15, No.
11
[8] Leo Grady and Eric L. Schwartz (2006): "Isoperimetric Graph Partitioning for Image Segmentation" (http:/ / www. cns. bu. edu/ ~lgrady/
grady2006isoperimetric. pdf), IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 469-475, Vol. 28, No. 3
[9] C. T. Zahn (1971): "Graph-theoretical methods for detecting and describing gestalt clusters" (http:/ / web. cse. msu. edu/ ~cse802/ Papers/
zahn. pdf), IEEE Transactions on Computers, pp. 68-86, Vol. 20, No. 1
[10] Witkin, A. P. "Scale-space filtering", Proc. 8th Int. Joint Conf. Art. Intell., Karlsruhe, Germany,1019--1022, 1983.
[11] A. Witkin, "Scale-space filtering: A new approach to multi-scale description," in Proc. IEEE Int. Conf. Acoust., Speech, Signal Processing
(ICASSP), vol. 9, San Diego, CA, Mar. 1984, pp. 150--153.
[12] Koenderink, Jan "The structure of images", Biological Cybernetics, 50:363--370, 1984
[13] Lifshitz, L. and Pizer, S.: A multiresolution hierarchical approach to image segmentation based on intensity extrema, IEEE Transactions on
Pattern Analysis and Machine Intelligence, 12:6, 529 - 540, 1990. (http:/ / portal. acm. org/ citation. cfm?id=80964& dl=GUIDE&
coll=GUIDE)
[14] Lindeberg, T.: Detecting salient blob-like image structures and their scales with a scale-space primal sketch: A method for
focus-of-attention, International Journal of Computer Vision, 11(3), 283--318, 1993. (http:/ / www. nada. kth. se/ ~tony/ abstracts/
Lin92-IJCV. html)
[15] Lindeberg, Tony, Scale-Space Theory in Computer Vision, Kluwer Academic Publishers, 1994 (http:/ / www. nada. kth. se/ ~tony/ book.
html), ISBN 0-7923-9418-6
[16] Gauch, J. and Pizer, S.: Multiresolution analysis of ridges and valleys in grey-scale images, IEEE Transactions on Pattern Analysis and
Machine Intelligence, 15:6 (June 1993), pages: 635 - 646, 1993. (http:/ / portal. acm. org/ citation. cfm?coll=GUIDE& dl=GUIDE&
id=628490)
[17] Olsen, O. and Nielsen, M.: Multi-scale gradient magnitude watershed segmentation, Proc. of ICIAP 97, Florence, Italy, Lecture Notes in
Computer Science, pages 613. Springer Verlag, September 1997.
[18] Dam, E., Johansen, P., Olsen, O. Thomsen,, A. Darvann, T. , Dobrzenieck, A., Hermann, N., Kitai, N., Kreiborg, S., Larsen, P., Nielsen, M.:
"Interactive multi-scale segmentation in clinical use" in European Congress of Radiology 2000.
[19] Vincken, K., Koster, A. and Viergever, M.: Probabilistic multiscale image segmentation (doi:10.1109/34.574787), IEEE Transactions on
Pattern Analysis and Machine Intelligence, 19:2, pp. 109-120, 1997.]
[20] M. Tabb and N. Ahuja, Unsupervised multiscale image segmentation by integrated edge and region detection, IEEE Transactions on Image
Processing, Vol. 6, No. 5, 642-655, 1997. (http:/ / vision. ai. uiuc. edu/ ~msingh/ segmen/ seg/ MSS. html)
[21] Florack, L. and Kuijper, A.: The topological structure of scale-space images, Journal of Mathematical Imaging and Vision, 12:1, 65-79,
2000.
[22] Bijaoui, A., Ru, F.: 1995, A Multiscale Vision Model, Signal Processing 46, 345 (http:/ / dx. doi. org/ 10. 1016/ 0165-1684(95)00093-4)
[23] Mahinda Pathegama & Gl (2004): "Edge-end pixel extraction for edge-based image segmentation", Transactions on Engineering,
Computing and Technology, vol. 2, pp 213-216, ISSN 1305-5313
[24] http:/ / www. mitk. org
[25] http:/ / grass. osgeo. org/ grass64/ manuals/ html64_user/ i. smap. html
[26] http:/ / pvnick. x10hosting. com/ segmentation/ segmentation_demo. htm
[27] http:/ / people. cs. uchicago. edu/ %7Epff/ papers/ seg-ijcv. pdf
[28] http:/ / www. ee. siue. edu/ CVIPtools/
[29] http:/ / megawave. cmla. ens-cachan. fr/ index. php
Segmentation (image processing) 92

[30] http:/ / csc. fsksm. utm. my/ syed/ projects/ image-processing. html

3D Entropy Based Image Segmentation (http:/ / instrumentation. hit. bg/ Papers/ 2008-02-02 3D Multistage Entropy.
htm)
Frucci, Maria; Sanniti di Baja, Gabriella (2008). " From Segmentation to Binarization of Gray-level Images
(http://www.jprr.org/index.php/jprr/article/view/54/16)". Journal of Pattern Recognition Research (http://
www.jprr.org/index.php/jprr) 3 (1): 113.
Article Sources and Contributors 93

Article Sources and Contributors

Image processing Source: http://en.wikipedia.org/w/index.php?oldid=362347544 Contributors: Abhinav.bal, Abmac, Adoniscik, Alanbly, Andre Engels, Andthu, AndyCharles1Wilson,
Animum, Apyule, Asc99c, Bact, Barticus88, Bravekermit, BrentN, C3o, Calltech, Carloscastro, Cburnett, Chintanbug, Dancter, DanielZhang27, Dicklyon, Dwachter, Dysprosia, EagleFan, Fangz,
Ferengi, Fgnievinski, Finn-Zoltan, Fletch1978, Flightsoffancy, Fordescort79, Giftlite, GillianSmith, HarisM, Hezarfenn, Huotfh, Inschanr, Iobuddha, James086, Jiuguang Wang, Johnuniq,
Jonawebb, Jorge Stolfi, Kamoshirenai, Karol Langner, Kipmaster, Komap, Kutabar, Lamoidfl, Lee Daniel Crocker, Mahogny, Malte singapore, Mani1, Marc Lacoste, Matt Britt, Mdd, Mdd4696,
Meera123, Melaneum, Michael Hardy, MichaelBillington, MightCould, Minghong, Mostafa.elhoushi, Mr Yarrow, MrMambo, Nbarth, Nitin.mehta, Odo1982, Orderud, Organizmo, Ossmann,
PAR, Panecasareccio, Pearle, Phillipsacp, Pollinator, Poor Yorick, Porieu, Predictor, Pvosta, Quuxplusone, Renatokeshet, Requestion, Rholton, Roesser, Sam Korn, Sbowers3, Seabhcan,
SensibleStuff, Smack, Solipsist, Stw, Sumbaug, Supten, The Anome, Theresa knott, Tpl, Tuxide, User A1, Verbal, Wafulz, Waxigloo, Wikipedie, Wipware, Ylloh, Yswismer, 163 anonymous
edits

Digital image processing Source: http://en.wikipedia.org/w/index.php?oldid=361813256 Contributors: 777sms, Abdull, Adoniscik, Alansohn, Alex Spade, Andre Engels, AngelovdS, Apm-gh,
Asc99c, Bact, Barticus88, Beetstra, BrentN, Canaima, Cburnett, Chunawalla, D hanbun, DBAllan, DavidCary, Dchristle, Dicklyon, Dysprosia, Eekerz, Either way, Ekotkie, Enselic, Enviroboy,
Fgnievinski, Flowerheaven, GRAHAMUK, Giftlite, Glogger, HairyWombat, Hermann.tropf, Hezarfenn, Hosse, Imagegod, Inputpersona, Isnow, Jamelan, Jehochman, JimmyTheWig, Jiuguang
Wang, Johnuniq, Jorge Stolfi, Keilana, Lamoidfl, Lbarajas, Ligulem, MER-C, Matt Britt, Mav, Mholland, Michael Hardy, MightCould, Mindmatrix, Mr Yarrow, MrOllie, Naturenet, NawlinWiki,
Nbarth, Nilx, Nmnogueira, Nneonneo, ONEder Boy, Orderud, Pedronavajo, Phillipsacp, Prashantsonar, Predictor, Pvosta, Quuxplusone, Rich Farmbrough, Roesser, Roshan baladhanvi,
Sanfranman59, Sannse, Satishpitchikala, Seabhcan, Shireesh13, Snowolf, Stillastillsfan, Stonor, TFOWR, Targetter, The Anome, Themcman1, Tobias Conradi, Tomkalayil, Tuxide, User A1,
Wharron, Wiki alf, Wikipedie, Wildknot, Yaroslavvb, Ylloh, Zvika, Zvn, , ., 120 anonymous edits

Digital imaging Source: http://en.wikipedia.org/w/index.php?oldid=360569779 Contributors: Adoniscik, Altenmann, Althepal, Andrew113, DMCer, David Fuchs, Dgies, Dicklyon,
DocWatson42, Dthomsen8, Evil saltine, Funandtrvl, Gracemobley, Hemanshu, Imroy, JTN, JeremyA, Jiang, Jim.henderson, Jorge Stolfi, L7HOMAS, LordSnow, MER-C, Malcolma, Mattisse,
Maxamegalon2000, Nbarth, Nitomarsiglio, Nslimak, Oicumayberight, Pa-merynaten, Pak21, Pdburns, Phantomsteve, Phillipsacp, Pvosta, RedWolf, Renesis, RenniePet, SimonTrew, Smack,
Snowolf, Sparkleforce, Stemonitis, TBarregren, TheMindsEye, Tide rolls, Tongue and Groove, Xsamix, 57 anonymous edits

Medical imaging Source: http://en.wikipedia.org/w/index.php?oldid=360755398 Contributors: 01bambern77, A314268, AED, AXu.VSI, Action potential, Adw2000, Alex.tan, Altenmann,
Altzinn, AndreasJS, Andrel, Anupam, Appraiser, Arcadian, Auntminnie, Aushulz, Austegard, Bazzargh, Beetstra, Beevee, Beland, Benjaminevans82, Berland, Beth.farrell, Bignoter, Binksternet,
Blowdart, Bme591wikiproject, Bradhegg, Btunell, CMD Beaker, Cameron.walsh, Can't sleep, clown will eat me, Canis Lupus, Cozmog, Cuthbertwong, DHN, DMacks, Darklilac,
David.Monniaux, Dicklyon, Dina, Discospinster, DocWatson42, Doczilla, Download, Dozen, Dr Zak, Draligus, Elsys, Erikp, Esprit15d, Ferengi, Flowerheaven, Freedomworks, Fumitol,
Fuzzball!, G3pro, Ganesha Freya, GargoyleMT, Gene Nygaard, Giftlite, Glitzy queen00, Goldencuboid, GordyB, Gradient11, GreatWhiteNortherner, GregorB, Gwernol, Hamanibrave,
Harrydevoil, HenryLi, Hoof Hearted, Hu12, IXICOexpert, Ian Pitchford, Ihealthbuzz, Imroy, Iweber2003, J04n, JVinocur, JaGa, Jamelan, JamesAM, Jengod, Jfdwolff, JimVC3, Jmanzanares,
Jmsc312, John.d.van.horn, Joshnpowell, Karada, Karnesky, Katiem222, Kauczuk, Khalid hassani, Khanafer, Kl4m-AWB, Klar sagen, Kram9, Ksiddiqui, Kubanczyk, LilHelpa, Loode,
Louislemieux, LuisIbanez, Lynnflint, MGlosenger, MMA Texas, Madmedea, Majeka, Mako098765, Mandy0147, Manwichosu, Marcika, Maximus Rex, Maxxicum, Mco44, Mdd, Mechanistic,
Med imaging, Melchoir, Mendaliv, Michael Snow, Miguel Andrade, Mikael Hggstrm, Mirada Medical, Morenus, MrOllie, Nachiketgokhale, NerdyNSK, Nevit, Oclatz, Ohnoitsjamie, Oli Filth,
Oxymoron83, Paranoid, Pde, Peterclipston, Peterkmoch, Petersam, Philipcosson, Philippe Nicolai-Dashwood, Pion, Publicoption, Pvosta, Radrounds, Rafaela.lamy, RedWolf, Rhcastilhos, Rich
Farmbrough, Rick Sidwell, Rjwilmsi, Robofish, Rphilipp, Rsabbatini, S0uj1r0, SYSS Mouse, Sabry hassouna, Sarahb87, Scottandrewhutchins, Seabhcan, ShakingSpirit, Shunpiker, Sietse Snel,
Simeon antonov, SimonaN, Sistromc, Sjschen, Slicky, Smocking, Snoogans1984, Solok7, Soundray, Spencer, Splmko, Stuartghall, Supten, Talented Mr Miller, The Anome, The Rationalist,
TheRealNightRider, Thekohser, Themfromspace, Thor Waldsen, Tpl, User A1, UzEE, Van helsing, Wavelength, Way4thesub, Weregerbil, Wilhelm Bauer, Xanzzibar, Xiggelee, Yoderj, Zereshk,
215 anonymous edits

Quantization (signal processing) Source: http://en.wikipedia.org/w/index.php?oldid=353095128 Contributors: Aadri, Beetelbug, C xong, Cat5nap, Cburnett, Charles Matthews, DeadTotoro,
DmitriyV, DopefishJustin, Engelec, Giftlite, Glenn, Hadiyana, JordiTost, Kku, Krash, Longhair, Matt Gies, Metacomet, Michael Hardy, Nbarth, Necromancer44, Oarih, Omegatron, OverlordQ,
Pak21, Paolo.dL, Petr.adamek, Qz, Radagast83, Rbj, RedWolf, Sapphic, Slady, Smack, SudoMonas, Sylenius, The Anome, Topbanana, WhiteHatLurker, 52 anonymous edits

Brightness Source: http://en.wikipedia.org/w/index.php?oldid=362999065 Contributors: 16@r, AED, Adoniscik, Andycjp, AuburnPilot, BeSherman, C3o, Captainreuben, CheshireKatz,
Ciphers, David Shear, Dbachmann, Dicklyon, Dysprosia, EagleFan, Epbr123, Gaba p, Geocachernemesis, Jacobolus, JeLuF, Johnkarp, KP-Adhikari, Lelouch.Angelo., LiDaobing, MartinHarper,
Mcld, Mhss, Michael Hardy, Neckelmann, Neelix, Neparis, Newone, Patrick, Rebornsoldier, Silence, Srleffler, StefanVanDerWalt, TheMindsEye, UBIQ, Unara, WinterSpw, Zephyr103,
ZeroOne, 32 anonymous edits

Luminance Source: http://en.wikipedia.org/w/index.php?oldid=357906472 Contributors: 213.253.39.xxx, Aboalbiss, Andres, Aulis Eskola, Bemoeial, BenFrantzDale, Bignose, ChrisJ, Cmglee,
Conversion script, Cookie4869, Darkmeerkat, Davidofithaca, Dicklyon, DmitTrix, ElfQrin, Foobaz, GD 666, Gaius Cornelius, Gary Cziko, Gary D, Glennchan, Grafen, GregorB, Headbomb,
Heron, Housed, Icairns, Icd, Jossi, Jyril, Martin Kozk, Michael Hardy, Mstasa, Nedim Ardoa, Nick125, Nuggetboy, Octahedron80, Paddyez, Quilbert, Rogper, Shape84, Srleffler, SudoMonas,
Teorth, Tobias Hoevekamp, Trounce, Uniqueuponhim, Uriyan, Wahrmund, Wiki alf, Zappaz, 35 anonymous edits

Contrast (vision) Source: http://en.wikipedia.org/w/index.php?oldid=360624809 Contributors: AED, AVB, Anonywiki, Dicklyon, Ecoleetage, Epbr123, Excirial, Famousdog, Firsfron, Fmalan,
Frap, Gianpan, H2g2bob, HisSpaceResearch, Jacobolus, JeffPerry, Jwe21, Kntrabssi, Leia, Leujohn, Mnmngb, Nbarth, Nuvitauy07, Pikez33, Pince Nez, R oshea, Radagast83, Rurik1, Smalljim,
Srleffler, Stemonitis, Stephanusst, Stiepan Pietrov, Telecineguy, Travelbird, Wikipelli, 54 anonymous edits

Color space Source: http://en.wikipedia.org/w/index.php?oldid=362830344 Contributors: 166.70.2.xxx, Aadnk, Aboalbiss, Aditsu, Adoniscik, AgentPeppermint, Alan Peakall, Alue, Arthena,
Asc99c, Ashishbhatnagar72, Aubrey Jaffer, Aursani, Avsa, Beetstra, Benandorsqueaks, Bevo, Bobo192, BorgQueen, Branko, Brianski, Brona, CALR, CesarB, Chrisc666, Chrumps, Colordoc,
Conversion script, Cpesacreta, Cxw, Cyp, Da monster under your bed, Damian Yerrick, DariusMonsef, Darrien, Dave3457, Denelson83, Dkroll2, Dmurphy, Dominus, Dwandelt, FishSpeaker,
Fredrik, Grendelkhan, Gutza, Hellisp, Heron, Hyacinth, ISteve, Jacobolus, Jay, Jonathan Webley, Josef Meixner, Kasper Hviid, Lambtron, Life of Riley, LittleDan, Lovibond, Marluxia.Kyoshu,
Mfc, Mindmatrix, Misst r, Moeron, NekoDaemon, Neparis, Noldoaran, Normankoren, Notinasnaid, Odo1982, Orderud, PAR, Poccil, Rebornsoldier, Retired username, Rhebus, Rich Farmbrough,
Richie, Rjwilmsi, Rythie, Sam Hocevar, Sardanaphalus, Seabhcan, Seattlenow, SharkD, Skarebo, Sluzzelin, Smyth, SocratesJedi, Sparkit, Sterrys, Suruena, Sven271, Ta bu shi da yu, Tletnes,
Twirligig, Uaxuctum, Umofomia, Wapcaplet, Zundark, 135 anonymous edits

Color mapping Source: http://en.wikipedia.org/w/index.php?oldid=356879527 Contributors: Odo1982, Orangemike, Ukexpat, 2 anonymous edits

Color management Source: http://en.wikipedia.org/w/index.php?oldid=362581100 Contributors: Aboalbiss, Adamdewitz, Adoniscik, Alfio, Anand07.s, Badly Bradley, BorgQueen,
CambridgeBayWeather, CesarB, ChestRockwell, Ciampix, Cje, Cms tglc, Colorhq, DavidForthoffer, Dicklyon, Docu, Dougweller, Droll, Ekotkie, Elkman, Erikvlie, Fotospeed, Freshacconci,
GWG, Glennrp, Gogo Dodo, Gutza, Hawky, Hemidemisemiquaver, Hyperhypo, Iainpike, Ian Dunster, Iedmonds, Islandboy99, Jacobolus, Jonas August, Joriki, KaiUwe, Kate, Kdau, Keenan
Pepper, Lemon-s, Max theprintspace, Mike U, Mikenolte, Mlewan, Musichacker, Nandesuka, NekoDaemon, Nettlehead, OnlineBG, Pat Herold, Pcordes, Pestling, RVJ, Radagast83,
Rebornsoldier, RedAndr, Rick Sidwell, Serge Cashman, Sinistrad, Squirepants101, Stuartyeates, Suguru@Musashi, Sven Boisen, Ta bu shi da yu, Telarc, TheGood, Thumperward, VMS Mosaic,
Wheaty, Wik, Wusspie, Zundark, 127 anonymous edits

Digital Imaging and Communications in Medicine Source: http://en.wikipedia.org/w/index.php?oldid=361265016 Contributors: Abdull, Alex.tan, Alphachimp, Amdescombes, Amelio
Vzquez, Aminzade, Andreala, Avernet, Beevee, Blakecaldwell, Boivie, Brighterorange, Bwr54, Ceejayoz, Charlesb, Chengyanliang, Chruck, Ciscoperin, Cuthbert8, Dclunie, Dglane001,
Dicomprogrammer, Dineshkoka, Dqmillar, DrAdamInCA, EscapingLife, Frap, Freedomworks, Ganesha Freya, GargoyleMT, Gleb, Glider87, IXICOexpert, Imran qau, Intellijean, JHKrueger,
Jfdwolff, Johnkmetha, JungleRob, KellyCoinGuy, Khalid hassani, Khanafer, KissL, Kolmigabrouil, Kolmodin, Kr-val, Ksaravanakumar, Kubanczyk, Kweeket, Lambiam, Lancetfluke,
Lovecreak, Markaud, Mathieu, Mco44, Medconn, Movses, MrOllie, MrSerif, MrSincere, Mwarf, Ncouture, Nereocystis, Okelm, Onorem, Oxymoron83, Pacsman, Phatom87, Pmg, Rbardwell,
Rhcoe, Rlee0001, Ronz, Rsabbatini, Sanesto, SeanAhern, Simeon antonov, Smelnikov, Soundray, Stephan Leeds, Supten, Surv1v4l1st, TRosenbaum, TastyPoutine, Test21, Threexk, Tuqui,
Umls04aa, Undermyumbrella, WLU, Wheelie, Zabdiel, Zoicon5, mer Cengiz elebi, , 190 anonymous edits

JPEG 2000 Source: http://en.wikipedia.org/w/index.php?oldid=362619614 Contributors: .:Ajvol:., 15.253, 19.154, Adoniscik, Aeons, Akisankaku, Alejo2083, AlisonW, AnonMoos, Anthony
Fok, Apoc2400, Arru, Ascnder, Astronaut, Aths, Baszoetekouw, Bender235, Bilboq, Blutfink, Boardhead, Bobblehead, Boemanneke, BonsaiViking, Brokentooth, Bryan Derksen, Can't sleep,
clown will eat me, Cantalamessa, Caspcasp, Cek, Charlesriver, Chris Roy, Ckomnick, Cluthi, CountingPine, Crissov, CyberSkull, DaBler, Daniel.Cardenas, Dark Shikari, David Ashley, David
Gerard, Davidhorman, Dbenbenn, Derek Ross, Dicklyon, Digital dinosaur, DmitriyV, DopefishJustin, Dougransom, Download, Dreiraumzimmer, Efa, Evands, Eyreland, Feedmecereal, Ft1,
Gadget850, Ganesha Freya, Gareth Owen, GeorgeMoney, Ghettoblaster, Giftlite, Gparker, Gregfr, GregorB, Guy Harris, Gwillis, Haeleth, Harryboyles, Henning Makholm, Hnc14, Ian Dunster,
Ianmacm, Inspector 34, Jbarrett, JeremyTraub, John, KJRehberg, Katimawan2005, Kauai68, Klingoncowboy4, Kovianyo, Mabodo, Mahanga, Mathieu, Mattabat, Mazin07, Mboverload,
Michele.alessandrini, Mikhailfranco, Minghong, Mjb, Mr.wikikaw, MrJones, Mxn, Nbarth, Nerd65536, Nikevich, Nneonneo, Ohms law, OlavN, Oli Filth, Olivier.amato, Omegatron, Onejaguar,
OverlordQ, Ozhiker, Ozuma, Pangolin, Patrickdepinguin, Peak, Pegase, Piet Delport, Pipatron, Pne, Quentar, Qutezuce, Rade Kutil, Rcooley, RedWolf, Reisio, Requestion, Rpresser, Rursus,
Article Sources and Contributors 94

SPUI, Saoshyant, Saxifrage, Seabhcan, Seano1, Shervinafshar, Shibboleth, Shishir0610, Shlomital, Shoujun, Skarebo, Smably, Ssd, SteveSims, Stolsvik, SudoMonas, Superm401, TWWhiting,
Tajymoid, TakuyaMurata, Tannenmann, Taral, Thomas Richter, Thorfdbg, Timonczesq, Tired Starling, TomViza, Tomia, Tompagenet, Travelster, Tristanb, Tuvas, Untill, VMS Mosaic,
Wdscxsj, Wikidemon, Wjl, Xander99, Xdc-evs, Xpclient, Zeneky, Zhatt, Zundark, 280 anonymous edits

Linear filter Source: http://en.wikipedia.org/w/index.php?oldid=348104905 Contributors: Aldie, Atlant, Belizefan, Bloodshedder, CPES, CanisRufus, Cburnett, Conversion script, Darkwind,
Dicklyon, Egil, Eric119, Fieldday-sunday, Freestyle, Glenn, Hangjian, Heron, Hooperbloob, Iain, Ianharvey, Iluvcapra, Jeff3000, Karada, Karl-Henner, Lexor, LouScheffer, Mako098765,
Michael Hardy, Oli Filth, Omegatron, PAR, Pjacobi, Poulpy, Reddi, SimonP, Spinningspark, Taw, TedPavlic, The Anome, The Thing That Should Not Be, Tim Starling, Tisakowi, Tobias
Hoevekamp, Yath, 31 anonymous edits

Histogram Source: http://en.wikipedia.org/w/index.php?oldid=362485241 Contributors: 160.5.82.xxx, 2D, 5 albert square, ABF, Aastrup, AbsolutDan, Aitias, Alansohn, AlekseyP, Asterion,
Auldglory, Aushulz, Avenue, AxelBoldt, Baa, Bart133, Barticus88, BenFrantzDale, Bettia, BlaiseFEgan, Bobo192, Bogey97, Borgx, Boxplot, Calvin 1998, Capt. James T. Kirk, Charles
Matthews, Chris53516, Chuunen Baka, Conversion script, Cristianrodenas, Ctbolt, Cyclopia, DVD R W, DanielPenfield, DarthVader, Deep316, Demus Wiesbaden, Den fjttrade ankan,
Disavian, Dkkicks, DmitTrix, Donarreiskoffer, Ehrenkater, Eng.ahmedm, Enviroboy, Epbr123, Evgeny, Faithlessthewonderboy, Fatka, FayssalF, FreplySpang, Furrykef, G716, Garglebutt, Gary
King, Gcjblack, GeordieMcBain, Giftlite, Glen, Hadleywickham, HenningThielemann, Hgberman, Hooperbloob, Hu12, Hydrogen Iodide, Immunize, Indon, Irishguy, Iwaterpolo, J.delanoy,
JRSP, James086, Jamesootas, Japanese Searobin, Jerry teps, Jiminyjimjim, Jni, Johan1298, JohnManuel, Jusdafax, Karol Langner, Kgasso, Kierano, Kjetil1001, Kku, KnightRider, Kuru, Lara
bran, Lathrop, LeaveSleaves, Leonam, Liftarn, Ligulem, Lindmere, Luna Santin, MBlakley, Magog the Ogre, Male1979, Manop, Martarius, Master of Puppets, Mattopia, Meekywiki, Melcombe,
Mendaliv, Metacomet, Mhym, Michael Hardy, Mihoshi, Moreschi, Mr coyne, Mwtoews, N5iln, Naravorgaara, Nbarth, NeilN, Neko-chan, Nield, Nijdam, Nsaa, Onionmon, Ossmann, P.B. Pilhet,
Pathoschild, Paul August, Philip Trueman, Piano non troppo, Pyrrhus16, Qwfp, RJaguar3, Red Thunder, Richjwild, Rjwilmsi, Robert P. O'Shea, RodC, Ronz, Rufous, SE7, Saxbryn, Sbwoodside,
Schutz, Seraphim, Shimazaki, Shoeofdeath, Smith609, Spaz4tw, Specs112, Steven J. Anderson, Steven Zhang, Straif, SuperHamster, Surya21, Taffykins, Tangent747, The Thing That Should Not
Be, The wub, TheNewPhobia, Thomas Arnhold, Tombrazel, Tomi, Tommy2010, Undisputedloser, Voice In The Wilderness, Wile E. Heresiarch, Woohookitty, Wwwwolf, YUL89YYZ,
Yamaguchi, Yurik, Zach1994, Zaharous, Zheric, Zhou yi777, Zondox, ZooFari, Zr40, Zvika, 364 anonymous edits

Image histogram Source: http://en.wikipedia.org/w/index.php?oldid=362895468 Contributors: Abdull, Andfriend, Cewvero, Dicklyon, Errantkid, Fatheroftebride, Gkay1500, HairyWombat,
IronTek, Jannic86, Lara bran, Llorenzi, Lolicious, Macphesa, Manamarak, Mdd, Mihirgokani007, ONEder Boy, RoninChris64, Thejapanesegeek, Wikipelli, 11 anonymous edits

Color histogram Source: http://en.wikipedia.org/w/index.php?oldid=358934845 Contributors: Adoniscik, Alexander Anoprienko, Althepal, Asc99c, Charles Matthews, Chenzw, Codepete,
Dicklyon, Evil saltine, GB fan, HenningThielemann, Jeff3000, Jhinman, Jorge Stolfi, KYN, Kostmo, Lkinkade, Neurolysis, Ollydbg, Petiatil, Petr.noha, R'n'B, Rebornsoldier, Rich Farmbrough,
Sjakkalle, So two Willys walk into a bar..., Strait, UkPaolo, Valentein, Wiki alf, Zundark, 32 anonymous edits

Affine transformation Source: http://en.wikipedia.org/w/index.php?oldid=358194926 Contributors: .:Ajvol:., Addshore, Agreeney, Ajgorhoe, Alaudo, Albmont, Altenmann, Archelon,
AxelBoldt, BenFrantzDale, Cacadril, CatherineMunro, Charles Matthews, Chas zzz brown, Cjfsyntropy, Damian Yerrick, Dave Foley, DavidCBryant, Dcoetzee, Discospinster, Dvgrn, Dysprosia,
Eliyak, EmilJ, Ezrakilty, Felix Wiemann, Flinx, Foobaz, Fredrik, Giftlite, Guy Peters, Ht686rg90, Hyacinth, Iridescent, IvanAndreevich, J-Wiki, Jitse Niesen, JoergenB, Josh Cherry, LOL,
Lantonov, Longhair, Lord Bruyanovich Bruyanov, MSGJ, Mets501, Mhaitham.shammaa, Michael Hardy, Michael Slone, Mmernex, Mohan1986, Mpatel, Mysidia, Nostraticispeak, Nwbeeson,
Oleg Alexandrov, Paolo.dL, Patrick, PerezTerron, Phys, Pleasantville, Rgdboer, Rich257, SanderEvers, Schneelocke, SimonP, Skeetch, Sawomir Biay, Templarion, Tong, Urdutext,
VadimIppolitov, Yoshigev, 75 anonymous edits

Scaling (geometry) Source: http://en.wikipedia.org/w/index.php?oldid=356773915 Contributors: Aboeing, BD2412, Burntsauce, Charles Matthews, Dbfirs, Explodicle, Giftlite, Glenn,
Imasleepviking, JaGa, Jaredwf, Kelly elf, Lantonov, Little Mountain 5, Martin von Gagern, Mhss, MichaelMcGuffin, Mike Rosoft, Paolo.dL, Patrick, Pleasantville, Qwfp, 27 anonymous edits

Rotation (mathematics) Source: http://en.wikipedia.org/w/index.php?oldid=361240184 Contributors: Alexdow, AlmostReadytoFly, AugPi, BWDuncan, Beland, Brandonqd92, Charles
Matthews, Crazyvas, Dreftymac, Dynaflow, Dysprosia, Enisbayramoglu, Epbr123, Ewlyahoocom, Fer.jose.nandes, Fmalan, Fropuff, Gerbrant, Giftlite, Glenn, GregorB, Haseldon, Hworld, Igor
Yalovecky, Infovarius, JohnBlackburne, Jorunn, Juansempere, KSmrq, Kalazart, Kapelson, Linas, MaxSem, Mdob, MechPhD6, Metroman, Mhss, Michael Hardy, Mwtoews, NickBush24,
Nn123645, Noe, Oleg Alexandrov, Parvons, Patrick, PhotoBox, Portalian, Reddi, Remember the dot, Riojajar, Simpsons contributor, Smyth, Spoon!, Stamcose, Sawomir Biay, TakuyaMurata,
Tetracube, Tobias Bergemann, TomViza, Trevor MacInnis, Tslocum, Waldir, Wwoods, Xanderdad, 78 anonymous edits

Color balance Source: http://en.wikipedia.org/w/index.php?oldid=359332281 Contributors: -Majestic-, Adoniscik, Ancheta Wis, Andreas Kaufmann, Andrewferrier, Betacommand, Brettr,
ChestRockwell, Dave.Dunford, Dicklyon, Fg2, Geekinspired, Jacobolus, Jfriedl, Jordash, Kjlewis, Lovibond, MCTales, NekoDaemon, NeonMerlin, Nilx, Noodle snacks, Okki, P00r, Rjwilmsi,
Rooh23, Sbisolo, Sdgjake, Sfiga, Sterrys, Suckindiesel, Tabletop, The Thing That Should Not Be, Thorseth, Wandell, William Graham, Ykhwong, 30 anonymous edits

Image registration Source: http://en.wikipedia.org/w/index.php?oldid=363008620 Contributors: Adoniscik, Alansohn, Art LaPella, Avsa, BenFrantzDale, Bensmith, Brspurri,
CallipygianSchoolGirl, Carloscastro, ComputerScience101, Demiurge, Dglane001, DonPMitchell, Duedilly, Fnielsen, Fountains of Bryn Mawr, Gfariello, Giancarlo Rossi, Gurduloo, Heron,
Infzy, J.delanoy, Kl4m-AWB, Kostmo, LuisIbanez, Lulutiki, Milkmandan, Miym, Modersitzki, Mxlplk, NielsenGW, Nisheethg, Nwbeeson, Okseduard, Oleg Alexandrov, Oli Filth, Onevalefan,
Orderud, Pgan002, Psiphiorg, Rriker dad, SanderEvers, Stfnhs65, Thangalin, Thebigalphamaler, Tkapur, Tomash, Tpl, VoteFair, YoungMasterbator, ZooFari, 121 anonymous edits

Segmentation (image processing) Source: http://en.wikipedia.org/w/index.php?oldid=360538508 Contributors: Alexander Anoprienko, Altenmann, AttoRenato, Ayasi, Bachrach44,
Barticus88, Bawolff, BrokenSegue, CIreland, Calltech, Cybercobra, Dannyxyz22, Dishayloo, Dmccreary, El C, Eng.ahmedm, Etienne Decenciere, FRAINVITCHE, Ffangs, Formol, Fox, Fredrik,
G N Newsam, Henryjlyons, Hongooi, Huku-chan, Iweber2003, J04n, JEIhrig, Jan Wassenberg, Jeph paul, Jorge Stolfi, Karnesky, Kingpin13, Kl4m, Kl4m-AWB, LuisIbanez, Mailseth,
MarcoGallotta, Medicsl, Mietchen, Mostafa.elhoushi, Orderud, Paavan vasudev, Qmwne235, Quuxplusone, RJFJR, Random Tree, Rednblu, Revolver, STDestiny, SabreMau, Salix alba,
Seabhcan, ShelfSkewed, Smartcat, Spike Wilbury, Steve Quinn, Steven Zhang, Stevenj, Tamfang, Tkapur, Tpl, W3bbo, Way4thesub, Wegsjac, Weishaupt00, Yushkevich, Zawersh, 139
anonymous edits
Image Sources, Licenses and Contributors 95

Image Sources, Licenses and Contributors

Image:Imagen binaria.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Imagen_binaria.jpg License: Public Domain Contributors: Cyborg, Ma-Lik, 2 anonymous edits
Image:Alleia Hamerops composite.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Alleia_Hamerops_composite.jpg License: unknown Contributors: Berillium, GoClick, Gorgo,
Ma-Lik, Mdd4696
Image:Brain Mri nevit.svg Source: http://en.wikipedia.org/w/index.php?title=File:Brain_Mri_nevit.svg License: GNU Free Documentation License Contributors: Nevit Dilmen
Image:Sampled.signal.svg Source: http://en.wikipedia.org/w/index.php?title=File:Sampled.signal.svg License: Public Domain Contributors: User:Petr.adamek, User:Rbj
Image:Quantized.signal.svg Source: http://en.wikipedia.org/w/index.php?title=File:Quantized.signal.svg License: Public Domain Contributors: User:Petr.adamek, User:Rbj
Image:Digital.signal.svg Source: http://en.wikipedia.org/w/index.php?title=File:Digital.signal.svg License: Public Domain Contributors: User:Petr.adamek, User:Rbj
Image:Photo editing contrast correction.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Photo_editing_contrast_correction.jpg License: Public Domain Contributors: Original
uploader was Toniht at en.wikipedia
Image:Contrast change photoshop.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Contrast_change_photoshop.jpg License: GNU Free Documentation License Contributors:
Akinom, Chrkl, Roke, WikipediaMaster
Image:Tour Eiffel Notre-Dame.JPG Source: http://en.wikipedia.org/w/index.php?title=File:Tour_Eiffel_Notre-Dame.JPG License: Public Domain Contributors: Bibi Saint-Pol, Millevache,
Pseudomoi, Tokle
Image:Tour Eiffel Notre-Dame+contrast.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Tour_Eiffel_Notre-Dame+contrast.jpg License: Public Domain Contributors:
User:Jacobolus, User:Tokle
Image:SinVibr.png Source: http://en.wikipedia.org/w/index.php?title=File:SinVibr.png License: Public Domain Contributors: User:Aleksey463
Image:Colorspace.png Source: http://en.wikipedia.org/w/index.php?title=File:Colorspace.png License: Creative Commons Attribution 2.5 Contributors: Original uploader was Cpesacreta at
en.wikipedia
File:RGB_and_CMYK_comparison.png Source: http://en.wikipedia.org/w/index.php?title=File:RGB_and_CMYK_comparison.png License: Public Domain Contributors:
User:Marluxia.Kyoshu
Image:AdditiveColor.svg Source: http://en.wikipedia.org/w/index.php?title=File:AdditiveColor.svg License: Public Domain Contributors: Original uploader was SharkD at en.wikipedia Later
versions were uploaded by Jacobolus at en.wikipedia.
Image:SubtractiveColor.svg Source: http://en.wikipedia.org/w/index.php?title=File:SubtractiveColor.svg License: unknown Contributors: Original uploader was SharkD at en.wikipedia Later
version uploaded by Jacobolus, Dacium at en.wikipedia.
Image:Img_HistMatch_Source.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Img_HistMatch_Source.jpg License: Public Domain Contributors: Odo1982
Image:Img_HistMatch_Target.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Img_HistMatch_Target.jpg License: Public Domain Contributors: Odo1982
Image:Img_HistMatch_SourceAfter.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Img_HistMatch_SourceAfter.jpg License: Public Domain Contributors: Odo1982
File:Corta transversal PET TAC keosys.JPG Source: http://en.wikipedia.org/w/index.php?title=File:Corta_transversal_PET_TAC_keosys.JPG License: Public Domain Contributors:
User:Mco44
Image:Acrnema1.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Acrnema1.jpg License: Creative Commons Attribution-Sharealike 2.5 Contributors: Krister Valtonen
Image:Knie mr.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Knie_mr.jpg License: GNU Free Documentation License Contributors: Amada44, Dbenzhuser, Shizhao
Image:JPEG JFIF and 2000 Comparison.png Source: http://en.wikipedia.org/w/index.php?title=File:JPEG_JFIF_and_2000_Comparison.png License: Creative Commons
Attribution-Sharealike 2.5 Contributors: CountingPine, Kilom691, Nachcommonsverschieber, Shlomital, Sylenius, WikipediaMaster, 1 anonymous edits
Image:JPEG 2000 Artifacts Demonstration.png Source: http://en.wikipedia.org/w/index.php?title=File:JPEG_2000_Artifacts_Demonstration.png License: Creative Commons
Attribution-Sharealike 2.5 Contributors: Nachcommonsverschieber, Shlomital, Sylenius, WikipediaMaster, 3 anonymous edits
Image:Wavelet Bior2.2.svg Source: http://en.wikipedia.org/w/index.php?title=File:Wavelet_Bior2.2.svg License: Public Domain Contributors: User:DaBler
Image:Jpeg2000 2-level wavelet transform-lichtenstein.png Source: http://en.wikipedia.org/w/index.php?title=File:Jpeg2000_2-level_wavelet_transform-lichtenstein.png License: GNU Free
Documentation License Contributors: User:Alejo2083
Image:Lichtenstein jpeg2000 difference.png Source: http://en.wikipedia.org/w/index.php?title=File:Lichtenstein_jpeg2000_difference.png License: Creative Commons Attribution-Sharealike
2.5 Contributors: User:Alejo2083
Image:Electronic linear filters.svg Source: http://en.wikipedia.org/w/index.php?title=File:Electronic_linear_filters.svg License: Creative Commons Attribution-Sharealike 2.5 Contributors:
User:Alejo2083
Image:Histogram of arrivals per minute.svg Source: http://en.wikipedia.org/w/index.php?title=File:Histogram_of_arrivals_per_minute.svg License: Creative Commons Attribution-Sharealike
3.0 Contributors: User:DanielPenfield
Image:Black cherry tree histogram.svg Source: http://en.wikipedia.org/w/index.php?title=File:Black_cherry_tree_histogram.svg License: GNU Free Documentation License Contributors:
Mwtoews, 2 anonymous edits
Image:Travel time histogram total n Stata.png Source: http://en.wikipedia.org/w/index.php?title=File:Travel_time_histogram_total_n_Stata.png License: GNU Free Documentation License
Contributors: User:Qwfp
Image:Travel time histogram total 1 Stata.png Source: http://en.wikipedia.org/w/index.php?title=File:Travel_time_histogram_total_1_Stata.png License: Creative Commons
Attribution-Sharealike 3.0 Contributors: User:Qwfp
Image:Cumulative vs normal histogram.svg Source: http://en.wikipedia.org/w/index.php?title=File:Cumulative_vs_normal_histogram.svg License: GNU Free Documentation License
Contributors: User:Kierano
Image:SunLou2.jpg Source: http://en.wikipedia.org/w/index.php?title=File:SunLou2.jpg License: Public Domain Contributors: Original uploader was Phillipsacp at en.wikipedia
Image:SunHistacp.jpg Source: http://en.wikipedia.org/w/index.php?title=File:SunHistacp.jpg License: Public Domain Contributors: Original uploader was Phillipsacp at en.wikipedia
Image:Fractal fern explained.png Source: http://en.wikipedia.org/w/index.php?title=File:Fractal_fern_explained.png License: Public Domain Contributors: Antnio Miguel de Campos -
Image:Rotation illustration2.svg Source: http://en.wikipedia.org/w/index.php?title=File:Rotation_illustration2.svg License: Public Domain Contributors: User:Oleg Alexandrov
Image:Rotation4.svg Source: http://en.wikipedia.org/w/index.php?title=File:Rotation4.svg License: Creative Commons Attribution-Sharealike 2.5 Contributors: HB
Image:Simx2=rotOK.png Source: http://en.wikipedia.org/w/index.php?title=File:Simx2=rotOK.png License: Creative Commons Attribution 2.5 Contributors: McSush, Toobaz
Image:Flight dynamics with text.png Source: http://en.wikipedia.org/w/index.php?title=File:Flight_dynamics_with_text.png License: Creative Commons Attribution 3.0 Contributors:
User:ZeroOne
Image:Lily-M7292-As-shot-and-manual.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Lily-M7292-As-shot-and-manual.jpg License: Public Domain Contributors: User:Fg2
File:Clifton Beach 5.jpg Source: http://en.wikipedia.org/w/index.php?title=File:Clifton_Beach_5.jpg License: GNU Free Documentation License Contributors: User:Noodle snacks
Image:Registrator_Demo2.png Source: http://en.wikipedia.org/w/index.php?title=File:Registrator_Demo2.png License: GNU Free Documentation License Contributors: User:Clock
License 96

License
Creative Commons Attribution-Share Alike 3.0 Unported
http:/ / creativecommons. org/ licenses/ by-sa/ 3. 0/