Noida Institute of Engineering and Technology, Greater Noida

IMAGE RESTORATION

Unit: 4
Image Processing and Pattern
Recognition
ACSAI0522
Suhail Rashid
ASSISTANT PROFESOR
IT 5th Sem CSE

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 1
 EVALUATION SCHEME

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 2

 Course Syllabus

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 3

 Applications of Image Processing

• Image sharpening and restoration.

• Medical field.
• Remote sensing.
• Transmission and encoding.
• Machine/Robot vision.
• Color processing.
• Pattern recognition.
• Video processing.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 4

 Course Objectives

1. Analyze general terminology of digital image processing.

2. Examine various types of images, illumination factors, intensity
transformations and spatial filtering.
3. Develop Fourier transform for image processing in frequency domain.
4. Evaluate the methodologies for image segmentation, restoration etc.
5. Implement image process and analysis of algorithms for de- noising and
de- blurring.
6. Apply image processing algorithms in practical applications for enhancing
the images.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 5

 Course Outcomes
Bloom’s
Course Outcomes (CO) Knowledge Level
(KL)

To become familiar with digital image fundamentals. K1, K2

CO1

To get exposed to simple image enhancement techniques in Spatial K2, K3

and Frequency domain .
CO2

To learn concepts of degradation function and restoration K2, K3

techniques .
CO3
To study the image segmentation and representation techniques. K2,K3

CO4

To become familiar with image compression and recognition K2,K3

method .
CO5
12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 6
 PROGRAM OUTCOMES
1 Engineering knowledge
2 Problem analysis:
3 Design/development of solutions
4 Conduct investigations of complex problems
5 Modern tool usage
6 The engineer and society
7 Environment and sustainability
8 Ethics
9 Individual and team work
10 Communication
11 Project management and finance
12/12/24 12 Life-long learning
MINI JAIN KCS-062 Image Processing Unit-4 7
 CO-PO Mapping

CO/PO PO1 PO2 PO3 PO4 PO5 PO6 PO7 PO8 PO9 PO10 PO11 PO12

1 KCS062.1 3 3 3 2 3 1 3 2 1 3

2 KCS062.2 3 3 3 3 3 3

3 KCS062.3 3 3 1 3 3 1 1 1

4 KCS062.4 3 3 2 3 1 3 1 3

5 KCS062.5 3 3 3 2 3 2 3

3-High, 2-Medium, 1- Low

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 8

 Program Specific Outcomes

On successful completion of graduation degree the Engineering graduates will be able to:

• PSO1: The ability to design and develop the hardware sensor device and related interfacing software
system for solving complex engineering problem.
• PSO2: The ability to understanding of Inter disciplinary computing techniques and to apply them in
the design of advanced computing .
• PSO 3:. The ability to conduct investigation of complex problem with the help of technical,
managerial, leadership qualities, and modern engineering tools provided by industry sponsored
laboratories.
• PSO 4: The ability to identify, analyze real world problem and design their solution using artificial
intelligence ,robotics, virtual. Augmented reality ,data analytics, block chain technology and cloud
computing.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 9

 CO-PSO Mapping

CO/PSO PSO1 PSO2 PSO3 PSO4

CO1 3 1 3 3

CO2 3 1 3 2

CO3 3 2 1

CO4 3 2 3 2

CO5 3 2 3 1

3-High, 2-Medium, 1- Low

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 10

 Prerequisites
Prerequisites:
• Basic understanding of images.
• Understanding of image sharpening and image smoothing.
• Understanding of types of noise in any image.

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 11
 Content

Mono-modal/multimodal
image registration;
Global/local registration;
Transform and similarity measures for registration;
Intensity/pixel interpolation.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 12

 Program Educational Objectives

The Program Educational Objectives (PEOs) of B.Tech (Computer science & Engineering) program are established and are
listed as follow:

• PEO 1: To have an excellent scientific and engineering breadth so as to comprehend, analyze, design and provide
sustainable solutions for real-life problems using state-of-the-art technologies.

• PEO 2: To have a successful career in industries, to pursue higher studies or to support entrepreneurial endeavors and to
face the global challenges.

• PEO 3: To have an effective communication skills , professional attitude, ethical values and a desire to learn specific
knowledge in emerging trends, technologies for research, innovation and product development and contribution to
society.

• PEO 4: To have life-long learning for up-skilling and re-skilling for successful professional career as engineer, scientist,
entrepreneur and bureaucrat for betterment of society.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 13

 Result Analysis

Subject Name Code Result

Image processing KCS 062 100%

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 14

 Introduction (CO4)
• Image registration is an image processing technique used to align
multiple scenes into a single integrated image. It helps overcome
issues such as image rotation, scale, and skew that are common when
overlaying images.
• Image registration is often used in medical and satellite imagery to
align images from different camera sources. Digital cameras use
image registration to align and connect adjacent images into a single
panoramic image

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 15

 Registering aerial photos using point mapping.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 16

 Automatic registration on multimodal medical images.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 17

 Image registration

• Image registration is the process that overlays two or more images

from different sources taken at different times and angles.
• The image registration process is an automated or manual
operation that attempts to discover matching spots between two
photos and spatially align them to minimize the desired error, i.e. a
uniform proximity measurement between two images. Medical
sciences, remote sensing, and computer vision all use image
registration.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 18

 Introduction

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 19

 Introduction
• Multimodal Image Registration Using Mutual Information
• Registration of multispectral/multisensory pictures is a difficult task. In general,
such pictures have varying grey level properties, and basic approaches based on
area correlation cannot be easily used.
• It is discovered that the image’s entropy remains constant even though the
histogram changes. Even after randomly shuffling the image’s pixels, the entropy of
the image stays constant. Natural photographs also have less ambiguity.
• In a realistic image, the pixel’s value is likely to be quite near to that of some of its
neighbours. As a result, this reliance decreases overall entropy.
• Drawbacks
• When pictures have a poor resolution or the overlap region is narrow, mutual
information causes misregistration.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 20

 How does image registration work?
• How does image registration work?
Alignment can be looked at as a simple coordinate transform. The algorithm works as
follows:
• Convert both images to grayscale.
• Match features from the image to be aligned, to the reference image and store the
coordinates of the corresponding key points. Keypoints are simply the selected few points
that are used to compute the transform (generally points that stand out), and descriptors
are histograms of the image gradients to characterize the appearance of a keypoint. In this
post, we use ORB (Oriented FAST and Rotated BRIEF) implementation in the OpenCV
library, which provides us with both key points as well as their associated descriptors.
• Match the key points between the two images. In this post, we use BFMatcher, which is a
brute force matcher. BFMatcher.match() retrieves the best match, while
BFMatcher.knnMatch() retrieves top K matches, where K is specified by the user.
• Pick the top matches, and remove the noisy matches.
• Find the homomorphy transform.
• Apply this transform to the original unaligned image to get the output image.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 21

 How does image registration work?

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 22

 How does image registration work?

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 23

 How does image registration work?

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 24

 How does image registration work?
import cv2
import numpy as np
# Open the image files.
img1_color = cv2.imread("align.jpg") # Image to be aligned.
img2_color = cv2.imread("ref.jpg") # Reference image.
# Convert to grayscale.
img1 = cv2.cvtColor(img1_color, cv2.COLOR_BGR2GRAY)
img2 = cv2.cvtColor(img2_color, cv2.COLOR_BGR2GRAY)
height, width = img2.shape
# Create ORB detector with 5000 features.
orb_detector = cv2.ORB_create(5000)
# Find keypoints and descriptors.
# The first arg is the image, second arg is the mask
# (which is not required in this case).
kp1, d1 = orb_detector.detectAndCompute(img1, None)
kp2, d2 = orb_detector.detectAndCompute(img2, None)
# Match features between the two images.
# We create a Brute Force matcher with
# Hamming distance as measurement mode.
matcher = cv2.BFMatcher(cv2.NORM_HAMMING, crossCheck = True)
12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 25
)
 How does image registration work?
# Match the two sets of descriptors.
matches = matcher.match(d1, d2)
# Sort matches on the basis of their Hamming distance.
matches.sort(key = lambda x: x.distance)
# Take the top 90 % matches forward.
matches = matches[:int(len(matches)*0.9)]
no_of_matches = len(matches)
# Define empty matrices of shape no_of_matches * 2.
p1 = np.zeros((no_of_matches, 2))
p2 = np.zeros((no_of_matches, 2))
for i in range(len(matches)):
p1[i, :] = kp1[matches[i].queryIdx].pt
p2[i, :] = kp2[matches[i].trainIdx].pt
# Find the homography matrix.
homography, mask = cv2.findHomography(p1, p2, cv2.RANSAC)
# Use this matrix to transform the
# colored image wrt the reference image.
transformed_img = cv2.warpPerspective(img1_color, homography, (width, height))
# Save the output.
cv2.imwrite('output.jpg', transformed_img
12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 26
 Pixel Based Method
For registration, a cross-correlation statistical methodology is employed in this procedure. It
is frequently used for template matching or pattern recognition, which involves finding the
location and orientation of a template or pattern in an image. Cross-correlation is a measure
of similarity or a match metric.
For example, the two-dimensional normalised cross-correlation function assesses the
similarity for each translation for template (an image for reference) and image, where the
template is tiny in comparison to the image.
If the template fits the image, the cross-correlation will be at its peak. Because the measure
might be influenced by local picture intensity, the cross-correlation should be adjusted.
Drawbacks
The key disadvantages of correlation approaches are the flatness of the similarity measure
maximum (owing to the self-similarity of the pictures) and the high processing complexity.
The maximum can be successfully sharpened by pre-processing or by applying edge or
vector correlation.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 27

 Summarizing the image registration

Image registration is a necessary step in integrating, fusing, and

evaluating data from numerous sensors (sources). It has a wide
range of applications in medical sciences, computer vision, and
remote sensing.
Image registrations with complicated nonlinear distortions, multi-
modal registration, and registrations of occluded images, among
other things, contribute to the robustness of the techniques
belonging to the hardest tasks in the current environment.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 28

 Transformation

To register two images, a transformation must be found so that each

point in one image can be mapped to a point in the second. This
mapping must "optimally" align the two images where optimality
depends on what needs to be matched.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 29

 Transformation

To register two images, a transformation must be found so that each point in

one image can be mapped to a point in the second. This mapping must
"optimally" align the two images where optimality depends on what needs to
be matched.
If the transformation needed is global but not affine, the primary approach
uses feature point mapping to define a polynomial transformation.
the techniques which use the simplest local transformation based on
piecewise interpolation are described. In the most complex cases, where the
registration technique must determine local transformation

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 30

 Global /Local Transformation

Transformations used to align two images may be global or local. A global

transformation is given by a single equation which maps the entire image.
Examples are the affine, projective, perspective and polynomial
transformations. Local transformations map the image differently depending
on the spatial location and are thus much more difficult to express succinctly.
The important distinction that needs to be understood is between global/local
transformations and methods, global/local distortions and global/local
computations.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 31

 Transform & similarity measures for registration(CO4)

The task of determining the best spatial transformation for the

registration of images can be broken down into three major
components
1. feature space
2. similarity metric
3. search space and strategy

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 32

 Transform & similarity measures for registration(CO4)

Feature Space The first step in registering two images is to decide upon the
feature space to use for matching. This may be the image itself, but other
common feature spaces include: edges, contours, surfaces, other salient
features such as corners, line intersections, and points of high curvature,
statistical features such as moment invariants or centroids, and higher level
structural and syntactic descriptions. The feature space is a fundamental
aspect of almost all computer vision tasks and influences:

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 33

 Transform & similarity measures for registration(CO4)

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 34

 Transform & similarity measures for registration(CO4)

Similarity Measure The second step made in designing or choosing a registration

method is the selection of a similarity measure. This step is closely related with
the selection of the matching feature since it measures the similarity between
these features. The intrinsic structure, i.e., the invariance properties of the
image are extracted by either the feature space or through the similarity
measure. Typical similarity measures for image or feature values are cross-
correlation with or without pre-filtering (e.g., matched filters or statistical
correlation), sum of absolute differences (for better efficiency), and Fourier
invariance properties such as phase correlation. Using curves and surfaces as a
feature space requires measures such as sum of squares of differences
between nearest points.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 35

 Transform & similarity measures for registration(CO4)
Similarity Measure
Structured or syntactic methods have measures highly dependent on their
properties. For example, the minimum change of entropy between "random"
graphs is used as a similarity criteria by (Wong 851 for noisy data in structural
pattern recognition.
The choice of similarity metric is one of the most important elements of how
the registration transformation is determined. Given the search space of
possible transformations, the similarity metric may be used to find the
parameters of the final registration transformation.
For cross-correlation or the sum of the absolute differences, the transformation
is found at the peak value.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 36

 Transform & similarity measures for registration(CO4)
Search Space:
The model underlying each registration technique determines the
characteristics of the search space. Models can be classified as allowing either
global or local transformations since this directly influences the size and
complexity of the search space.
Global methods are typically either a search for the allowable transformation
which maximizes some similarity metric or a search for the parameters of the
transformation, typically a low order polynomial which fit matched control
points.
This search space exploits knowledge of the imaging process and its effects on
three dimensional structures.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 37

 Intensity/pixel interpolation (CO4)

Interpolation works by using known data to estimate values at unknown points.

Image interpolation works in two directions, and tries to achieve a best
approximation of a pixel's intensity based on the values at surrounding pixels.
Common interpolation algorithms can be grouped into two categories: adaptive
and non-adaptive. Adaptive methods change depending on what they are
interpolating, whereas non-adaptive methods treat all pixels equally. Non-
adaptive algorithms include: nearest neighbor, bilinear, bicubic, spline, sinc,
lanczos and others. Adaptive algorithms include many proprietary algorithms in
licensed software such as: Qimage, PhotoZoom Pro and Genuine Fractals

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 38

 pixel interpolation (CO4)

Pixel interpolation algorithms are crucial for processes involving geometric

transformations, with the exception of non-interpolating procedures such
as Integer Resample (pixel binning) and FastRotation (rotate ±90°, 180° and
specular transformations). The available pixel interpolation algorithms are
directly selectable by the user in standard tools such
as Resample, Rotation, DynamicCrop and StarAlignment. They are also used
internally by the PixInsight Core application and many tools and processes

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 39

 pixel interpolation (CO4)

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 40

 pixel interpolation (CO4)

Nearest Neighbor Interpolation

This method is the simplest technique that re samples the pixel values
present in the input vector or a matrix. In MATLAB, ‘imresize’ function
is used to interpolate the images.
In next page, pictorial representation depicts that a 3x3 matrix is
interpolated to 6x6 matrix. The values in the interpolated matrix are
taken from the input matrix (i.e) no new value is added.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 41

 pixel interpolation (CO4)

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 42

 pixel interpolation (CO4)

MATLAB CODE:

%3x3 Matrix
A = zeros(3,3);
A(:)=[10,2,9,4,18,14,22,7,25];

display(A); %Before Interpolation

C = imresize(C,[6,6],'nearest');
display(C);

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 43

 Transformation

the registration methods are helpful in many research areas

i) Computer Vision and Pattern Recognition - for numerous different tasks such
as segmentation, object recognition, shape reconstruction, motion tracking,
stereomapping and character recognition
ii) Medical Image Analysis - including diagnostic medical imaging such as tumor
detection and disease localization, and biomedical research including
classification of microscopic images of blood cells, cervical smears and
chomosomes, and
iii) Remotely Sensed Data Processing - for civilian and military applications in
agriculture, geology, oceanography, oil and mineral exploration

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 44

 Need for morphological processing

► Mathematical morphology is a very powerful tool for analysing the shapes

of objects present in images.
► Theory of Mathematical morphology is based on set theory
► In this, Binary image is visualized as a set.
► Morphological operators take a binary image and a mask known as
structuring element as inputs.
► Then set operations like union, intersection etc can be applied to images.
► A set is a group of objects or elements that share some common
attributes.
► In a binary image, set is a collection of coordinates where the pixel may be
black or white.
► Morphological operators follow set theory laws of union, intersection,
complement etc

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 45

 Need for morphological processing

► Mathematical morphology is a very powerful tool for analysing the shapes

of objects present in images.
► Theory of Mathematical morphology is based on set theory
► In this, Binary image is visualized as a set.
► Morphological operators take a binary image and a mask known as
structuring element as inputs.
► Then set operations like union, intersection etc can be applied to images.
► A set is a group of objects or elements that share some common
attributes.
► In a binary image, set is a collection of coordinates where the pixel may be
black or white.
► Morphological operators follow set theory laws of union, intersection,
complement etc

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 46

 Properties of sets
► Associative law
AUBUC = (AUB)UC = AU(BUC)
AՈB ՈC = (A ՈB) ՈC= A Ո(B ՈC)
Ø Commutative law
AUB= BUA
Ø Identity law
A=B if A ᶜ B and B ᶜ A
Ø De-morgan’s law
(AUB)c = Ac Ո Bc and (A Ո B)c = Ac U Bc
Ø Set differenece
A-B = A Ո Bc = A-(A Ո B)

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 47

 cont..
► Concepts of sets can be easily extended to binary images
► In binary image, background is represented by black pixels
and foreground by white pixels
► All set theory operations can be applied to binary images
• Sets in mathematical morphology represent objects in an
image

Example
– Binary image: the elements of a set is the coordinate (x,y)
of the pixels, in Z2
– Gray-level image: the element of a set is the triple, (x, y,
gray-value), in Z3

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 48

 Operations on binary images
►Complement of binary image f(x,y) is
g(x,y) = { 1 if f(x,y) = 0 and 0 if f(x,y) = 1}
Ø Union of two binary images f and g is
h=fUg where h(x,y) = { 1 if f(x,y) =1 OR if g(x,y)=1,
otherwise 0}
Ø Intersection gives the common non zero pixels of
both images A and B. denoted by h=f Ո g and
described as
h(x,y)={1 if f(x,y) =1 AND if g(x,y)=1, otherwise 0}

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 49

 Operations on binary images
• Reflection

The reflection of a set B, denoted B , is defined as

 = {w | w = -b, for b Î B}
B

• Translation
The translation of a set B by point z = ( z1 , z2 ), denoted ( B ) Z ,
is defined as
( B) Z = {c | c = b + z , for b Î B}

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 50

 Structuring elements (SE)
• Small sets or sub-images used to probe an image under study for
properties of interest
• small binary image or a flat file that consists of coordinate values
• It is a matrix of size 3X3
• Has its origin in the centre of matrix
• It is shifted over image and at each pixel of image its elements are
compared with the set of underlying pixels.
• For basic morphological operators, Structuring element has ones,
zeroes, and don’t cares
• These element select or suppress features of a certain shape eg
removing noise from image or selecting objects with a particular
direction.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 51

 Structuring elements (SE)
Some standard structuring elements are:
1. Line segment: characterized by the length and
orientation
2. Disk: popular structuring element and can be
described for digital grids
3. Set of points: a set of coordinate points
4. Composite structuring elements: two structuring
elements that share a common origin
5. Elementary structured mask: They are square grids.
Smallest square grid is of size 2X2.

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 52

 Examples: Structuring Elements (1)

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 53

 Basic morphological operators
► Two operators: Dilation and erosion
► Dilation can be applied to binary as well as grey scale images
► Basic effect of this operator on a binary image is that it gradually
increases the boundaries of the region while the small holes present in
the image becomes smaller.
► The dilation operator takes two pieces of data as inputs.
► The first is the image which is to be dilated.
► The second is a (usually small) set of coordinate points known as
a structuring element (also known as a kernel). It is this structuring
element that determines the precise effect of the dilation on the input
image.
► Dilation is usually represented by ⊕

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 54

 Basic morphological operators

►Assume that A and B are two set of pixels

►The dilation of A by B is defined by:
►A ⊕B= Ux Є B Ax
►This means that A is translated by every point of
set B.
►An equivalent definition is:
A ⊕B= {(x,y)+(u,V): (x,y) Є A, (u,v) Є B}

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 55

 Basic morphological operators

► Eg: A consists of 4 points {(3,4),(3,5),(4,4,),(4,5)} and structuring element

is (0,1).
Thus coordinates of structuring element are {(0,0),(0,1)}.
► A ⊕B = {A+(0,0)U A+(0,1)}
► The result of A+(0,0) operation would be
(3,4)+(0,0) = (3,4) (4,4)+(0,0) = (4,4)
(3,5)+(0,0) = (3,5) (4,5)+(0,0) = (4,5)
► The result of A+(0,1) operation would be
(3,4)+(0,1) = (3,5) (4,4)+(0,1) = (4,5)
(3,5)+(0,1) = (3,6) (4,5)+(0,1) = (4,6)
Ø Final result is union of all these pixels i.e.
{(3,4),(3,5),(3,6),(4,4),(4,5),(4,6)}

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 56

 Basic morphological operators

► The presence of a pixel to the right of the origin grows a layer of pixels on the
right side of the object.
► The presence of a 1 on the left side of the origin of the structuring element
grows a layer of pixels on the left side of resultant image.
► Dilation operator is commutative i.e. A ⊕B = B ⊕A
► Also known as Minkowski addition
► It increases the size of object as well as increases the thickness of object

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 57

 Faculty Video Links, You tube & NPTEL Video Links and Online Courses
Details

You tube video-

•https://www.youtube.com/watch?v=DSGHkvQBMbs&list=PLuv3GM6-
gsE08DuaC6pFUvFaDZ7EnWGX8
•https://www.youtube.com/watch?v=T0bgf3V7u-E&list=PLuv3GM6-
gsE08DuaC6pFUvFaDZ7EnWGX8&index=3

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 58
 MCQs

1. Which of the following is the valid response when we apply a first derivative?
a) Non-zero at flat segments
b) Zero at the onset of gray level step
c) Zero in flat segments
d) Zero along ramps
Answer: c

2 Which of the following is not a valid response when we apply a second

derivative?
a) Zero response at onset of gray level step
b) Nonzero response at onset of gray level step
c) Zero response at flat segments
d) Nonzero response along the ramps
Answer: b

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 59
 Assignment 4

§ Explain the advanced Techniques for Edge Detection in detail.

§ Explain Region Based Segmentation.
§ Explain Boundary Segments.
§ Explain Point,Line and Edge Detection.
§ Explain the Use of Motion in Segmentation.

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 60
 Old Question Papers

§ Differentiate Prewitt and Sobel edge detection operators.

§ Explain region based segmentation.
§ Explain the principal approaches used in image processing
to describe the texture of a region.
§ Explain image degradation/restoration model.

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 61
 Old University Exam Question Paper

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 62
 Old University Exam Question Paper

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 63
 Old University Exam Question Paper

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 64
 Old University Exam Question Paper

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 65
 Expected Questions for University Exam

• What is image registration in digital image processing?

• Explain the Use of Motion in Segmentation in detail.
• What is the purpose of image registration?
§ What is feature-based image registration?.
§ What are the Fundamentals of Image Segmentation?

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 66
 References

Textbooks:

1. Digital Image Processing – by Rafael.C.Gonzalez & Richard E.Woods,

3rd edition, Pearson Education, 2008.

Others:

2. Digital Image Processing – by S.Shridhar, 2nd edition, Oxford Higher

Education, 2016.

MINI JAIN KCS-062 Image Processing Unit-4

12/12/24 67
 Summary

• The purpose of image segmentation is to partition an image

into meaningful regions with respect to a particular application
• Segmentation should stop when the objects of interest in an
application have been isolated.
• The segmentation is based on measurements taken from the
image and might be greylevel, colour, texture, depth or motion

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 68

12/12/24 MINI JAIN KCS-062 Image Processing Unit-4 69