ISBN-13: 978-1537313573

Proceedings of ICTPRE-2016
IMAGE FUSION OF BRAIN IMAGES USING DISCRETE WAVELET TRANSFORM
S.Azhagumeena, S.Rajakumari, SHEEBA RACHEL
UG, Department of Information Technology
Sri Sairam Engineering College
[email protected], [email protected]

ABSTRACT
The ultimate aim of the project is to develop an image that
is a combination of Scanned images of MRI (Magnetic
Resonance Imaging) and CT (Computed Tomography) of
brain. CT scan is used to view the bone and MRI scan is used
to view the tissues. therefore when a physician need to
diagnose he/she should view both the CT scan and MRI scan
at the same time side by side which is quite difficult for
human eye and this can also lead to error and mismatch. In
order to make it easier these two images can be fused to
form a single image without affecting the contents of the
image. The process of fusion of these two images is done by
means of Discrete Wavelet Transform (DWT). Thereby
making it easy for the physician to view the two images in a
single image.
INTRODUCTION
1. IMAGE FUSION
In computer vision image fusion is the process of
combining relevant information from two or more images
into a single image. The resulting image will be more
informative than any of the input images. It is also defined
as the set of methods, tools and means of using data from
two or more different images to improve the quality of
information.
In remote sensing applications, the increasing
availability of space borne sensors gives a motivation for
different image fusion algorithms. Several situations in
image processing require high spatial and high spectral
resolution in a single image. Image fusion techniques allow
the integration of different image sources.
In satellite imaging, two types of images are available.
The panchromatic image acquired by satellites is
transmitted with the maximum resolution available and the
multispectral data are transmitted with courser resolution
which is two or four times lower. At the receiver station, the
panchromatic images are merged with the multispectral
data to convey more information.
In medical imaging image fusion has become a
common term used within medical diagnostics and
treatment. The term is used when multiple images of a
patient are registered and overlaid or merged to provide
additional information. Fused images may be created from
multiple images from the same imaging modality or
combining images from multiple modalities. For accurate
diagnosis, radiologist must integrate information from
multiple image formats. Fused, anatomically consistent
images are especially beneficial in diagnosing and treating
cancer. With the advent of these new technologies
radiation oncologist can take full advantage of intensity
modulated radiation therapy (IMRT).

www.iirdem.org

1.2 IMAGE FUSION ALGORITHMS

Lens it is often not possible to get an image that contains all
relevant objects in Due to the limited focus depth of the
optical focus. To obtain an image with every object in focus
a multi-focus image fusion process is required to fuse the
images giving a better view for human or machine
perception. Pixel-based, region-based and wavelet based
fusion algorithms were implemented.
1.2.1 SIMPLE AVERAGE
It is a well-documented fact that the regions of
image that are in focus tend to be of higher pixel intensity.
Thus this algorithm is a simple way of obtaining an output
image with all regions in focus. The value of the pixel P(i,j)
of each image is taken and added. This sum is then divided
by 2 to obtain the average. The average value is assigned to
the corresponding pixel of the output image.
1.2.2 SELECT MAXIMUM
The greater the pixel values the more in focus the
image. Thus this algorithm chooses the in-focus regions
from each input image by choosing the greater value for
each pixel, resulting in highly focused output. The value of
the pixel P (i, j) of each image is taken and compared to each
other. The greatest pixel values is assigned to the
corresponding pixel
1.2.3 MULTIPLICATIVE ALGORITHM
The multiplication model combines data set by
multiplying each pixel in each band of the input data by
source data. To compensate for the increased brightness
value, the square root of the mixed data is taken.
1.2.4 SUBTRACTIVE METHOD
Subtractive Resolution Merge uses a subtractive
algorithm to sharpen multi-spectral images. It produces
highly preserved spatial and spectral resolution. It is limited
to dual sensor platforms with specific band ratios between
the high resolution panchromatic image and the low
resolution multi spectral image.
1.2.5 BROVEY ALGORITHM
Brovey algorithm is a ratio method where the data
values of each band of the MS data are divided by the sum
of the MS data set and then multiplied by the input data set.
It increases the contrast in the low and high ends of an
image histogram.
1.2.6 DISCRETE WAVELET TRANSFORM (DWT)
Wavelets are finite duration oscillatory functions
with zero average value. They have finite energy. They are
suited for analysis of transient signal. The irregularity and

91

IIRDEM 2016

ISBN-13: 978-1537313573

Proceedings of ICTPRE-2016

good localization properties make them better basis for

analysis of signals with discontinuities. Wavelets can be
3.1 WAVELET-BASED APPROACH:
Described by using two functions viz. the scaling function
f(t), also known as father wavelet and wavelet function or
mother wavelet. Mother wavelet undergoes translation
and scaling operations to give self-similar wavelet families.
The wavelet transform decomposes the image
into low-high, high-low, high-high spatial frequency bands
at different scales and low-low band at the coarsest scale.
The L-L band contains the average image information
whereas the other band contains directional information
due to spatial orientation. Higher absolute values of
wavelet coefficients in the high bands correspond to salient
features such as edges or lines.

The two input images were fused in the wavelet

domain, and an inverse transformation was applied to
produce the result. Other variations of this technique
include discrete wavelet transform, additive wavelet
decomposition, the contourlet transform, the curvelet
transform and the complex wavelet transform.
ADVANTAGE
In wavelet based transform edges becomes sharper. Images
can be compressed by setting the useless data.
Decomposition can continue until the size of the image is as
small as we need. When we want to detect the edges of the
image, we can simply set the diagonal sub-images to zero,
and then we can obtain the output image with edges
clearly.

1.2.7 PRINCIPLE COMPONENT ANALYSIS (PCA)

PCA is a mathematical tool which transforms a number of
correlated variables into a number of uncorrelated
variables. The PCA is used extensively in image compression
and image classification. The PCA involves a mathematical
procedure that transform a number of correlated variables
into a number of uncorrelated variables called principle
components. It computes a compact and optimal
description of the data set. The first principle component
accounts for as much of the variance in the data as possible
and each succeeding component accounts for as much of
the remaining variance as possible. First principle
component is taken to be along the direction with the
maximum variance. The second principle component is
constrained to lie in the subspace perpendicular of the first.
Within this Subspace, this component points the direction
in the subspace perpendicular to the first two and so on.
The PCA is also called as Karhunen-Loeve transform or the
Hotelling transform. The PCA does not have a fixed set of
basis vectors like FFT, DCT and wavelet etc. and its basis
vectors depend on the data set.

3.2 VARIATIONAL FUSION METHODS:

These methods consist of finding the optimum of
energy functional, often via standard continuous
optimization techniques, e.g., gradient descent.
DRAWBACK:
For these variation approaches, the main difficulties come
from the limitations of the optimizers.
3.3 GRADIENT-DESCENT METHOD:
It yield suboptimal solutions and have a very highcomputational load.
4. PROPOSED SYSTEM
Image fusion can be done using Discrete Wavelet Transform
through which the images can be viewed as a single image
which will reduce the noise level and computational load.
They offer a simultaneous localization in time and
frequency domain fine details in a signal can be separated
easily using wavelet transform. Wavelet theory is capable
of revealing aspects of data that other signal analysis
techniques miss the aspects like trends, breakdown points,
and discontinuities in higher derivatives and self-similarity.
Discrete wavelet transform is used for fusing multimodal
images from CT and MRI of brain images.

3. EXISTING SYSTEM
For brain disease patients have to receive both the MRI scan
images and CT scan images for their benefits.
CT (COMPUTERIZED TOMOGRAPHY): Also called X-ray
Computerized Axial Tomography make use of computer
processed combinations of many X-ray images taken from
different angles to produce cross sectional images of
specific areas of a scanned object. CT is suited for examining
bone details.

4.1 PROPOSED METHOD ADVANTAGES

The advantages of proposed method are:
1) Soft tissue and bony details are available in single image
2) Reduction in storage cost

MRI (MAGNETIC RESONANCE IMAGING): It is a medical

imaging technique used in radiology to image the anatomy
and the physiological processes of the body in both health
and disease. MRI scanners use strong magnetic fields, radio
waves and field gradients to form image of the body.

www.iirdem.org

92

IIRDEM 2016

ISBN-13: 978-1537313573

Proceedings of ICTPRE-2016

CT (Bony
details)
Patient
undergoing
radio
therapy
treatment
for brain

Fused CT
and MRI
images
MRI (soft
tissues
details)
Fig1: Block diagram of proposed method

MRI image

Registratio
n

Preprocessing

Filters image

CT image

Fused image
using
discrete
wavelet
transform

Default
registration

Preprocessing

Registration

Fig2: Block diagram of steps involved

x (n)

h0(n)

g0(n)

h1(n)

g1(n)

Analysis

x*(n)

Synthesis

Fig 3: Block diagram of sub band coding technique

5.1 INPUT IMAGES

www.iirdem.org

93

IIRDEM 2016

ISBN-13: 978-1537313573

Proceedings of ICTPRE-2016

The MR scans can be acquired using a 3-D T2 weighted pulse

sequence, and the CT scans were acquired from either
helical or axial slice CT images. No contrast should be used
in either scan. T2-weighted 3-D MR images are used
because they clearly present the cerebral fluid, occipital
lobe and parietal lobe. Finally the images are fused as sets
of 2-D images because radiologists typically view 3-D
volumes as stacks of 2-D images. The 3-D MR/CT images are
then pre-processed and registered.

SSIM =

(2 + 1 )(2 )+2
( 2 + 2 +1 )( + +2 )

where x, y, x, y, xy represent the means in the

x and y images, the variances in the x and y images and the
covariance of the two images, respectively. C1 = 0.01 and
C2 = 0.03 are positive constants.

5.2 PREPROCESSING AND REGISTRATION

Literature Survey:

The given input image are resized from resolution to

256*256. In preprocessing filtering of the image is done
were median filter is applied to remove salt and pepper
noise from the medical image. The filtered images are given
as the input to registration. Image registration is the
process of transforming different sets of data into one
coordinate. system Registration is mapping between two
images both spatially and with respect to intensity. Two
types of Image registration are Multimodal and
Monomodal registration. Multi-modal registration
methods are often used in medical imaging as images of a
subject are frequently obtained from different scanners.
Under Multimodal registration, default registration method
is used It is intensity based (minimize intensity difference
over entire image). After registration the images are fused.

The purpose of the literature survey is to give the brief

overview and also to establish complete information about
the reference papers. The goal of literature survey is to
completely specify the technical details related to the main
project in a concise and unambiguous manner.
1) TITLE- Multiresolution-based image fusion with additive
Wavelet decomposition
AUTHOR- J. Nunez, X. Otazu, O. Fors, A. Prades, V. Pal`a,
and R. Arbiol
The standard data fusion methods may not be
satisfactory to merge a high-resolution panchromatic image
and a low-resolution multispectral image because they can
distort the spectral characteristics of the multispectral data.
In this paper, we developed a technique, based on multi
resolution wavelet decomposition, for the merging and
data fusion of such images. This method presented here
consists of adding the wavelet coefficients of the highresolution image to the multispectral (low resolution) data.
We have studied several possibilities concluding
that the method which produces the best results consists in
adding the high order coefficients of the wavelet transform
of the panchromatic image to the intensity component
(defined as L =R+G+B3 ) of the multispectral image. The
method is, thus, an improvement on standard intensityhue-saturation (IHS or LHS) mergers. We used the `a trous
algorithm which allows using a dynamic wavelet to merge
non dynamic data in a simple and efficient scheme. We
used the method to merge SPOT and LANDSAT(TM) images.
The technique presented is clearly better than the IHS and
LHS mergers in preserving both spectral and spatial
information.

5.3 DISCRETE WAVELET TRANSFORM

Wavelet transform in two dimensions is used in
image processing. For two-dimensional wavelet transform,
we need one two-dimensional scaling function (x, y) and
three two-dimensional wavelet 1(x, y), 2 (x, y) and 3(x,
y) each is the product of a one-dimensional scaling function
and corresponding wavelet.
These are as follows
(x, y) = (x) (y)
1(x, y) = (x) (y)
2 (x, y) = (x) (y)
3(x, y) = (x) (y)
For image processing, these functions measure the
variation of intensity for the image along different
directions 1 measures variation along columns, 2
measures variations along rows 3 measures variations
along diagonals.
5.3.1 SUBBAND CODING
In sub band coding, an image is decomposed into a
set of band limited components, called sub bands. It can be
proved that we can perfectly reconstruct the original image
using the sub bands. Because the bandwidth of the
resulting sub bands y0 (n) and y1 (n) are smaller than the
original x n, it can be down sampled without loss of
information. Reconstruction of the original image is
accomplished by up sampling, filtering, and summing the
individual sub bands.

2) TITLE- Multimodality medical image fusion based on

multiscale geometric
Analysis of contourlet transforms
AUTHOR- L. Yang, B. L. Guo, and W. Ni
As a novel multiscale geometric
analysis tool, contourlet has shown many advantages over
the conventional image representation methods. In this
paper, a new fusion algorithm for multimodal medical
images based on contourlet transform is proposed. All
fusion operations are performed in contourlet domain.

6. SSIM (STRUCTURAL SIMILARITY) CALCULATION

www.iirdem.org

94

IIRDEM 2016