Proceedings of the 2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 
INTERNATIONAL JOURNAL OF COMPUTER ENGINEERING & 
17 – 19, July 2014, Mysore, Karnataka, India 
TECHNOLOGY (IJCET) 
ISSN 0976 – 6367(Print) 
ISSN 0976 – 6375(Online) 
Volume 5, Issue 9, September (2014), pp. 20-26 
© IAEME: www.iaeme.com/IJCET.asp 
Journal Impact Factor (2014): 8.5328 (Calculated by GISI) 
www.jifactor.com 
20 
 
IJCET 
© I A E M E 
PROPERTY BASED FUSION FOR MULTIFOCUS IMAGES 
Pradeep kumar H S1, Shruthi2, Vaishnavi N L3, Ranjini M G4 
1Asst. professor, Dept of CSE, MIT Mysore/VTU Belgaum, India 
2, 3, 4Student –UG, Dept of CSE, MIT Mysore/VTU Belgaum, India 
 
ABSTRACT 
Here we concentrated on multifocus image fusion by adopting discrete wavelet transform 
(DWT) technique. This paper highlights on pixel based image fusion, which yields highly focused 
image as outcome. This paper aims to implement pixel-level image fusion based on mathematical 
and wavelet transform fusion methods and find out their capacity to improve spatial and spectral 
information. For this purpose different methods such as Averaging method, Minimum method, 
Maximum method are used. Pixel level methods are affected by blurring effect which directly affects 
the quality of the image; hence in order to overcome blurring effect, we are measuring the quality of 
the image. Based on that value, quality of the image can be improved. Before applying any of the 
fusion methods the primary task is to register the images. 
Keywords: Image Registration, Image fusion, Contrast, Entropy, Visibility. 
I. INTRODUCTION 
The images that are already captured at different focus levels are fused to produce an image 
that appears sharp everywhere. The fusion of image requires registration of the image prior to fusion. 
Image fusion means the combining of two images into a single image that has the maximum 
information content without producing details that are non-existent in the given images. Image 
Fusion provides a mechanism to improve the quality of information from a set of images. Important 
applications of the fusion of images include medical imaging, microscopic imaging, remote sensing, 
computer vision, and robotics [1]. 
Image registration is the process of spatially aligning two or more images of a scene. The 
process brings into correspondence individual pixels in the images. Therefore, given a point in one 
image, the registration process will determine the positions of the same point in all the images. For 
the image registration we are considering two geometric transformations like rotation and scaling.
Proceedings of the 2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 
17 – 19, July 2014, Mysore, Karnataka, India 
Registration used here is landmark based image fusion which is concerned with the concept called as 
spatial transformation. 
21 
 
The actual fusion process can take place at different levels of information representation, a 
generic categorization of these levels are signal, pixel, feature and symbolic level. In order to achieve 
pixel level image fusion we are considering simple minimum, simple maximum, simple average 
method. For property measurements we are considering image properties like entropy, contrast, 
kurtosis, visibility in ordered to increase the information that is particular application. 
Combination of registration and image fusion technique improves the performance as 
compared to use of individual fusion method as a result the partially blurred image can be converted 
into a highly detailed image, facilitating automated processes that rely on image details to understand 
a scene. 
II. RELATED WORK 
The lowest possible technique in image fusion is the pixel level, is also called as non linear 
method, in which intensity values of sources image are used for merging the images. The next level 
is feature level, which operates on characteristics such as size, shape, edge etc. At further level, 
called decision level fusion, deals with symbolic representation of images[3]. 
An efficient pixel level Multifocus image fusion algorithm based on artificial neural networks 
is proposed. The fusion method originated from human visual perception principle is suitable to 
merge images with diverse focuses. Two spatially registered images with different focuses are 
decomposed into several blocks. Then three features reflecting the clear level of every block are 
calculated. Finally artificial neural networks are used to recognize the clear level of corresponding 
block to decide which blocks should be used to construct the fusion result[4]. 
III. PROPOSED METHOD 
Fusion process consists of two basic steps: 
1. Image Registration 
2. Image Fusion 
Image registration, which brings the input images to spatial alignment, and image fusion 
combines the image features (intensities, colors, etc) in the area of frame overlap. Image registration 
works in four steps. 
Feature detection 
Attention is paid on the effect of fusion on corners, line intersections, edges, contours, closed 
boundary, regions, etc. whether they are clearly detected. For further processing, these features can 
be represented by their point representatives (distinctive points, line endings, centers of gravity), 
called in the literature control points. 
Feature matching 
Features detected in the image that is to be registered are compared with those detected in the 
reference image. Various feature descriptors and similarity measures along with spatial relationships 
among the features are used here.
Proceedings of the 2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 
17 – 19, July 2014, Mysore, Karnataka, India 
22 
Transform model estimation 
 
The type and parameters of the so-called mapping functions, aligning the sensed image with 
the reference image, are estimated. The parameters of the mapping functions are computed by means 
of the established feature correspondence. 
Image re-sampling  transformation 
The sensed image is transformed by means of the mapping functions. Image values in non-integer 
coordinates are estimated by an appropriate interpolation technique. 
The problem faced in spatial domain method can be very well handled by Transform domain 
image fusion methods. The fusion methods that can be used are discrete wavelet transform, complex 
wavelet transform, curvelet transform and Laplacian pyramid based methods[6], in this paper we 
mainly focus on DWT based fusion method. 
Pixel-level image fusion means fusion at the lowest processing level referring to the merging 
of measured physical parameters. It generates a fused image in which each pixel is determined from 
a set of pixels of various images, and serves to increase the useful information content of an image. 
Pixel Level Image fusion method can be divided into two groups: 
1. Spatial domain fusion method 
2. Transform domain fusion 
Spatial domain fusion method directly deals with pixels of input images. We assume the 
input images are spatially and temporally aligned, semantically equivalent and radio metrically 
calibrated. The fusion methods such as simple maximum, simple minimum and averaging based 
methods fall under spatial domain approaches. In transform domain method image is first transferred 
into frequency domain. The fusion method such as DWT falls under transform domain method. 
3.1 SIMPLE AVERAGE 
It is a well-documented fact that regions of images that are in focus tend to be of higher pixel 
intensity. This algorithm is a simple way of obtaining an output image with all regions in focus. In 
this method the resultant fused image is obtained by taking the value of the pixel P(i, j) of each 
image and the values are 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 which is given in equation. This is 
repeated for all pixel values. 
F (i, j) = {(X (i, j)) + (Y (i, j))}/2 (1) 
Where, ( X (i , j)) and ( Y ( i, j) ) in the equation (1) are two registered input images and F (i, j) is 
fused image. 
3.2 SIMPLE MAXIMUM 
Greater the pixel value more is the focus of the image. Thus this algorithm chooses the in-focus 
regions from each input image by choosing the greatest value for each pixel, resulting in highly 
focused output. The value of the pixel P (i, j) of each image is taken and compared with each other. 
The greatest pixel value is assigned to the corresponding pixel of the output image which is given in 
equation. 
F( i ,j) = MAX{(X( I ,j )),( Y(i,j))} (2) 
Where, (X(i,j)), (Y(i,j)) in the equation (2) are input images and F(i,j) is fused image.
Proceedings of the 2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 
17 – 19, July 2014, Mysore, Karnataka, India 
23 
3.3 SIMPLE MINIMUM 
 
Lesser pixel value provides critical information about the image. In this method, the resultant 
fused image is obtained by selecting the minimum intensity of corresponding pixels from both the 
input images which is given in equation. 
F( i,j) = MIN{(X(i,j )),( Y( i,j))} (3) 
Where, (X( i,j)), (Y(i ,j)) in the equation (3) are input images and F ( i,j) is the fused image. 
3.4 WAVELET BASED IMAGE FUSION 
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 good localization 
properties make them better basis for analysis of signals with discontinuities. 
The wavelet transform decomposes the image into low-high, high-low, high-high spatial frequency 
bands at different scales and the low-low band at the coarsest scale which is shown in fig:3.1. The L-L 
band contains the average image information whereas the other bands contain 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. 
Figure 3.1: Wavelet Based Image Fusion 
The wavelets-based approach is appropriate for performing fusion tasks for the following reasons:- 
(1) It is a multi scale (multi resolution) approach well suited to manage the different image 
resolutions. Useful in a number of image processing applications including the image fusion. 
(2) The discrete wavelets transform (DWT) allows the image decomposition in different kinds of 
coefficients preserving the image information. Such coefficients coming from different images 
can be appropriately combined to obtain new coefficients so that the information in the original 
images is collected appropriately. 
(3) Once the coefficients are merged the final fused image is achieved through the inverse discrete 
wavelets transform (IDWT), where the information in the merged coefficients is also 
preserved.
Proceedings of the 2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 
17 – 19, July 2014, Mysore, Karnataka, India 
= − ( ) log P i 
(4) 
F m n μ N 
−   a μ 
s (7) 
= − 
24 
IV. PERFORMANCE EVALUATION 
 
The main goal of any image fusing process is that it should preserve all valid and useful 
pattern information from the source images and must provide the most detailed and reliable 
information possible, while at the same time it should not introduce any artifacts that could interfere 
with subsequent analysis. Thus output of fused image will be evaluated in terms of Entropy, 
Visibility, Kurtosis, and Contrast 
4.1 ENTROPY 
The contrast-enhancement performance is measured by calculating the second-order entropy. 
Entropy is an index to evaluate the information quantity contained in an image. If the value of 
entropy becomes higher after fusing, it indicates that the information increases and the fusion 
performances are improved. 
[ ( ) ] 
255 
E  P i 
0 2 
i 
= 
Where, p in the equation (4) is the probability distribution of each level of image, and its 
value are {p0, p1. . . p−1}[1][3][7]. 
4.2 VISIBILITY 
This feature is inspired from the human visual system, and is defined as 
VI= 1 
1 1 
| ( , ) | 
+ 
= n 
= 
M 
m 
(5) 
Where, μ in the equation (5) is the mean intensity value of the image and µ is a visual 
constant ranging from 0.6 to 0.7[4][5]. 
4.3 KURTOSIS 
Kurtosis is a measure of the degree of peakness of a histogram and is represented by K.It is 
calculated as follows. 
s2 
T = S i ( i - μT )2p (6) 
Where,s2 
T in the equation (6) is the total variance of levels 
K= (μ4/s4)-3 
K = 0, the curve is normal. 
If K is positive, the curve is more peaked. If K is negative, the curve is more flat topped. 
4.4 CONTRAST 
A measure of the clarity with which objects or regions in the image can be identified 
estimated by the value of the standard deviation of pixel intensities in the image. 
2 
[ ( ) ] 
1 1 
, 
1 
 
= 
 
= 
m 
i 
n 
j 
f i j M 
mn 
Where, M in the equation (7) is the mean of the image.[5].
Proceedings of the 2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 
17 – 19, July 2014, Mysore, Karnataka, India 
Method Entropy Contrast Visibility Kurtosis 
DWT 7.7349 0.2810 190.3346 20.4227 
MIN 7.3963 62.4051 95.7323 23.4828 
MAX 7.7582 74.2615 170.313 12.1345 
AVG 4.8991 42.6342 200.1764 249.5814 
Method Entropy Contrast Visibility Kurtosis 
DWT 6.7056 0.2219 285.8140 10.9263 
MIN 6.9507 39.4405 59.3326 5.9643 
MAX 7.0025 40.2220 48.5587 4.7579 
AVG 6.5411 38.2088 44.4673 133.4365 
25 
 
Table 1: Performance comparision of multifocus image fusion 
Table 2: Performance comparision of MRI-CT image fusion 
Type of 
Image 
MRI-CT 
Type of 
Image 
MULTI 
FOCUS 
V. EXPERIMENT AND RESULT 
The input images used in all the algorithms were registered images of equal size. Images are 
first undergone through the registration process then they are fed to fusion. 
In each iteration, values of the image properties such as visibility, contrast, kurtosis and 
entropy are documented. Experiment is repeated for various types of images such as multifocal, 
multisensor and MRI-CT (medical images) images fig 5.1 and 5.2. 
5.1. Multifocus Images 
. 
Figure 5.1: (a) (b) (c) (d) 
5.2. Medical Images 
Figure 5.2: (a) (b) (c ) (d)
Proceedings of the 2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 
17 – 19, July 2014, Mysore, Karnataka, India 
26 
CONCLUSION 
 
Although selection of fusion algorithm is problem dependent but this review results that 
spatial domain provide high spatial resolution. But spatial domain have image blurring problem. The 
wavelet transforms is the very good technique for the image fusion provide a high quality spectral 
content. But a good fused image have both quality so the combination of DWT  spatial domain 
fusion method (like avg, min, max) fusion algorithm improves the performance as compared to use 
of individual DWT and pixel level fusion algorithm. 
REFERENCES 
[1]. Deepak Kumar Sahu and M.P.Parsai, Different Image Fusion Techniques –A Critical 
Review,International Journal of Modern Engineering Research (IJMER) www.ijmer.com 
Vol. 2, Issue. 5, Sep.-Oct. 2012. 
[2]. Wang Anna, Wu Jie, Li Dan and Chen Yu, Research on Medical Image Fusion Based on 
Orthogonal Wavelet Packets Transformation Combined with 2v-SVM, IEEE/ICME 
International Conference on Complex Medical Engineering 2007. 
[3]. S.Chithra, J.B.Bhattacharjee and B.Thilakavathi, Image Fusion using Re-modified SPIHT for 
fused image. 
[4]. Shutao LI, Yaonan WANG and Boris Lohmann, Multi-focus Image Fusion using Artificial 
Neural Networks. 
[5]. Amina Saleem, Azeddine Beghdadi and Boualem Boashash. Image fusion-based contrast 
enhancement. EURASIP Journal on Image and Video Processing 2012. 
[6]. R.Maruthi and Dr.K.Sankarasubramanian, Multi Focus Image Fusion Based on the 
Information Level in the Regions of the Images. Journal of Theoretical and Applied 
Information Technology (JATIT). 
[7]. Kusum Rani and ReechaSharma, Study of Different Image fusion Algorithm,International 
Journal of Emerging Technology and Advanced Engineering (ISSN 2250-2459, 
ISO 9001:2008 Certified Journal, Volume 3, Issue 5, May 2013). 
[8]. Dr.S.S.Bedi, Mrs.JyotiAgarwal and PankajAgarwal, Image Fusion Techniques and Quality 
AssessmentParameters for Clinical Diagnosis: A ReviewInternational Journal of Advanced 
Research in Computer and Communication Engineering Vol. 2, Issue 2, February 2013. 
[9]. Shaveta Mahajan and Arpinder Singh, A Comparative Analysis of Different Image Fusion 
Techniques, IPASJ International Journal of Computer Science (IIJCS) Volume 2, Issue 1, 
January 2014.

Property based fusion for multifocus images

  • 1.
    Proceedings of the2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 INTERNATIONAL JOURNAL OF COMPUTER ENGINEERING & 17 – 19, July 2014, Mysore, Karnataka, India TECHNOLOGY (IJCET) ISSN 0976 – 6367(Print) ISSN 0976 – 6375(Online) Volume 5, Issue 9, September (2014), pp. 20-26 © IAEME: www.iaeme.com/IJCET.asp Journal Impact Factor (2014): 8.5328 (Calculated by GISI) www.jifactor.com 20 IJCET © I A E M E PROPERTY BASED FUSION FOR MULTIFOCUS IMAGES Pradeep kumar H S1, Shruthi2, Vaishnavi N L3, Ranjini M G4 1Asst. professor, Dept of CSE, MIT Mysore/VTU Belgaum, India 2, 3, 4Student –UG, Dept of CSE, MIT Mysore/VTU Belgaum, India ABSTRACT Here we concentrated on multifocus image fusion by adopting discrete wavelet transform (DWT) technique. This paper highlights on pixel based image fusion, which yields highly focused image as outcome. This paper aims to implement pixel-level image fusion based on mathematical and wavelet transform fusion methods and find out their capacity to improve spatial and spectral information. For this purpose different methods such as Averaging method, Minimum method, Maximum method are used. Pixel level methods are affected by blurring effect which directly affects the quality of the image; hence in order to overcome blurring effect, we are measuring the quality of the image. Based on that value, quality of the image can be improved. Before applying any of the fusion methods the primary task is to register the images. Keywords: Image Registration, Image fusion, Contrast, Entropy, Visibility. I. INTRODUCTION The images that are already captured at different focus levels are fused to produce an image that appears sharp everywhere. The fusion of image requires registration of the image prior to fusion. Image fusion means the combining of two images into a single image that has the maximum information content without producing details that are non-existent in the given images. Image Fusion provides a mechanism to improve the quality of information from a set of images. Important applications of the fusion of images include medical imaging, microscopic imaging, remote sensing, computer vision, and robotics [1]. Image registration is the process of spatially aligning two or more images of a scene. The process brings into correspondence individual pixels in the images. Therefore, given a point in one image, the registration process will determine the positions of the same point in all the images. For the image registration we are considering two geometric transformations like rotation and scaling.
  • 2.
    Proceedings of the2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 17 – 19, July 2014, Mysore, Karnataka, India Registration used here is landmark based image fusion which is concerned with the concept called as spatial transformation. 21 The actual fusion process can take place at different levels of information representation, a generic categorization of these levels are signal, pixel, feature and symbolic level. In order to achieve pixel level image fusion we are considering simple minimum, simple maximum, simple average method. For property measurements we are considering image properties like entropy, contrast, kurtosis, visibility in ordered to increase the information that is particular application. Combination of registration and image fusion technique improves the performance as compared to use of individual fusion method as a result the partially blurred image can be converted into a highly detailed image, facilitating automated processes that rely on image details to understand a scene. II. RELATED WORK The lowest possible technique in image fusion is the pixel level, is also called as non linear method, in which intensity values of sources image are used for merging the images. The next level is feature level, which operates on characteristics such as size, shape, edge etc. At further level, called decision level fusion, deals with symbolic representation of images[3]. An efficient pixel level Multifocus image fusion algorithm based on artificial neural networks is proposed. The fusion method originated from human visual perception principle is suitable to merge images with diverse focuses. Two spatially registered images with different focuses are decomposed into several blocks. Then three features reflecting the clear level of every block are calculated. Finally artificial neural networks are used to recognize the clear level of corresponding block to decide which blocks should be used to construct the fusion result[4]. III. PROPOSED METHOD Fusion process consists of two basic steps: 1. Image Registration 2. Image Fusion Image registration, which brings the input images to spatial alignment, and image fusion combines the image features (intensities, colors, etc) in the area of frame overlap. Image registration works in four steps. Feature detection Attention is paid on the effect of fusion on corners, line intersections, edges, contours, closed boundary, regions, etc. whether they are clearly detected. For further processing, these features can be represented by their point representatives (distinctive points, line endings, centers of gravity), called in the literature control points. Feature matching Features detected in the image that is to be registered are compared with those detected in the reference image. Various feature descriptors and similarity measures along with spatial relationships among the features are used here.
  • 3.
    Proceedings of the2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 17 – 19, July 2014, Mysore, Karnataka, India 22 Transform model estimation The type and parameters of the so-called mapping functions, aligning the sensed image with the reference image, are estimated. The parameters of the mapping functions are computed by means of the established feature correspondence. Image re-sampling transformation The sensed image is transformed by means of the mapping functions. Image values in non-integer coordinates are estimated by an appropriate interpolation technique. The problem faced in spatial domain method can be very well handled by Transform domain image fusion methods. The fusion methods that can be used are discrete wavelet transform, complex wavelet transform, curvelet transform and Laplacian pyramid based methods[6], in this paper we mainly focus on DWT based fusion method. Pixel-level image fusion means fusion at the lowest processing level referring to the merging of measured physical parameters. It generates a fused image in which each pixel is determined from a set of pixels of various images, and serves to increase the useful information content of an image. Pixel Level Image fusion method can be divided into two groups: 1. Spatial domain fusion method 2. Transform domain fusion Spatial domain fusion method directly deals with pixels of input images. We assume the input images are spatially and temporally aligned, semantically equivalent and radio metrically calibrated. The fusion methods such as simple maximum, simple minimum and averaging based methods fall under spatial domain approaches. In transform domain method image is first transferred into frequency domain. The fusion method such as DWT falls under transform domain method. 3.1 SIMPLE AVERAGE It is a well-documented fact that regions of images that are in focus tend to be of higher pixel intensity. This algorithm is a simple way of obtaining an output image with all regions in focus. In this method the resultant fused image is obtained by taking the value of the pixel P(i, j) of each image and the values are 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 which is given in equation. This is repeated for all pixel values. F (i, j) = {(X (i, j)) + (Y (i, j))}/2 (1) Where, ( X (i , j)) and ( Y ( i, j) ) in the equation (1) are two registered input images and F (i, j) is fused image. 3.2 SIMPLE MAXIMUM Greater the pixel value more is the focus of the image. Thus this algorithm chooses the in-focus regions from each input image by choosing the greatest value for each pixel, resulting in highly focused output. The value of the pixel P (i, j) of each image is taken and compared with each other. The greatest pixel value is assigned to the corresponding pixel of the output image which is given in equation. F( i ,j) = MAX{(X( I ,j )),( Y(i,j))} (2) Where, (X(i,j)), (Y(i,j)) in the equation (2) are input images and F(i,j) is fused image.
  • 4.
    Proceedings of the2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 17 – 19, July 2014, Mysore, Karnataka, India 23 3.3 SIMPLE MINIMUM Lesser pixel value provides critical information about the image. In this method, the resultant fused image is obtained by selecting the minimum intensity of corresponding pixels from both the input images which is given in equation. F( i,j) = MIN{(X(i,j )),( Y( i,j))} (3) Where, (X( i,j)), (Y(i ,j)) in the equation (3) are input images and F ( i,j) is the fused image. 3.4 WAVELET BASED IMAGE FUSION 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 good localization properties make them better basis for analysis of signals with discontinuities. The wavelet transform decomposes the image into low-high, high-low, high-high spatial frequency bands at different scales and the low-low band at the coarsest scale which is shown in fig:3.1. The L-L band contains the average image information whereas the other bands contain 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. Figure 3.1: Wavelet Based Image Fusion The wavelets-based approach is appropriate for performing fusion tasks for the following reasons:- (1) It is a multi scale (multi resolution) approach well suited to manage the different image resolutions. Useful in a number of image processing applications including the image fusion. (2) The discrete wavelets transform (DWT) allows the image decomposition in different kinds of coefficients preserving the image information. Such coefficients coming from different images can be appropriately combined to obtain new coefficients so that the information in the original images is collected appropriately. (3) Once the coefficients are merged the final fused image is achieved through the inverse discrete wavelets transform (IDWT), where the information in the merged coefficients is also preserved.
  • 5.
    Proceedings of the2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 17 – 19, July 2014, Mysore, Karnataka, India = − ( ) log P i (4) F m n μ N − a μ s (7) = − 24 IV. PERFORMANCE EVALUATION The main goal of any image fusing process is that it should preserve all valid and useful pattern information from the source images and must provide the most detailed and reliable information possible, while at the same time it should not introduce any artifacts that could interfere with subsequent analysis. Thus output of fused image will be evaluated in terms of Entropy, Visibility, Kurtosis, and Contrast 4.1 ENTROPY The contrast-enhancement performance is measured by calculating the second-order entropy. Entropy is an index to evaluate the information quantity contained in an image. If the value of entropy becomes higher after fusing, it indicates that the information increases and the fusion performances are improved. [ ( ) ] 255 E P i 0 2 i = Where, p in the equation (4) is the probability distribution of each level of image, and its value are {p0, p1. . . p−1}[1][3][7]. 4.2 VISIBILITY This feature is inspired from the human visual system, and is defined as VI= 1 1 1 | ( , ) | + = n = M m (5) Where, μ in the equation (5) is the mean intensity value of the image and µ is a visual constant ranging from 0.6 to 0.7[4][5]. 4.3 KURTOSIS Kurtosis is a measure of the degree of peakness of a histogram and is represented by K.It is calculated as follows. s2 T = S i ( i - μT )2p (6) Where,s2 T in the equation (6) is the total variance of levels K= (μ4/s4)-3 K = 0, the curve is normal. If K is positive, the curve is more peaked. If K is negative, the curve is more flat topped. 4.4 CONTRAST A measure of the clarity with which objects or regions in the image can be identified estimated by the value of the standard deviation of pixel intensities in the image. 2 [ ( ) ] 1 1 , 1 = = m i n j f i j M mn Where, M in the equation (7) is the mean of the image.[5].
  • 6.
    Proceedings of the2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 17 – 19, July 2014, Mysore, Karnataka, India Method Entropy Contrast Visibility Kurtosis DWT 7.7349 0.2810 190.3346 20.4227 MIN 7.3963 62.4051 95.7323 23.4828 MAX 7.7582 74.2615 170.313 12.1345 AVG 4.8991 42.6342 200.1764 249.5814 Method Entropy Contrast Visibility Kurtosis DWT 6.7056 0.2219 285.8140 10.9263 MIN 6.9507 39.4405 59.3326 5.9643 MAX 7.0025 40.2220 48.5587 4.7579 AVG 6.5411 38.2088 44.4673 133.4365 25 Table 1: Performance comparision of multifocus image fusion Table 2: Performance comparision of MRI-CT image fusion Type of Image MRI-CT Type of Image MULTI FOCUS V. EXPERIMENT AND RESULT The input images used in all the algorithms were registered images of equal size. Images are first undergone through the registration process then they are fed to fusion. In each iteration, values of the image properties such as visibility, contrast, kurtosis and entropy are documented. Experiment is repeated for various types of images such as multifocal, multisensor and MRI-CT (medical images) images fig 5.1 and 5.2. 5.1. Multifocus Images . Figure 5.1: (a) (b) (c) (d) 5.2. Medical Images Figure 5.2: (a) (b) (c ) (d)
  • 7.
    Proceedings of the2nd International Conference on Current Trends in Engineering and Management ICCTEM -2014 17 – 19, July 2014, Mysore, Karnataka, India 26 CONCLUSION Although selection of fusion algorithm is problem dependent but this review results that spatial domain provide high spatial resolution. But spatial domain have image blurring problem. The wavelet transforms is the very good technique for the image fusion provide a high quality spectral content. But a good fused image have both quality so the combination of DWT spatial domain fusion method (like avg, min, max) fusion algorithm improves the performance as compared to use of individual DWT and pixel level fusion algorithm. REFERENCES [1]. Deepak Kumar Sahu and M.P.Parsai, Different Image Fusion Techniques –A Critical Review,International Journal of Modern Engineering Research (IJMER) www.ijmer.com Vol. 2, Issue. 5, Sep.-Oct. 2012. [2]. Wang Anna, Wu Jie, Li Dan and Chen Yu, Research on Medical Image Fusion Based on Orthogonal Wavelet Packets Transformation Combined with 2v-SVM, IEEE/ICME International Conference on Complex Medical Engineering 2007. [3]. S.Chithra, J.B.Bhattacharjee and B.Thilakavathi, Image Fusion using Re-modified SPIHT for fused image. [4]. Shutao LI, Yaonan WANG and Boris Lohmann, Multi-focus Image Fusion using Artificial Neural Networks. [5]. Amina Saleem, Azeddine Beghdadi and Boualem Boashash. Image fusion-based contrast enhancement. EURASIP Journal on Image and Video Processing 2012. [6]. R.Maruthi and Dr.K.Sankarasubramanian, Multi Focus Image Fusion Based on the Information Level in the Regions of the Images. Journal of Theoretical and Applied Information Technology (JATIT). [7]. Kusum Rani and ReechaSharma, Study of Different Image fusion Algorithm,International Journal of Emerging Technology and Advanced Engineering (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 5, May 2013). [8]. Dr.S.S.Bedi, Mrs.JyotiAgarwal and PankajAgarwal, Image Fusion Techniques and Quality AssessmentParameters for Clinical Diagnosis: A ReviewInternational Journal of Advanced Research in Computer and Communication Engineering Vol. 2, Issue 2, February 2013. [9]. Shaveta Mahajan and Arpinder Singh, A Comparative Analysis of Different Image Fusion Techniques, IPASJ International Journal of Computer Science (IIJCS) Volume 2, Issue 1, January 2014.