Sei sulla pagina 1di 4

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 5, May 2013)

Study of Different Image fusion Algorithm


Kusum Rani1, Reecha Sharma2
1 2
M. Tech Student, Assistant Professor, UCoE, Punjabi University, Patiala, India.
Abstract-- Image fusion is a technique that integrate In transform domain method image is first transferred
complimentary details from multiple input images such that in to frequency domain. The fusion method such as DWT
the new image give more information and more suitable for fall under transform domain method .
the purpose of human visual perception. This paper Some of these pixel-level algorithms are described
presents a review on some of the image fusion techniques
below:
(simple average, simple minimum, simple maximum, PCA,
DWT). Comparison of all the techniques concludes the a) Simple Maximum Method: In this method, the resultant
better approach for its future research. fused image is obtained by selecting the maximum
intensity of corresponding pixels from both the input
Keywords-- Image Fusion, Discrete Wavelet Transform
(DWT), Principal Component Analysis (PCA).
image.
( ) ∑ ∑ ( ) ( )......(1)
I. INTRODUCTION A(i,j), B(i,j) are input images and F(i,j) is fused image.
Image Fusion is a process of combining the relevant
b) Simple Minimum Method: In this method, the resultant
information from a set of images of the same scene into a
fused image is obtained by selecting the minimum
single image and the resultant fused image will be more
intensity of corresponding pixels from both the input
informative and complete than any of the input images.
image.
Input images could be multi sensor, multimodal,
multifocus or multi temporal. There are some important ( ) ∑ ∑ ( ) ( ).......(2)
requirements for the image fusion process [1]: A(i,j), B(i,j) are input images and F(i,j) is fused image.
- The fused image should preserve all relevant c) Simple Average Method: In this method the resultant
information from the input images fused image is obtained by taking the average intensity of
- The image fusion should not introduce artifacts which corresponding pixels from both the input image.
can lead to a wrong diagnosis
( ) ( )⁄
One of the important pre-processing steps for the ( ) ........(3)
fusion process is image registration.Image registration is A(i,j), B(i,j) are input images and F(i,j) is fused image.
the process of transforming different sets of data into one
coordinate system.Image fusion find application in the d) Weighted Average Method: in this method the
area of navigation guidance, object detection and resultant fused image is obtained by taking the weighted
recognition, medical diagnosis, satellite imaging for average intensity of corresponding pixels from both the
remote sensing, military and civilian surveillance, input image.
etc.Image fusion algorithms can be categorized into ( ) ∑ ∑ ( ) ( ) ( ).......(4)
different levels: pixel, feature, and decision levels.Pixel
A(i,j), B(i,j) are input images and F(i,j) is fused image
level fusion works directly on the pixels of source images
and W is weight factor.
while feature level fusion algorithms operate on features
extracted from the source images. e) Principal Component Analysis (PCA) method:
Principal Component Analysis is a sub space method,
II. IMAGE FUSION ALGORITHMS which reduces the multidimensional data sets into lower
Image fusion method can be divided into two groups dimensions for analysis. This method determines the
–1.Spatial domain fusion method 2.Transform domain weights for each source. image using the eigenvector
fusion corresponding to the largest eigen value of the covariance
Spatial domain fusion method directly deal with matrix of each source image.
pixels of input images. The fusion methods such as
simple maximum, simple minimum, averaging, principal
component analysis (PCA) and IHS based methods fall
under spatial domain approaches.

288
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 5, May 2013)
Stepwise procedure for PCA method is given below: let s(n1,n2) is input image with size N1xN2 then
scaling and wavelet function are

( ) ⁄ ∑ ∑ ( ) ( )

....(5)

( ) ⁄ ∑ ∑ ( ) ( )

....(6)
General process of image fusion using DWT
Step 1. Implement Discrete Wavelet Transform on both
the input image to create wavelet lower decomposition.
Step 2. Fuse each decomposition level by using different
fusion rule .
Step 3. Carry Inverse Discrete Wavelet Transform on
fused decomposed level, which means to reconstruct the
image, while the image reconstructed is the fused image
F.

e) Discrete Wavelet Transform Method :Wavelet


transforms are multi-resolution image decomposition tool
that provide a variety of channels representing the image
feature by different frequency subbands at multi-scale. It
is a famous technique in analyzing signals. When
decomposition is performed, the approximation and
detail component can be separated 2-D Discrete Wavelet
Transformation (DWT) converts the image from the Fig 2. Wavelet Based image fusion
spatial domain to frequency domain. The image is
divided by vertical and horizontal lines and represents the
first-order of DWT, and the image can be separated with
four parts those are LL1, LH1, HL1 and HH1. [10]

Fig 3. Source Images


a)CT image b) MRI image

Fig 1. Wavelet decomposition

289
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 5, May 2013)
Resultant fused images are shown in figure 2 [11] III. IMAGE QUALITY METRICS
The general requirement of an image fusing process is
to preserve all valid and useful information from the
source images, while at the same time it should not
introduce any distortion in resultant fused image.
Performance measures are used essential to measure the
possible benefits of fusion and also used to compare
results obtained with different algorithms.
a) Peak Signal to Noise Ratio: The PSNR is use to
c) Simple average d) Simple minimum calculate the similarity between two images. The
PSNR between the reference image R and the fused
image F is defined as

( ) .......(7)

For better fused image PSNR value should high.


b) Normalized Cross Correlation: The Normalized Cross
Content between the reference image R and the fused
image F is defined as
e) Simple maximum f) Simple weighted average
∑ ∑ ∑ ∑ .....(8)
c) Entropy (EN): Entropy is used to calculate the
amount of information. Higher value of entropy indicates
that the information increases and the fusion
performances are improved.
∑ .....(9)

g) PCA h) DWT

290
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 5, May 2013)
TABLE 1
COMPARISON OF DIFFERENT IMAGE FUSION TECHNIQUES

S.No Fusion technique Domain Measuring Advantage Disadvantage


parameter

1 Simple Spatial PSNR-25.48 Simple in implementation Resultant fused image


Average [6] [8] EN-7.22 is not clear

2 Simple maximum Spatial PSNR-26.86 Fused image is highly Blurring effect is more
[4] [6] EN-7.20 focused image

3 PCA [7] Spatial NC-0.998 PCA is a tools which Spectral degradation


PSNR-76.44 transforms number of is present
correlated variable into
number of uncorrelated
variables, this property can
be used in image fusion

4 DWT [5][8] Transform NC-0.998 Provide better signal to Less spatial resolution
PSNR-76.44 noise ratio

5 Combine DWT, Transform NC-0.998 Output image contained Complex method


PCA[9] PSNR-76.44 both high spatial resolution .
with high quality spectral
content

[6] Anjali Malviya, S. G. Bhirud .” Image Fusion of Digital Images”


IV. CONCLUSION International Journal of Recent Trends in Engineering, Vol 2, No.
Spatial domain image fusion techniques provide high 3, November 2009
spatial resolution. But spatial domain have image [7] V.P.S. Naidu and J.R. Raol, “Pixel-level Image Fusion using
blurring problem. The Wavelet transforms provide a high Wavelets and Principal Component Analysis”. Defence Science
Journal, Vol. 58, No. 3, May 2008, pp. 338-352 Ó 2008,
quality spectral content. Combination of DWT and PCA DESIDOC
provide better performance and improve the image fusion [8] Chetan K. SolankiNarendra M. Patel, “Pixel based and Wavelet
quality. based Image fusion Methods with their Comparative Study”.
National Conference on Recent Trends in Engineering &
REFERENCES Technology. 13-14 May 2011
[1] Kirankumar Y., Shenbaga Devi S. “ Transform-based medical [9] M .Chandana,S. Amutha, and Naveen Kumar, “ A Hybrid Multi-
image fusion” Int. J. Biomedical Engineering and Technology, focus Medical Image Fusion Based on Wavelet Transform”.
Vol. 1, No. 1, 2007 101 International Journal of Research and Reviews in Computer
[2] ShivsubramaniKrishnamoorthy, K P Soman “Implementation and Science (IJRRCS) Vol. 2, No. 4, August 2011, ISSN: 2079-2557
Comparative Study of Image Fusion Algorithms” International [10] Hong ZHENG ,Dequan ZHENG, Yanxiang Sheng “ Study on the
Journal of Computer Applications (0975 – 8887) Volume 9–No.2, Optimal Parameters of Image Fusion Based on Wavelet
November 2010 Transform” Journal of Computational Information
[3] Sukhdipkaur, Kamaljitkaur “Study and Implementation of Systems6:1(2010) 131-137,January 2010.
ImageFusion Methods” International Journal of Electronics and [11] Dr. M. Sumathi , R. Barani “Qualitative Evaluation of Pixel Level
Computer Science Engineering ,2010 Image Fusion Algorithms” IEEE transaction on Pattern
[4] YufengZheng, Edward A. Essock and Bruce C. Hansen, “An Recognition, Informatics and Medical Engineering, March 21-23,
Advanced Image Fusion Algorithm Based on Wavelet Transform 2012
– Incorporation with PCA and Morphological Processing”
[5] Chetan K. SolankiNarendra M. Patel, “Pixel based and Wavelet
based Image fusion Methods with their Comparative Study”.
National Conference on Recent Trends in Engineering &
Technology. 13-14 May 2011

291

Potrebbero piacerti anche