Enhanced Image Fusion Algorithm Centered on Pyramidal Organization

(1)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 5, Issue 9, September 2015)

147

Enhanced Image Fusion Algorithm Centered on Pyramidal

Organization

Sandeep Singh

1

, Navneet Bawa

2

1_{M.Tech. Scholar,}2_{Associate Professor, Department of CSE, PTU Regional Centre ACET, Amritsar, India} Abstract-- The pyramidal algorithm’s outcomes are

superior over the five layer algorithm of wavelet transform. The Discrete wavelet transform fundamentally based upon the mathematical basis of fourier transform. In this transform the size of the window is static as well in time window and the frequency window .As a result this analysis has improved resolution but worse time resolution in low frequency band and vice-versa. But PCA technique used in pyramidal algoritham is based on mathematical decomposition of images.

Keywords - PCA, fusion , filter , blur , panchromatic

I. INTRODUCTION

Image fusion is a method to combine two relevant images. So the resulting image that is attained has to be more instructive. In computer vision, Multisensor Image fusion is the method of combining relevant information from two or more images into a single image .The resultant image will be more informative than any of the input images[1]. In remote sensing applications, the increasing availability of space borne sensors gives a motivation for different image fusion algorithms. Several situations in image processing need high spatial and high spectral resolution in a single image. Most of the available equipment is not capable of providing such data convincingly. Image fusion techniques allow the integration of dissimilar information sources. The fused image can have complementary spatial and spectral resolution characteristics. However, the standard image fusion techniques can distort the spectral information of the multispectral data however merging .In satellite imaging, two kinds of images are available. The panchromatic image acquired by satellites is transmitted with the maximum resolution available and the multispectral data are transmitted with coarser resolution[2]. This will usually be two or four times lower. At the receiver station, the panchromatic image is merged with the multispectral data to convey more information[6] .Numerous procedures exist to implement image fusion. The very basic one is the high pass filtering technique. Later techniques are based on Discrete Wavelet Transform, uniform rational filter bank, and Laplacian pyramid.

II. LITRATURE SURVEY

Wang,W et al.(2011) [9] This paper presented a simple and efficient algorithm for multi-focus image fusion, which used a multiresolution signal decomposition scheme called Laplacian pyramid method. The method mainly composed of three steps. Firstly, the Laplacian pyramids of each source image are deconstructed separately, and then each level of new Laplacian pyramid is fused by adopting different fusion rules. To the top level, it adopts the maximum region information rule; and to the rest levels, it adopts the maximum region energy rule. Finally, the fused image is obtained by inverse Laplacian pyramid transform. Two sets of images are applied to verify the fusion approach proposed and compared it with other fusion approaches. By analyzing the experimental results, it showed that this method has good performance, and the quality of the fused image is better than the results of other methods.

Yang,Y et al.(2010)[10] A novel wavelet-based approach for medical image fusion is presented, which is developed by taking into not only account the characteristics of human visual system (HVS) but also the physical meaning of the wavelet coefficients. After the medical images to be fused are decomposed by the wavelet transform, different-fusion schemes for combining the coefficients are proposed: coefficients in low-frequency band are selected with a visibility-based scheme, and coefficients in high-frequency bands are selected with a variance based method.

(2)

International Journal of Emerging Technology and Advanced Engineering

148 III. PCA(PRINCIPAL COMPONENT ANALYSIS)

PCA is a mathematical tool which transforms a number of correlated variables into a number of uncorrelated variables. The PCA is used widely in image compression and image grouping. The PCA involves a mathematical procedure that transforms a number of correlated variables into a number of uncorrelated variables called principal components[4]. It computes a compact and optimal description of the data set. The first principal 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 principal module is taken to be along the direction with the maximum variance. The second principal module is constrained to lie in the subspace perpendicular of the first. Inside this Subspace, this component points the direction of maximum variance[5]. The third principal component is taken in the maximum variance direction in the subspace perpendicular to the first two and so on. The PCA is also called as Karhunen-Loeve transform or the Hostelling transform. The PCA does not have a static set of basis vectors similar FFT, DCT and wavelet etc. and its basis vectors depend on the data set.

3.1 Gaussian Pyramid

A Gaussian pyramid is a procedure used in image processing, particularly in texture synthesis. The technique includes building a series of images which are weighted down using a Gaussian average (Gaussian blur) and scaled down[8]. When this procedure is used multiple times, it creates a stack of successively smaller images, with every pixel containing a local average that corresponds to a pixel neighborhood on a lower level of the pyramid.

3.2 Laplacian Pyramid

Laplacian pyramid methodology holds dissimilar level of original image .These levels are attained recursively by filtering the low level image with low-pass filter .Before applying this methodology have to apply Gaussian pyramid for filtering every level of image by using low-pass filter and down sampling is done and as the level rises image is getting smaller and smaller[3]. The equation of attaining upper level of Gaussian pyramid from the lower level is as follows:

GK = [ ѡ * GK-1]↓2

WHERE W IS LOW PASS FILTER THAT WE USE

1.Principle of laplacian pyramid

One powerful and pellucid organization used to depict picture with multi-determination is the picture pyramid suggested by Burt and Adel child in 1983. The fundamental rule of this system is to deteriorate the first picture into bits of sub-pictures with diverse spatial resolutions through some numerical operations. The Laplacian pyramid is acquired from the Gaussian pyramid, which is a multi-scale representation acquired through a recursive low-pass sifting and demolition[7]. Thus, the Laplacian pyramid deterioration is partitioned into two phases: the first is Gaussian pyramid disintegration; the second is from Gaussian pyramid to Laplacian pyramid.

IV. METHODLOGY

Algorithm used in proposed work is:

1. As we know that for fusing images we have to select two images which we want to fuse and the images which are to be chosen can be blurred images. So, choose the first image for fusing.

2. Now choose the second blurred image of your choice from the any folder respectively.

3. For fusing, the Gaussian pyramids of both the images are needed to be known. For that apply Gaussian Pyramids on both the images.

4. Create the different pyramids or level of both the images.

5. The Gaussian pyramid created will further help in creating the Laplacian Pyramids of the images selected. So, with the help of Gaussian Pyramids create Laplacian Pyramid of both the images.

6. From the Laplacian pyramid created the different levels of first image, select the top level of the pyramid for the first image selected.

7. Similarly, select the top level of the Laplacian Pyramid created for the second image selected. 8. The image fusing is done using PCA (Principal

Component Analysis), so using this concept based on PCA of the top level of Laplacian Pyramid.

9. As now the blurred part of both the images are removed and a new image is formed which is not blur, so now reconstruct original image from Laplacian Pyramid.

(3)

International Journal of Emerging Technology and Advanced Engineering

149 V. RESULTS

Entropy:

[image:3.612.322.573.186.607.2]

Entropy is an index to evaluate the information size confined in an image. If the value of entropy becomes higher after fusing, it indicates that the information growths and the fusion performances are enhanced as shown in Table 1.

Table 1: Entropy Evaluation

0 1 2 3 4 5 6 7 8 9

1 2 3 4 5

Ent

ropy

V

al

ue

No. of Images Entropy Evaluation

Existing Technique

Proposed Technique

Mean Squared Error(MSE) :

Mean square error is quantity of image quality index. Larger value of mean square error means that the image is poor quality as shown in Table 2.

Table 2

Images

Existing

techniques

Proposed

techniques

1 1312

943

2 2627

1727

3 1178

596

4

784

390

5

874

484

Standard Deviation(SD) :

Standard Deviation measures the contrast in the fused image. Fused image having high contrast would have high standard deviation as shown in Table 3.

Images

Existing

techniques

Proposed

techniques

1

6.09

7.57

2

6.14

7.52

3

6.46

7.89

4

6.21

7.59

(4)

International Journal of Emerging Technology and Advanced Engineering

150

Table 3 VI. SCREEN SHOTS

[image:4.612.334.551.137.368.2]

Fig I.First Image(Uncleared Right side)

Fig II.Second Image(Uncleared Left Side)

Images

Existing

techniques

Proposed

techniques

1

34.12

58.39

2

29.31

52.06

3

31.34

56.12

4

33.54

57.12

[image:4.612.337.552.395.577.2]

(5)

International Journal of Emerging Technology and Advanced Engineering

[image:5.612.48.291.131.356.2]

151 Fig III. Final fused image by proposed system

VII. CONCLUSION

Laplacian image pyramids centered on the bilateral filter provide a good structure for image detail enhancement and manipulation. The difference images among every layer are reformed to exaggerate or reduce details at different scales in an image. Specific image compression file formats use the Adam7 algorithm or some other interlacing method. These can be seen as a kind of image pyramid.

Because those file format store the "large-scale" features first, and fine-grain details later in the file, a particular viewer showing a small "thumbnail" or on a small screen can quickly download just sufficient of the image to display it in the available pixels .So one file can support many viewer resolutions, rather than having to store or create a dissimilar file for every resolution.

REFERENCES

[1] Sukhpreet Singh(12 December 2014) Multiple Image Fusion Using Laplacian Pyramid

[2] Yang Li(7, July 2014 ) IMAGE FUSION: A NOVEL APPROACH [3] Sweta K. Shah(3, March 2014) Comparative Study of Image Fusion

Techniques based on Spatial and Transform Domain

[4] Sonali Mane(February 2014) Image Fusion of CT/MRI using DWT , PCA Methods and Analog DSP Processor

[5] Mirajkar Pradnya P.( 5, May 2012) Image Fusion Method Based On WPCA

[6] S. Zebhi(4, August 2012) IMAGE FUSION USING PCA IN CS DOMAIN

[7] Abhishek Singh(December 2012) Implementation & comparative study of different fusion techniques (WAVELET, IHS, PCA) [8] Rakesh Aditya Mitra Hirve(2014) A Comparative Analysis of

Wavelet Transform & PCA Image Fusion Techniques.

[9] Wencheng Wang (12, DECEMBER 2011) Pyramid A Multi-focus Image Fusion Method Based on Laplacian.

[10] Yong Yang (Hindawi Publishing CorporationVolume 2010.)Medical Image Fusion via an Effective Wavelet-Based Approach.