[image fusion] image fusion based on directional discrete cosine transform and principal component analysis with matlab code

1 brief introduction

Abstract ：

The purpose of image fusion is to merge relevant information from multiple images right into a single image. In this paper, by conducting the review it has been discovered that the majority of the existing techniques are based upon transform domain therefore it could results in some artifacts which might decrease the execution of the transform based vision fusion techniques. Moreover it is already been discovered that the issue of the uneven illuminate has already been neglected in the absolute most of existing focus on fusion. Therefore to overcome these issues, a fresh method which integrates the larger valued Alternating Current (AC) coefficients calculated in iterative block level principal component averaging (IBLPCA) domain base fusion with illuminate normalization and fuzzy enhancement has been proposed in this paper. The experimental results show the efficiency of proposed algorithm over existing work.

​2 Part of the code

%function[] = DDCTIFdemo()% DDCT (Directional Discrete Cosine Transform) based image fusion - demo% VPS Naidu, MSDF Lab, CSIR-NAL, March 2014% Reference: "? %       Journal of Optics, Vol. 43, No.1, pp.48-61, March 2014.​%%close all;clear all;home;​%%dflg = 1; % 0: no display OR 1: displayaflag = 1; % 1: Average, 2: max rule OR 3: energy rulebs = 4; %[4 8 16 32 64 128 256];  block size​%%% insert imagesimt = im2double(imread('saras9t.jpg'));im1 = im2double(imread('saras91.jpg'));im2 = im2double(imread('saras92.jpg'));​if dflg == 1    figure(1);    subplot(121);imshow(im1);title('image to be fused - im1');    subplot(122);imshow(im2);title('image to be fused - im2');    pause(1);end​%%mode = [0 1 3 4 5 6 7 8]; % directional modelmode = length(mode);​%% if aflag == 1 % fusion by DDCT average rule    h1 = waitbar(0,'Please wait...');    for j=1:lmode        imf1{j} = DDCTIFav(im1,im2,bs,mode(j));        waitbar(j/lmode,h1);    end    close(h1);end​%%if aflag == 2 % fusion by DDCT max rule    h1 = waitbar(0,'Please wait...');    for j=1:lmode        imf1{j} = DDCTIFmax(im1,im2,bs,mode(j));        waitbar(j/lmode,h1);    end    close(h1);end​%%if aflag == 3 % fusion by DDCT energy rule    h1 = waitbar(0,'Please wait...');    for j=1:lmode        imf1{j} = DDCTIFek(im1,im2,bs,mode(j));        waitbar(j/lmode,h1);    end    close(h1);end%%% fusion by PCAimf = fuse_pca(imf1{1},imf1{2},imf1{3},imf1{4},imf1{5},imf1{6},imf1{7},imf1{8});​%%% Performance evaluation metrics[RMSE,SF] = im_fuse_per_eval(imt,imf);​%%% display resultsif dflg == 1    figure(2);    subplot(121); imshow(imf); title('fused image');        imd = imt-imf;    subplot(122); imshow(imd); title('error image');endfprintf('\nRMSE :  %3f2', RMSE);fprintf('\nSF   :  %3f2', SF);fprintf('\n\n');%%

3 Simulation results

 

4 reference

[1] Zhao Xiaolei . be based on IHS Image fusion of transform and principal component analysis transform [J]. Science and technology and Engineering , 2010(20):4.

[2] Kaur P . Hybrid PCA-DCT Based Image Fusion For Medical Images. 

About bloggers ： Good at intelligent optimization algorithms 、 Neural networks predict 、 signal processing 、 Cellular automata 、 The image processing 、 Path planning 、 UAV and other fields Matlab Simulation , relevant matlab Code problems can be exchanged by private letter .

Some theories cite network literature , If there is infringement, contact the blogger to delete .