[image fusion] image fusion based on NSST and PCNN with matlab code

1 brief introduction

Purpose : The fusion PET/CT/MRI Medical images , Make the resulting image contain as much information as possible, such as edge and texture features , In order to better distinguish the pathological changes , Tumors and normal tissues and organs , Provide more useful information for disease diagnosis . Method : A shear wave transform based on non down sampling is proposed (NSST) And pulse coupled neural network (PCNN) Model fusion method . First , According to the energy sum of the local area of the image , To image NSST Weighted fusion of low-frequency coefficients ; then , according to PCNN Firing times of neurons , Choose an image NSST High frequency directional coefficient ; Last , Through inverse NSST Transformation , Get the fused image . result : Respectively for 7 Group MRI/PET and CT/PET Image fusion experiment , The resulting image has a good visual effect , And mutual information , Edge similarity , Gradient similarity and spatial frequency 4 It is better than other algorithms in the comprehensive evaluation of three indexes . Conclusion : This method can adaptively capture edge and texture information , It has good fusion effect .​

2 Part of the code

clear all;close all;clc;%% NSST tool boxaddpath(genpath('shearlet'));%%A=imread('sourceimages/s02_MR.tif');  %anatomical imageB=imread('sourceimages/s02_PET.tif'); %functional image​img1 = double(A)/255;img2 = double(B)/255;img2_YUV=ConvertRGBtoYUV(img2);img2_Y=img2_YUV(:,:,1);[hei, wid] = size(img1);​% image fusion with NSST-PAPCNN imgf_Y=fuse_NSST_PAPCNN(img1,img2_Y); ​imgf_YUV=zeros(hei,wid,3);imgf_YUV(:,:,1)=imgf_Y;imgf_YUV(:,:,2)=img2_YUV(:,:,2);imgf_YUV(:,:,3)=img2_YUV(:,:,3);imgf=ConvertYUVtoRGB(imgf_YUV);​F=uint8(imgf*255);figure;subplot(131);imshow(A);title(' chart 1')subplot(132);imshow(B);title(' chart 2')​subplot(133);imshow(F);title(' Fusion map ');imwrite(F,'results/fused.tif');​​

3 Simulation results

4 reference

[1] Tianjuanxiu , Liuguocai . be based on NSST Transform sum PCNN Medical image fusion method based on [J]. Chinese Journal of medical physics , 2018, 35(8):7.

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 .