【图像融合】基于平移不变小波变换实现CT图像融合matlab源码

 

登录后复制

%% Compressed sensing MRI reconstruction based SIDWT

 clear all
 close all
 clc
 currentFolder = pwd;
 addpath(genpath(currentFolder));
%%  data
 load im1
 figure,imshow(abs(im1)); title('fully-sampled image') % fully sampled image.
 [row,column]=size(im1); 
 %% masks
 load mask1 
 paramSIDWT.Fu=Fu_downsample(mask1,row,column);
 paramSIDWT.U=mask_downsample(mask1,row,column);   
 y1 = paramSIDWT.Fu * im1;   
 zerofilling = ifft2c(paramSIDWT.U'*y1);
 figure,imshow(abs(ifft2c(paramSIDWT.U'*y1))); title('undersampled k-space data') 
%% SIDWT paramters  analysis
 filterType = 'Daubechies';   filterSize = 4;
 wavScale = 4;  complex_Yes = 1;
 paramSIDWT.psi = SIDWT(filterType,filterSize,wavScale,complex_Yes); % define the shift-invariant wavelets operator
 paramSIDWT.lamda=5*1e3;
 tic;im1_Rec=solver_ADMC(y1,mask1,paramSIDWT);time_SIDWT_CS1=toc % solve the L1 norm minimization problem with ADMC
 Evaluation_ana = Evaluation_CS_MRI(im1,im1_Rec)
 %%   synthesis
 for lamda=10:20:1e3
     paramSIDWT.lamda=lamda;
     alpha=solver_ADMC_syn(y1,mask1,paramSIDWT);
     Rim = paramSIDWT.psi' *alpha;
     Evaluation_syn = Evaluation_CS_MRI(im1,Rim)
 end
 Evaluation_zerofilling = Evaluation_CS_MRI(im1,zerofilling)
 figure,imshow(abs(im1_Rec));title('Reconstructed image')

• 1.
• 2.
• 3.
• 4.
• 5.
• 6.
• 7.
• 8.
• 9.
• 10.
• 11.
• 12.
• 13.
• 14.
• 15.
• 16.
• 17.
• 18.
• 19.
• 20.
• 21.
• 22.
• 23.
• 24.
• 25.
• 26.
• 27.
• 28.
• 29.
• 30.
• 31.
• 32.
• 33.
• 34.