matlab-(含教程)基于维纳滤波的图像去运动模糊算法matlab仿真

clc; clear; close all; warning off; addpath(genpath(pwd)); % Read image and convert into double img=imread('Signs_degraded.jpg'); temp_img=double(img); % Choose a random flat region in the image to identify noise standard % deviation and Variance flat=temp_img(70:93,1:80); figure,imshow(img) title('Original image') figure,imshow(uint8(flat)) title('Patch to calculate noise') Sd=std2(flat); Var=Sd^2; % Fourier transform of image to get magnitude fft=fft2(img); mag=log(abs(fft)); mag_shift=fftshift(mag); figure,imshow(mag_shift,[]) title('Shifted Magnitude image') % Identify the motion and its direction from shifted magnitude plot H = fspecial('motion',35,90); % Temporary matrix with same size of image temp=zeros(600,600); % Place detected motion at center of temporary matrix [l n]=size(H); center_temp=round(length(temp)/2); temp(center_temp-round(l)/2:center_temp+(round(l)/2)-1,center_temp-round(n)/2:center_temp+(round(n)/2)-1)=H; % Shifted Magnitude of the temporary matrix fft_temp=fft2(temp); mag_temp=(abs(fft_temp)); magshift_temp=fftshift(mag_temp); figure,imshow(magshift_temp,[]) title('Shifted magnitude Wiener') % Multiply image magnitude with magnitude of temporary matrix I=mag_shift.*magshift_temp; figure,imshow(I,[]) title('image+wiener') % Obtain the output image NSR = 10/var(temp_img(:)); Z = deconvwnr(temp_img,H,NSR); figure,imshow(Z,[]) title('Output image')