Image denoising using laplacian prior

Description

function [out, psnr] = dn_wiener2 (noisy_image, clean_image, sigman )

% denoising using laplacian prior

[wc] = fwt_image(noisy_image,6); % 6-level DWT: lowest band is 16x16
 [CA1, CH1, CV1, CD1] = get_subbands(wc);
 [CA2, CH2, CV2, CD2] = get_subbands(CA1);
 [CA3, CH3, CV3, CD3] = get_subbands(CA2);
 [CA4, CH4, CV4, CD4] = get_subbands(CA3);
 [CA5, CH5, CV5, CD5] = get_subbands(CA4);

% Reconstruction algorithm
 % ca5 .... ch3_w cv3_w cd3_w ch2_w
 % cv2_w cd2_w ch1_w cv1_w cd1_w
 % -------------------------------

ch5_w = wiener2(CH5,[5 5], sigman^2);
 cv5_w = wiener2(CV5,[5 5], sigman^2);
 cd5_w = wiener2(CD5,[5 5], sigman^2);

ch4_w = wiener2(CH4,[5 5], sigman^2);
 cv4_w = wiener2(CV4,[5 5], sigman^2);
 cd4_w = wiener2(CD4,[5 5], sigman^2);

ch3_w = wiener2(CH3,[5 5], sigman^2);
 cv3_w = wiener2(CV3,[5 5], sigman^2);
 cd3_w = wiener2(CD3,[5 5], sigman^2);

ch2_w = wiener2(CH2,[5 5], sigman^2);
 cv2_w = wiener2(CV2,[5 5], sigman^2);
 cd2_w = wiener2(CD2,[5 5], sigman^2);

ch1_w = wiener2(CH1,[5 5], sigman^2);
 cv1_w = wiener2(CV1,[5 5], sigman^2);
 cd1_w = wiener2(CD1,[5 5], sigman^2);

% reconstructed image
 % ---------------------------

ca5 = CA5;

wc = [ca5 ch5_w
 cv5_w cd5_w ];
 wc = [wc ch4_w
 cv4_w cd4_w ];
 wc = [wc ch3_w
 cv3_w cd3_w ];
 wc = [wc ch2_w
 cv2_w cd2_w ];
 wc = [wc ch1_w
 cv1_w cd1_w ];

out = iwt_image(wc,6);
 psnr = feval('psnr',clean_image,out);

https://matlab1.com/shop/matlab-code/image-denoising-via-dictionary-learning-and-structural-clustering/