ayushbaid/dipproject

Residual Image based denoising

DIPProject

• The use of residuals in image denoising
• Implementation of a paper by Brunet et. al.

Code Environment Requirements

• This code will run on Linux OS only due to the use of file specific paths.

Code Licence and use of 3rd party toolbox

The Total Variation Minimization filter code has been used as a black box. The toolbox is located at ./toolbox_optim/.

Test Images

The test images are located in ./USC_SIPI_DATABASE

Code details

• ms1.m executes the algo over 5 test images for 6 different values of noise standard deviation and plots psnr and ssim figure for each standard deviation. Denoiser 1 is Wiener filter and Denoiser 2 is TVD

• ms2.m executes the algo over the lenna image for a s.d value mentioned in the code. Plots the residues and images for 2 iterations. Denoiser 1 is TVD and Denoiser 2 is Wiener filter.

• ms3.m Similar to ms2.m except that Denoiser 1 is Wiener filter and Denoiser 2 is TVD.