Optimizing Shrinkage Curves and Application in Image Denoising

Algorithm 2

Optimizing shrinkage curves based denoising algorithm.

Input: Noisy image and the shrinkage parameters

()    Initialize the intermediate image and adjusted image ;

()      ;

()      ;

()    for  

()    Set a counter to zeros, ;

()    Obtain an adjusted estimate, ;

()    According to Definition 1, build the patch-set , associated with image ;

()    for each patch in   do

()      According to Definition 2, obtain a ROSM ;

()      Singular value decomposition, ;

()      Find the parameter corresponding to from Table ;

()      Obtain the estimation ;

()      if  , then ;

()      if  , then ;

()      Put back into the image canvas;

()      In counter , the entries associated with pixels in is added by 1;

()    end for

()    Obtain an estimated image, obtained canvas image is entry-wisely divided by ;

()  end for

Output: The final estimated image .