mmse_filter_cvip() - minimum mean squared error restoration filter.



new_image = mmse_filter_cvip( imageP, noise_var,  kernel_size)

Input Parameters include :

Output parameter include :


This function performs the minimum mean squared error restoration filtering operation on an input image that is noise to restore it to its originla state. MMSE filter is an adaptive filter that alters its behavior based on the local statistics. The filter mitigates the noise from the image while still retaining the image details. MMSE works best with Gaussian or Uniform noise.


1. Scott E Umbaugh. DIGITAL IMAGE PROCESSING AND ANALYSIS: Applications with MATLAB and CVIPtools, 3rd Edition.


% Read image

 imageP = imread('noise_mmse.bmp');

% Kernel size

 kernel_size = 9;

% Noise variance

 noise_var = 300;

% Call function

 new_image = mmse_filter_cvip( imageP,noise_var,kernel_size);

% Display input image

 figure;imshow(imageP);title('Input image');

% Display output images

 figure; imshow(new_image/255);title('Output image');


Author: Deependra Mishra, March 2017
Copyright © 2017-2018 Scott E Umbaugh
For updates visit CVIP Toolbox Website