%Let me explain with the same example. %To the input image Iimg applied the 'Salt & Pepper noise' with the noise ratio of 20% and named as 'Nimg'. %Now for the 'Nimg' i applied median filter and named as Mimg. %Now i need to find the 'Mean Square Error' MSE comparison between Mimg and Nimg. clear all close all clc; Iimg=imread('moon.tif'); [r c]=size( Iimg ); d=ndims( Iimg ); if d == 3 Iimg =rgb2gray( Iimg ); end Iimg=double( Iimg ); Iimg= Iimg /225; Nimg=imnoise( Iimg ,'salt & pepper',0.2); Mimg=medfilt2( Nimg ,[5 5]); clc; Diff= Nimg - Mimg ; MSE= sum(sum(Diff.* Diff)) / (r * c); fprintf('\n\nThe Mse value is: %d',MSE1); subplot(1,2,1);imshow( Nimg );title('Noised image'); subplot(1,2,2);imshow( Mimg );title('Denoised image'); RESULT: The Mse value is: 2.521604e-002 i.e, MSE ~ 0.0252
Comments
Post a Comment