Change detection using Mean ratio operator

    change detection is nothing but process of detecting difference between two different images . this can also be performed  by using Mean Ratio operator  give below 

µ1, µ2 is local mean values of images A and B.

the underlying idea of the optimal difference image is that unchanged pixels exhibit small values,whereas changed areas exhibit larger values. it show  changed region  but it doesn't enhance it.result is better then that of log ratio operator

procedure is giveb below

STEP1 : clear previous project variable, clear history , and close all the window

STEP2 : Read the two source images as A and B

STEP3: Resize the image B to the size of image A to avoid the matrix mismatch error

STEP4: convert to gray scale image to if it is color image (if dimension is 3 then it is color image )

STEP5: Fing the difference between image using Mean Ratio operator

STEP6: Display the source and Rasult image

Coding is given below

%%step1

close all

clear all

clc

%%step2

[filename pathname]=uigetfile({'*.bmp';'*.jpg';'*.png';'*.tif'},'Select  SAR image 1');

a=imread([pathname filename]);

[filename pathname]=uigetfile({'*.bmp';'*.jpg';'*.png';'*.tif'},'Select  SAR image 2');

b=imread([pathname filename]);

%%step3

b=imresize(b,[size(a,1) size(a,2)]);

%%step4

dim=ndims(a);

if(dim==3)

    a=rgb2gray(a);

end

dim1=ndims(b);

if(dim1==3)

    b=rgb2gray(b);

end

%%step5

c=mean2(im2double(a));

d=mean2(im2double(b));

m=(a/c)-(b/d);

m=im2bw(m,.5);

%%step6

figure(1),imshow(a);

title('SAR image1');

figure(2),imshow(b);

title('SAR image2');

figure(3),imshow(m);

title('Mean Ratio image');

screen shot

source image A

source image B

result image