How to compare the pixel of two image?

How to turn the image into pure 0(white) and 1(black) ?

After turn into pure 0 and 1, how I sum the image so I can get total number pixel of 1?

count=0;
[x,y]=size(IM);
for i=1:x
    for j=1:y
        if IM(i,j)==1
            if IM2(i,j)==1
                count=count+1;
            end
        end
    end

I want to compare both image and get how many % similarity.

similarity= count/totalobjectpixel

1.1 is total object pixel and 1.2 is result of segmentation i get.

Answer 1

[Stephen23]

Write vectorized code, it makes you life much easier:

thresh = 0.5; % threshold for converting to B&W
A_clr = rand(5,5,3); % random image
B_clr = rand(5,5,3); % random image
A_bw = A_clr>thresh; % convert to B&W
B_bw = B_clr>thresh; % convert to B&W
idx = A_bw==B_bw;    % compare
idy = all(idx,3);    % R==G==B
out = sum(idy(:));   % count

This code assumes that the image is a gray-scale RGB. Of course if you have the images toolbox you can use the more robust im2bw to convert to a black-and-white image:

[Tan Wen Kun]

A_clr = imread('1.1.jpg'); %segment

B_clr = imread('1.2.jpg'); %original

How to superimpose the segment image to original image?