code does not seem to work but is not giving any errors

I=imread('phonebox2_noisy.bmp');
a=I;
[row, col, channel]=size(a);
red=a(:,:,1);
green=a(:,:,2);
blue=a(:,:,3);
a1 = red;
a2 = green;
a3 = blue;
for i=2:1:row-1
    for j = 2:1:col-1
        
        a1(i,j)= (red(i-1,j-1)+red(i-1,j)+red(i-1,j+1)+red(i,j-1)+red(i,j)+red(i,j+1)+...
            red(i+1,j-1)+red(i+1,j)+red(i+1,j+1)); 
        
        a1=sort(a1);
        red(i,j)= a1(5);
        
        
        a2(i,j)= (green(i-1,j-1)+green(i-1,j)+green(i-1,j+1)+green(i,j-1)+green(i,j)+green(i,j+1)+...
            green(i+1,j-1)+green(i+1,j)+green(i+1,j+1)); 
        
        a2=sort(a2);
        green(i,j)= a2(5);
        
        
        a3(i,j)= (blue(i-1,j-1)+blue(i-1,j)+blue(i-1,j+1)+blue(i,j-1)+blue(i,j)+blue(i,j+1)+...
            blue(i+1,j-1)+blue(i+1,j)+blue(i+1,j+1)); 
        
        a3=sort(a3);
        blue(i,j)= a3(5);
        
    end
end
final=cat(3,red,green,blue);
figure;
imshow(a);
title('latte noisy');
figure;
imshow(final);
title('latte');

[Benjamin Thompson]

Can you post your bitmap? It runs on the coloredChips.png sample image that comes with MATLAB, though I don't know what the output is supposed to look like. Sometimes bitmaps are not 24 bit color, try using imfinfo on your image.

Answer 1

[Walter Roberson]

        a1(i,j)= (red(i-1,j-1)+red(i-1,j)+red(i-1,j+1)+red(i,j-1)+red(i,j)+red(i,j+1)+...
            red(i+1,j-1)+red(i+1,j)+red(i+1,j+1)); 
        

As written, that is a problem because red is going to be an integer data type such as uint8, and adding uint8 gives you a uint8 -- which will very likely overflow.

        a1=sort(a1);

You are doing that within the loop. So you are replacing one element of a1 at a time, and sorting the entire a1 array.

        red(i,j)= a1(5);

a1 is a 2D array the same size as red so a1(5) is likely to be the same as a1(5,1)

I get the impression that your code is attempting to do a median filter. To do a median filter "by hand" (instead of using medfilt2() ) you should not add those elements of red, you should extract them. For example,

        a1 = red(i-1:i+1, j-1:j+1);
        a1 = sort(a1(:));
        

[Lewis Laurence]

yes I'm trying to create a median algorithm. I need to include all the i,j values as it required for a paper I'm writing. Could you recommend another way, similar to the way I'm currently writing it ?

thanks

[Walter Roberson]

I would recommend just using medfilt2() .

If for some reason you can't use that, then

a1 = red(i-1:i+1, j-1:j+1);
red(i,j) = median(a1(:));

and if you are not even permitted to use median() then

a1 = red(i-1:i+1, j-1:j+1);
a1 = sort(a1(:));
red(i,j) = a1(5);

[Lewis Laurence]

thank you