Why do I get an error for my erosion code of an image?

My code is having an error, but I do not know where it is coming from. Thanks in advance!

function [img_out] = myErosion(img_in, N)

if img_in >=(2*N+1)^2

    for k=0:N

        for i=1:img_in(1)

            for j=1:img_in(2)

                if img_in(i+k,j)== 0

                    if img_in(i-k,j)== 0

                        if img_in(i,j+k)== 0

                            if img_in(i,j-k)== 0

                                if img_in(i+k,j+k)== 0

                                    if img_in(i+k,j-k)== 0

                                        if img_in(i-k,j+k)== 0

                                            if img_in(i-k,j-k)== 0

                                                img_out(i,j) = 0;

                                            end
                                        end
                                    end
                                end
                            end
                        end
                    end
                end
            end
        end
    end
end
end

Answer 1

[Image Analyst]

Again, you're just not thinking this through. Erosion is a local min. So do that instead of whatever you're trying to do. Try this: (untested)

[rows, columns] = size(img_in);
% N is an odd number that is the window size
n2 = floor(N/2);
img_out = zeros(size(img_in));
for c = n2 : columns - n2
  for r = n2 : rows - n2
    subImage = img_in(r-n2:r+n2, c-n2 : c+n2);
    img_out(r, c) = min(subImage(:));
  end
end

[Christopher]

I understand what the problem is getting at now! however, don't the indices have to be real positive integers? of r-n2:r+n2

[Image Analyst]

Yeah. You might have to add or subtract 1 here and there to make sure it doesn't go outside the range 1:rows or 1:columns.

[Christopher]

Thanks I got it to work but it's only eroding only one time through so there are still black specks on my image. How do I fix that?

function [img_out] = myErosion(img_in, N)

[rows, columns] = size(img_in);

% N is an odd number that is the window size

n2 = floor(N/2);

img_out = zeros(size(img_in));

for c = n2 : (columns - n2)

for r = n2 : (rows - n2)

    subImage = img_in(((r+1)-(n2)):((r-1)+(n2)), ((c+1)-(n2)) : ((c-1)+(n2)));

    img_out(r, c) = min(subImage(:));

end

end

end

[Image Analyst]

You can use a larger window size. Try 9 or 19 or 51 or whatever it takes to get rid of noise.