Hi Paul,
You can use the MATLAB ‘conv2’ function to compute the Max-Delta of Sliding 3x3 Window by subtracting the center pixel with all 8 neighbours and taking the absolute difference.
Using an appropriate kernel can be helpful while dealing with the ‘conv2’ function, the following kernels can be used for the calculation:
% Kernels of neighbours are as follows: Right, Left, Up, Down, Top-Left, Top-Right, Bottom-Left, Bottom-Right
kernels = {
[0 0 0; 0 -1 1; 0 0 0],
[0 0 0; 1 -1 0; 0 0 0],
[0 1 0; 0 -1 0; 0 0 0],
[0 0 0; 0 -1 0; 0 1 0],
[1 0 0; 0 -1 0; 0 0 0],
[0 0 1; 0 -1 0; 0 0 0],
[0 0 0; 0 -1 0; 1 0 0],
[0 0 0; 0 -1 0; 0 0 1]
};
To handle edge cases and ensure accurate convolution, you can add the padding to the input matrix as follows:
A = magic(3)
padPt = mean([min(A(:)), max(A(:))]);
% Pad the matrix with mean value
Apad = padarray(A, [1, 1], padPt, 'both');
Refer to the following example code snippet to apply Convolution to compute differences and obtain Max-Delta result:
diffs = zeros([size(Apad), 8]);
% Apply each kernel with conv2 and store the absolute difference
for k = 1:8
diffs(:,:,k) = abs(conv2(Apad, kernels{k}, 'same'));
end
% Using the maximum difference across all 8 neighbors
Ctmp = max(diffs, [], 3);
% Output after removing padding
res = Ctmp(2:end-1, 2:end-1)
Refer to the following MathWorks Documentation to learn more about ‘conv2’ function: https://www.mathworks.com/help/releases/R2024a/matlab/ref/conv2.html
I hope this helps you in resolving your query.
