%this is the image
%[0 0 0 0 0 0;
% 0 0 0 0 0 0;
% 0 0 0 1 0 0;
% 0 0 0 0 0 0;
% 0 0 0 0 0 0];
% FilterCoefficientes [0 1 0] is to not smooth the image
% next is the answer from MATLAB

 M = cornermetric([0 0 0 0 0 0;0 0 0 0 0 0;0 0 0 1 0 0;0 0 0 0 0 0;0 0
0 0 0 0], "FilterCoefficients", [0 1 0])
M =
         0         0         0   -0.0400         0         0
         0         0         0   -0.0400         0         0
         0         0   -0.0400         0   -0.0400   -0.0400
         0         0         0   -0.0400         0         0
         0         0         0   -0.0400         0         0

% The calculation to go from 1 to -0.04 is:
%SensitivityFactor = 0.04 (optional parameter to the function)
%A = horizontal_border; % in the original paper the filter is [-1 0 1], maybe this filter is used twice in the image somehow. Any idea how to pass twice the filter?
%B = vertical_border;
%C = diagonal_border;
%Result = A.*B - C.*C - SensitivityFactor*(A + B);
% A.*B - C.*C is equal to 0, then
% - SensitivityFactor*(A + B)  == -0.04