Gradient Image Normalization Process

Hello Team,

I need to follow the below equation (attached as an image) in order to normalize the gradient image after calculating the absolute difference between horizontally adjacent pixels.

This normalization method was proposed in "A no-reference blocking artifact measure for adaptive video processing" for Muijs, R. and Kirenko in 2005.

So I have written this code but I wonder how can I verify if this will work correctly? It did not display any error, but how can I confirm if the results are correct?

if true
  % code
  imageArray = imread('index.jpg');
  Diff_H = abs(diff(imageArray,1,2))
  [rows_H cols_H] = size(Diff_H)
  sum = 0
  Diff_H_norm = zeros(rows_H, cols_H)
  for j=1:1:rows_H
      for i=1:1:cols_H
          for z=1:1:cols_H
              if (z ~= i)
                  sum = sum + power(Diff_H(z,y),2)
              end
          end
       Diff_H_norm(i,j) = Diff_H(i,j)/sqrt((1/(2*cols_H))*sum)
       end
   end 
end