how to calculate glcm ?

5 views (last 30 days)
Oman Wisni
Oman Wisni on 15 Oct 2018
Commented: Oman Wisni on 17 Oct 2018
Hi, here I have code for calculate glcm using graycoprops(); and manual calculated. When I run in comman window using graycoprops() I get result. But in manual code the result is 0. Any explain and have will good. thanks !
[fname, pathname]=uigetfile({'*.bmp*','Input image'}); fname = fullfile(pathname, fname); im=imread(fname);
[~, foldername, ext] = fileparts(fname);
gray = rgb2gray(im);
%using toolbox
glcm = gray;
stats = graycoprops(glcm);
%manual calculated
S = size(gray, 1);
if any(gray(:))
prob = gray./sum(gray(:));
end
prob_norm = gray ./ sum(gray(:));
u = mean2(prob_norm);
py = sum(prob_norm,1);
px = sum(prob_norm,2);
contr = 0; energ = 0; homom = 0; %corr = 0;
for i = 1:S for j = 1:S
energ = energ + prob_norm(i,j)^2;
contr = contr + prob_norm(i,j).*(i-j)^2;
homom = homom + prob_norm(i,j)/(1+(i-j)^2);
%co = i*j*prob(i,j);
end
end
% Correlation
%ux = mean(px); sx=std(px);
%uy = mean(py); sy=std(py);
%corr =(sum(co(:))-(ux*uy))/(sx*sy);
Stat.Contrast = contr;
Stat.Homogeneity = homom;
Stat.Energy = energ;
%Stat.Correlation = corr;
disp ('========== Fitur Tekstur GLCM ==========')
disp(strcat(['Contrast = ',num2str(contr)]))
disp(strcat(['Homogenitas = ',num2str(homom)]))
disp(strcat(['Energy = ',num2str(energ)]))
  3 Comments
Oman Wisni
Oman Wisni on 16 Oct 2018
yes sir I will replay to ask and attached my code. I want calculate GLCM using formula which I get from haralick to know did the result same or not.
I already using graycoprops(); But when I using formula to calculate the result 0.
In below I have an image and my code :
S = size(gray, 4);
if any(gray(:))
prob = gray./sum(gray(:));
end
prob_norm = gray ./ sum(gray(:));
u = mean2(prob_norm);
py = sum(prob_norm,1);
px = sum(prob_norm,2);
contr = 0;
energ = 0;
homom = 0;
corr = 0;
for i = 1:S
for j = 1:S
energ = energ + prob_norm(i,j)^2;
contr = contr + prob_norm(i,j).*(i-j)^2;
homom = homom + prob_norm(i,j)/(1+(i-j)^2);
co = i*j*prob(i,j);
end
end
%Correlation
ux = mean(px); sx=std(px);
uy = mean(py); sy=std(py);
corr =(sum(co(:))-(ux*uy))/(sx*sy);
Stat.Contrast = contr;
Stat.Homogeneity = homom;
Stat.Energy = energ;
Stat.Correlation = corr;
Oman Wisni
Oman Wisni on 17 Oct 2018
Which one I wrong declarad that code? Or my code not complete so the result was 0

Sign in to comment.

Answers (0)

Categories

Find more on Sparse Matrices in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!