Calculating Halarick/GLCM features for just an image segment.

Hello, I have an image mask that I would like to apply to an image. I then want to calculate the Halarick texture features (GLCM features) for just this region. Is this possible? If so, what would the code look like?

I currently have a segmented image that contain my area of interest, but then has a large black region on the outside (see below). I'm worried this large black outside region will give me incorrect texture measurements if I input just this image into a graycomatrix.

I tried using code like:

GLCM = graycomatrix(I(I~=0));

but this doesn't keep the right orientation of the image.

Is there an easy solution? Thanks in advance!

Answer 1

[Image Analyst]

Just use the whole image. Then to get rid of the effect that the zero gray level has, just zero out the first row and first column of the GLCM.

[Jake]

The current code I am using is:

        gray = rgb2gray(imageSegment);
        offsets = [0 1; 0 2; 0 3; ...
                  -1 1; -2 2; -3 3; ...
                  -1 0; -2 0; -3 0; ...
                  -1 -1; -2 -2; -3 -3; ...
                  0 -1; 0 -2; 0 -3; ...
                  1 1; 2 2; 3 3; ...
                  1 0; 2 0; 3 0; ...
                  1 -1; 2 -2; 3 -3];
        GLCM2 = graycomatrix(gray,'Offset',offsets);

Would this implementation be correct? :

        GLCM2(1,:) = 0;
        GLCM2(:,1) = 0;

My goal is to compare across multiple segmented images. So I don't want to calculate any features based on the background black of these images.

Thanks again.