Dear Huda

I'm not sure exactly what you want to do but It sounds like the "Toolbox Graph" library might be useful. See:

https://www.ceremade.dauphine.fr/~peyre/matlab/graph/content.html

and

http://www.mathworks.com/matlabcentral/fileexchange/5355-toolbox-graph

Hope that helps

Tim

There is a script on mathworks wich allows for Canny edge detection on 2D and 3D data. maybe this helps you?

http://www.mathworks.com/matlabcentral/fileexchange/45459-canny-edge-detection-in-2-d-and-3-d

Dear Hayder;

Which code do you mean? that is exactly what I need . I need the code on Mathwork.

The one i mentioned above?

By 3D image, do you mean a scalar 3D grid like a CT-scan medical image?

Another question is how do you isolate the features of interest you want to measure (contrast, edges, 3D texture, etc)?  What characterizes them? This will also orient your choice of "segmentation" method. As I said, if it's like a medical image volume, I would look at medical oriented software/libraries that are specialized in this kind of algorithms. See the following links:

http://www0.cs.ucl.ac.uk/opensource_mia_ws_2012/links.html

http://www.mitk.net/