3d skeleton branchpoints

Hi all,

I have a 3d skeleton of which I want to calculate the number of branchpoints. Is it ok to use this code?

bp = bwmorph3(skel, 'branchpoints');

cc = bwconncomp(skel,18);

numberOfBranches = cc.NumObjects; % 1object = 1 branchpoint??

I attach an image of my 3d skeleton and of the branchpoints

Thanks in advance for any help!

Answer 1

[Constantino Carlos Reyes-Aldasoro]

Something does not look right. Specifically, the skeleton should be a line, and you do not seem to have any lines, just disconnected points.

Why don't you first detect the skeleton and display it, then manually count your branching points and then you can validate that the result of the branching points obtained with matlab is correct.

[Marco Festugato]

Hi Constantino and thank u for your answer. I don't know why the skeleton looks like that (maybe it's because of the bad quality of the original images?), I just used the MATLAB function (for what I know, it's the only 3D skeletonization function in MATLAB, is that correct?) and I dont know how to improve the skeletonization!

Any suggestions?

Best Regards

Answer 2

[Constantino Carlos Reyes-Aldasoro]

I would suggest that you do a maximum intensity projection into 2D, i.e.

data_2D = max(data_3D, [ ] ,3);

and then check the skeleton. In 2D you should have something standard like the ones shown here

http://homepages.inf.ed.ac.uk/rbf/HIPR2/skeleton.htm

Then you know if it is the data or the code that you are using