Simple shape recognition in a 3D image

112 views Asked by At

I have a 3D X-ray tomography image of a few thousand particles (each with a length scale of 20-30 voxels) jammed together, all the same shape (spheres, cubes, dolos, etc.), but scattered about randomly (touching each other) and in random orientations. I read about template matching in 2D for various libraries, but haven't been able to find anything good in 3D, and I'm not even sure if template matching is the way to go. In the case of spheres, I've located each particle through continuous erosion of a binarized image, but for more complex shapes this doesn't work too well. Any ideas? I'd prefer to stick to python, but am up to use/learn anything needed. Thanks!

0

There are 0 answers