search

siavashk / pycpd

Pure Numpy Implementation of the Coherent Point Drift Algorithm
MIT License
510 stars 115 forks source link

Does it work for higher dimensional point? Like 4D or more? #50

Closed wbensvage closed 3 years ago

wbensvage commented 3 years ago

Thank you for sharing the code. Appreciate your work. Just a quick question: does the code work for higher dimensional point sets? Currently I have 2 datasets containing 4d data (may be the dimension would be higher in the future). Can I use the code? Thank you so much!

gattia commented 3 years ago

Definitely. I think the only exception is that the current implementation won't do >3d for rigid transforms - though I might be miss-remembering. I've used it for more than 3 dimensions using affine, and non-rigid methods. E.g., in my example, I added additional features to the XYZ points like the curvature of a surface mesh associated with the points.

siavashk commented 3 years ago

It seems that @gattia answered this. Closing due to inactivity.

QiqiXian commented 10 months ago

@gattia Hi, I am wondering why rigid registration doesn't work for >3d. It seems that the algorithm is dimensionally independent according to the paper.