from Hacker News

This is a utility library I slowly built up during my PhD and has become my swiss army knife for processing 3D data. It's super easy to install (only depends on NumPy and SciPy).

The goal of the library to have an extremely simple API for geometry processing which uses NumPy arrays as a core data structure (so it can be dropped into whatever numerical codebase you're working with).

Most of the library is written in C++ using a custom binding framework (https://github.com/fwilliams/numpyeigen) that I wrote which avoids copies when converting NumPy arrays to Eigen Matrix types.

Happy to answer any questions you might have about it and I hope Point Cloud Utils is useful to you!