/scds-pb

Spatial cache data structures with python binding implementation.

Primary LanguageC++GNU General Public License v3.0GPL-3.0

SCDS-PB

Spatial Cache Data Structures with Python Bind implementation. The C++ code in this repo is fully templated and I try to incorporate as much modern C++ programming (like perfect-forwarding, rvalue references, conditional compilation and SFINAE), as well as macro programming. This repo is expected to develop high performance CPU/GPU end spatial caching structures for point clouds and graphics primitives, etc.

Currently supported features:

  • CPU end: Static-Multi-Tree (like Quad, Oct-Tree): dynamic insert (deletion not supported yet), fully templated (with visualization). C++/Python API
  • CPU end: k-d Tree: tested, dynamic insert (deletion not supported yet). C++/Python API. Even faster than Quad-tree (5-10x).
quad-tree space subdivision k-d tree space space subdivision Nearest neighbor search