This is a python based implementation of the OctoMap datastructure described in [OctoMap: An Efficient Probabilistic 3D Mapping Framework Based on Octrees] (http://www.informatik.uni-freiburg.de/~hornunga/pub/hornung13auro.pdf). If you want to use this library, checkout the OctoMap project first.

1. See whether python based implementation of an octomap is feasible.
2. See whether octomap principles could be of use in our autonomic drone research.

This library only makes sense if you can do some sort of raytracing. Remains future work though.

Ideally, the library offers an API that allows you to do record distance d measured from point (x,y,z) with sensor inaccuracy i in direction (yaw,pitch,roll). This would then update the odds of occupancy around the measured distance.

The current implementation doesn't store any of the geometry data in the OccupancyOctoNodes. It simply passes all required information as arguments down the chain. This should lead to some memory consumption improvements. It would be interesting to see the computational penalty involved though.