GIRA3D Benchmarking

Tools to compare with existing mapping frameworks via python bindings.

Currently supports:

1. Octomap [1]: Wrapper over ColorOcTree class.
2. NDTmap [2]: Wrapper over a custom NDT Cell that includes color in each cell.

Get this repository by:

git clone git@github.com:gira3d/gira3d-benchmarking.git --recursive

Dependencies

This meta-package assumes you are in a virtual environment set up via the meta-package gira3d-reconstruction. If you have not set this up yet, please head to https://github.com/gira3d/gira3d-reconstruction and set up the python virtual environment along with the Open3D, Pybind11, and Eigen3 dependencies.

libpcl-dev = 1.11 is required for NDTMap. On Ubuntu, install via sudo apt install libpcl-dev.

Build

Just run

./build

MacOS and Windows are not supported.

Usage

To update the PYTHONPATH, use

source workon

NDTMap

There is a helper script provided in wet/src/ndt_map_py/test/test_ndt.py. Reads the point cloud in copier.npz (make sure you use Git LFS to pull this point cloud). Visualizes the reconstructed intensity map from NDTMap using Open3D.

Octomap

There is a helper script provided in wet/src/octomap_py/test_scripts/test_color_pcld.py. Same functionality as in NDTMap script above.

References

[1] A. Hornung, K. M. Wurm, M. Bennewitz, C. Stachniss, and W. Burgard, “OctoMap: an efficient probabilistic 3D mapping framework based on octrees,” Autonomous Robots, vol. 34, no. 3, pp. 189–206, Apr. 2013, doi: 10.1007/s10514-012-9321-0.

[2] J. Saarinen, H. Andreasson, T. Stoyanov, J. Ala-Luhtala, and A. J. Lilienthal, “Normal Distributions Transform Occupancy Maps: Application to large-scale online 3D mapping,” in 2013 IEEE International Conference on Robotics and Automation, May 2013, pp. 2233–2238. doi: 10.1109/ICRA.2013.6630878.