Implement of pcl gicp, pcl ndt, fast_gicp single thread, fast_gicp multi threads, fast_vgicp single thread, fast_vgicp multi threads and their CUDA version.
git clone https://github.com/M-Evanovic/fast_gicp_cpp.git
cd path/to/fast_gicp_cpp
Put your target cloud and source cloud to /data and then compile.
mkdir build && cd build
cmake ..
make
./gicp_align target_pcd_path source_pcd_path det_range(optional) voxel_size(optional) max_correspondence_distance(optional)
example
./gicp_align ./data/map.pcd ./data/scan.pcd 30 0.3 3
Thank the authors of fast-gicp for open-sourcing their outstanding works.