LIO-SAM-GPU-ScanToMapOpt
LIO-SAM-GPU-ScanToMapOpt copied to clipboard
A CUDA reimplementation of the line/plane odometry of LIO-SAM. A point cloud hash map (inspired by iVox of Faster-LIO) on GPU is used to accelerate 5-neighbour KNN search.
In any case, nice work. I personally optimized LIO-SAM with nanoflann_pcl and replaces all instances of radius search with knn search to reach a 2x speedup. GPU based knn search...
Thans for your great work. While the robot will drift immediately when using the dataset "Campus dataset (small)" like this. What should I modify the project. 
I encountered an error while running the GPU version of LIO-SAM, but the original version works fine. Does anyone know how to resolve this issue? [pcl::VoxelGrid::applyFilter] Invalid filter field name....
Is it possible to run this repository with CUDA 10.2? I have tried it but I only get an error message: `CMake Error at /usr/share/cmake-3.25/Modules/FindPackageHandleStandardArgs.cmake:230 (message): Could NOT find CUDA...