Notes

1. It's a lightweight head-only 3d kdtree library based on nanoflann: https://github.com/jlblancoc/nanoflann.
2. Both StaticKdTree3d and DynamicKdTree3d supports for nearestKSearch and radiusSearch.
3. StaticKdTree3d is similar to pcl::KdTreeFLANN<pcl::PointXYZ> but faster.
4. DynamicKdTree3d supports for dynamic updating(i.e. 3d pointcloud map building) without rebuilding the whole tree.
5. Both StaticKdTree3d and DynamicKdTree3d supports for float/double/etc point type.

Building

1. install the Eigen3
2. install pcl if you want to build and run time_compare.cpp
3. cd ~/catkin_ws/src
   git clone https://github.com/G3tupup/KdTree.git
   cd ..
   catkin_make -DCMAKE_BUILD_TYPE=release or
   you can compile dynamic_3d_tree_example.cpp/static_3d_tree_example.cpp/time_compare in anyway you like.

Run

1. cd ~/catkin_ws/devel/lib/kdtree
2. ./dynamic_3d_tree_example and ./dynamic_3d_tree_example
3. ./time_compare to compare StaticKdTree3d with pcl::KdTreeFLANN