VoxelMap is an efficient and probabilistic adaptive(coarse-to-fine) voxel mapping method for 3D LiDAR. Unlike the point cloud map, VoxelMap uses planes as representation units. A scan of LiDAR data will generate or update the plane. Each plane contains its own plane parameters and uncertainties that need to be estimated. This repo shows how to integrate VoxelMap into a LiDAR(-Inertial) odometry.
The plane map constructed by VoxelMap on KITTI Odometry sequence 00.
The plane map constructed by VoxelMap in the park environment with Livox Avia LiDAR.
The plane map constructed by VoxelMap in the mountain environment with Livox Avia LiDAR.
Related paper available on arxiv:
Efficient and Probabilistic Adaptive Voxel Mapping for Accurate Online LiDAR Odometry
Our accompanying videos are now available on YouTube.
Our paper is currently under review, and our code will be released after the first round of review if the reviewer's comments are positive.