VoxelMap

Introduction

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 odometry.

The plane map constructed by VoxelMap on KITTI Odometry sequence 00.

Developers:

Chongjian Yuan 袁崇健， Wei Xu 徐威

Related paper

Related paper available on arxiv:
Efficient and Probabilistic Adaptive Voxel Mapping for Accurate Online LiDAR Odometry

Related video

Our accompanying videos are now available on YouTube.

1. Prerequisites

1.1. PCL && Eigen

PCL>= 1.8, Follow PCL Installation.

Eigen>= 3.3.4, Follow Eigen Installation.

1.2. livox_ros_driver

Follow livox_ros_driver Installation.

2. Build

Clone the repository and catkin_make:

    cd ~/$A_ROS_DIR$/src
    git clone https://github.com/hku-mars/VoxelMap.git
    cd ..
    catkin_make
    source devel/setup.bash

• Remember to source the livox_ros_driver before build (follow 1.2 livox_ros_driver)

3. Run on Dataset

Current version of VoxelMap does not support IMU and requires undistorted point cloud.

3.1 Run on KITTI Odometry dataset

Step A: Setup before run Edit config/velodyne.yaml to set the below parameters:

1. LiDAR point cloud topic name: lid_topic
2. If you want to show the voxel map, set pub_voxel_map to true
3. If you want to show the accumulated point cloud map, set pub_point_cloud to true

Step B: Run below

    cd ~/$VOXEL_MAP_ROS_DIR$
    source devel/setup.bash
    roslaunch voxel_map mapping_velodyne.launch

Step C: Play rosbag.

If want to save the trajectory result (camera pose), set the write_kitti_log to true and change the result_path to your own path.

3.2 Run on L515 dataset

Step A: Download our bags here: Voxel Map L515 Datasets

Then the same step as 3.1

4.Acknowledgments

Thanks for Fast-LIO2 (Fast Direct LiDAR-inertial Odometry)

5. License

The source code is released under GPLv2 license.For any technical issues, please contact us via email ycj1@connect.hku.hk. For commercial use, please contact Dr. Fu Zhang fuzhang@hku.hk.