M-LOAM

Robust Odometry and Mapping for Multi-LiDAR Systems with Online Extrinsic Calibration

M-LOAM is a robust system for multi-LiDAR extrinsic calibration, real-time odometry, and mapping. Without manual intervention, our system can start with several extrinsic-uncalibrated LiDARs, automatically calibrate their extrinsics, and provide accurate poses as well as a globally consistent map.

Authors: Jianhao Jiao, Haoyang Ye, Yilong Zhu, Linxin Jiang, Ming Liu from RAM-LAB, HKUST

Project website: https://ram-lab.com/file/site/m-loam

Videos:

(Video link in for mainland China friends: Video)

Related Papers

• Robust Odometry and Mapping for Multi-LiDAR Systems with Online Extrinsic Calibration, Jianhao Jiao, Haoyang Ye, Yilong Zhu, Ming Liu, under review. pdf
• Greedy-Based Feature Selection for Efficient LiDAR SLAM, Jianhao Jiao, Yilong Zhu, Haoyang Ye, Huaiyang Huang, Peng Yun, Linxin Jiang, Lujia Wang, Ming Liu, under review.
• MLOD: Awareness of Extrinsic Perturbation in Multi-LiDAR 3D Object Detection for Autonomous Driving, Jianhao Jiao*, Peng Yun*, Lei Tai, Ming Liu, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS, 2020). pdf

If you use M-LOAM for your academic research, please cite one of our paper. bib

1. Prerequisites

1.1 Ubuntu and ROS

Ubuntu 64-bit 16.04 or 18.04. ROS Kinetic or Melodic. ROS Installation

1.2. Ceres Solver && Eigen3 && GLOG

 ./setup/install_eigen3_ceres.sh

1.3. OpenMP

  sudo apt install libomp-dev

2. Build M-LOAM on ROS

  mkdir -p ~/catkin_ws/src
  cd ~/catkin_ws/src
  git clone https://github.com/gogojjh/M-LOAM.git
  catkin build mloam
  source ~/catkin_ws/devel/setup.bash

3. Example

• Datasets collected with different platforms:

  1. Simulation Robot (SR)
  2. Real Handheld Device (RHD)
  3. Real Vechile (RV)
  4. Oxford RoboCar (OR)

• Run M-LOAM and baseline methods

  1. We provide a script to perform batch testing of M-LOAM with baseline methods
  2. Enter the script folder: roscd mloam/script/
  3. Modify the python script: run_mloam.py for specific platforms with correct path
  4. Modify the shell files for methods in xx_main.sh
  5. Run the python script: python2 run_mloam.py -program=single_test -sequence=xx -start_idx=0 -end_idx=0

4. System pipeline

This could help you to understand the pipeline of M-LOAM. Note that mloam_loop is in development.

5. Issues

I have modified the code with several times and tried different new features during the journal review process. The code style is not very perfect. Also in some sequeneces, M-LOAM may not achieve the best performence. Hope you can understand and I will try to fix them.

6. TODO

1. Parameter tunning, and a more detailed tutorial .
2. loop closure.
3. Docker support. The initial Docker file is in the folder: docker/Dockerfile
4. etc.

6. Acknowledgements

Thanks for these great works from which we learned to write M-LOAM

• LOAM (J. Zhang and S. Singh. LOAM: Lidar Odometry and Mapping in Real-time) and its advanced version: A-LOAM;
• LEGO-LOAM
• LIO-MAPPING
• VINS-MONO
• Lidar Perception Library

7. Licence

The source code is released under GPLv3 license.

We are still working on improving the code reliability. For any technical issues, please contact Jianhao Jiao [email protected].

For commercial inquiries, please contact Prof.Ming Liu [email protected].