kafeiyin00 / MSF_LOAM

Multi-Sensor Fusion SLAM Based on A-LOAM.

MSF_LOAM

[toc]

Multi-Sensor Fusion SLAM

MSF_LOAM is a Multi-Sensor Fusion SLAM implementation based on A-LOAM.

Modifier Keke Liu

1. Prerequisites

1.1 Ubuntu and ROS

Recommend: Ubuntu 18.04 and ROS melodic.

1.2. Ceres Solver

sudo apt install libceres-dev

1.3. PCL

sudo apt install libpcl-dev

2. Build MSF_LOAM

Clone the repository and catkin_make.

3. Run

3.1 Velodyne VLP-16 Example

Download NSH indoor outdoor to YOUR_DATASET_FOLDER.

roslaunch msf_loam_velodyne msf_loam_velodyne_VLP_16.launch
rosbag play ${YOUR_DATASET_FOLDER}/nsh_indoor_outdoor.bag

3.2 Use self-collected data

Sensor	ROS topic	Frequency	Remark
LiDAR (Required)	/velodyne_points	10
GPS	/odometry_gt	1
IMU	/imu	400	higher is better, use xsens_ros_mti_driver to record IMU data with high time precision

4. Acknowledgements

Thanks for LOAM(J. Zhang and S. Singh. LOAM: Lidar Odometry and Mapping in Real-time) and A-LOAM.

5. TODO

For some reasons, this repo will be massively updated, aiming to add features as following:

• IMU intrinsic and extrinsic paramerter estimation
• IMU LiDAR fusion localization (tightly-coupled)
• LiDAR scan undistortion
• Online temporal calibration for system

6. Related paper

• Qin, T., Li, P. and Shen, S., 2018. Vins-mono: A robust and versatile monocular visual-inertial state estimator. IEEE Transactions on Robotics, 34(4), pp.1004-1020.
• Qin, T. and Shen, S., 2018, October. Online temporal calibration for monocular visual-inertial systems. In 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (pp. 3662-3669). IEEE.
• Wu, Y., 2019. Formula Derivation and Analysis of the VINS-Mono. arXiv preprint arXiv:1912.11986.