BEV-LSLAM:

A Novel and Compact BEV LiDAR SLAM for Outdoor Environment

Shaocong Wang, Fengkui Cao, Xieyuanli Chen, Ting Wang and Lianqing Liu

IEEE Robotics and Automation Letters

News

• 1 March 2025: Code updata
• 7 January 2025: Accepted by IEEE RAL!

Getting Started

Instructions

BEV-LSLAM requires an input point cloud of type sensor_msgs::PointCloud2

Dependencies

• Ubuntu 18.04 or 20.04
• ROS Melodic or Noetic (roscpp, std_msgs, sensor_msgs, geometry_msgs, pcl_ros)
• cv_bridge
• Opencv
• C++ 14
• OpenMP
• Point Cloud Library
• Eigen >=3.3.4
• Ceres >=1.14

Compiling

Create a catkin workspace, clone the imagecloud_msg and orb_lio repository into the src folder, and compile via the catkin_tools package (or catkin_make if preferred):

mkdir ws && cd ws && mkdir src && catkin init && cd src
git clone https://github.com/ROBOT-WSC/BEV-LSLAM.git
catkin_make

Execution

For your convenience, KITTI, Urbanloco and Groundrobot can be test on BEV-LSLAM.(Different datasets have different intensity range， please check the 222 line in scantoscan.cpp before your experiment.) For Groundrobot, we provide example test data here. To run, first launch BEV-LSLAM via:

roslaunch orb_lio orb_lo.launch

In a separate terminal session, play back the downloaded bag:

rosbag play bag's name --clock

Citation

If you find BEV-LSLAM is useful in your research or applications, please consider giving us a star 🌟 and citing it by the following BibTeX entry.

@ARTICLE{10845798,
  author={Cao, Fengkui and Wang, Shaocong and Chen, Xieyuanli and Wang, Ting and Liu, Lianqing},
  journal={IEEE Robotics and Automation Letters}, 
  title={BEV-LSLAM: A Novel and Compact BEV LiDAR SLAM for Outdoor Environment}, 
  year={2025},
  volume={10},
  number={3},
  pages={2462-2469},
  keywords={Laser radar;Feature extraction;Simultaneous localization and mapping;Point cloud compression;Visualization;Tracking loops;Robots;Optimization;Pose estimation;Pipelines;SLAM;localization;mapping},
  doi={10.1109/LRA.2025.3531727}}

Acknowledgements

We thank the authors of the FastGICP, orb-slam and A-LOAM open-source packages.