LL-Localizer

A Lifelong Localization System based on Dynamic i-Octree

Xinyi Li, Shenghai Yuan, Haoxin Cai, Shunan Lu, Wenhua Wang, and Jianqi Liu

This work has been accpeted at IEEE Transactions on Instrumentation and Measurement (TIM).

Relative experiments videos on real AGV: indoor and outdoor

Statement

All experiments, including both simulation and real-world experiments in paper and on real AGV, were conducted on Ubuntu 22.04 with ROS2 Humble. Due to commercial applications and ongoing academic research, the complete code deployed on the real vehicle is not fully open-sourced. Instead, we provide a simplified demo implementation based on Faster-LIO.

Requirements

• Ubuntu 20.04
• ROS Noetic
• Eigen
• PCL

Build

mkdir -p /workspace/src
cd /workspace/src
git clone https://github.com/M-Evanovic/LL-Localizer.git
cd /workspace
catkin build
source devel/setup.bash

Run SLAM Mode

• Set the path /localization_param/map_file_name to " " in /config/config.yaml. When no prior map is available, the system automatically switches to SLAM mode.
• Derectly launch ll_localizer:

roslaunch ll_localizer localizer_velodyne.launch

• And play your rosbags:

rosbag play yourbag.bag

Run Localization Mode

• If you want to use a prior map to run the localization mode, please put the map (.pcd file) in /map, and modify the path /localization_param/map_file_name in /config/config.yaml.
• Then launch ll_localizer:

roslaunch ll_localizer localizer_velodyne.launch

• After loading the prior map, provide a rough 2D Pose Estimate in rviz. (Notice: Since 2D Pose Estimation can only provide the initial x and y coordinates, users can specify the initial z coordinate by adjusting the /localization_param/init_height in /config/config.yaml.)
• Finally play your rosbags:

rosbag play yourbag.bag

Datasets

• NCLT: The University of Michigan North Campus Long-Term Vision and LIDAR Dataset
• M2DGR: A Multi-modal and Multi-scenario SLAM Dataset for Ground Robots
• BotanicGarden: A high-quality dataset for robot navigation in unstructured natural environments

Acknowledgements

• We thank the authors of Faster-LIO, fast-gicp and robin-map for their great contributions.
• And please cite us and support us with a star if you find this work useful:

@ARTICLE{11123901,
  author={Li, Xinyi and Yuan, Shenghai and Cai, Haoxin and Lu, Shunan and Wang, Wenhua and Liu, Jianqi},
  journal={IEEE Transactions on Instrumentation and Measurement}, 
  title={LL-Localizer: A Lifelong Localization System Based on Dynamic i-Octree}, 
  year={2025},
  volume={74},
  number={},
  pages={1-11},
  doi={10.1109/TIM.2025.3598401}}