liorf_localization

This repo is a simple location system based on a priori map, which is based on the lio-sam framework. The current version only supports small scene maps, large scene maps may have some problems.

------------------- Update Date: 2022-11-20 -------------------

• First verison

------------------- Update Date: 2022-12-13 -------------------

• Re-derivation the LM optimization, don't need coordinate transformation.

------------------- Update Date: 2023-03-18 -------------------

• Fixed a bug and was more robust for large scene maps.

------------------- Update Date: 2024-04-30 -------------------

• Add liorf-localization ROS2 version (Foxy、Galactic).

Video: 基于LIO-SAM框架SLAM算法开发（三）：定位之小场景定位算法

Prepare priori map

• You can get a priori map by running liorf or lviorf or liorf-ros2 or any open source SLAM algorithm(e.g. A-LOAM or LOAM or LIO-SAM);

- Make sure the priori map name is "GlobalMap.pcd";
- Change the "savePCDDirectory" parameter in "localization.yaml" to your priori map path.

Dependency

• gtsam(Georgia Tech Smoothing and Mapping library)

    sudo add-apt-repository ppa:borglab/gtsam-release-4.0
    sudo apt install libgtsam-dev libgtsam-unstable-dev
  

Install

1. Use the following commands to download and compile the package. (The available ROS2 implementation see branch liorf_localization-ros2)

  mkdir -p ~/liorf_localization/src && cd ~/liorf_localization/src
  git clone https://github.com/YJZLuckyBoy/liorf_localization.git
  cd ../
  catkin_make

How to use

1. Run the launch file

  roslaunch liorf_localization run_localization.launch

2. Use "2D Pose Estimate" on rviz to set the initial pose

3. Play bag files

  rosbag play your_localization.bag

Examples

1. Urban

Scalable Examples

1. Building(Avia)

2. Underground Parking(Avia)

Acknowledgments

Thanks for LIO-SAM, MulRan and UrbanNavDataset.