LIMOT

• A tightly-coupled multi-object tracking and LiDAR-inertial odometry system, allowing for joint estimation of the poses of both the ego-vehicle and surrounding objects.

Dependency

We developed LIMOT on Ubuntu 20.04.

• ROS (Noetic)

  sudo apt-get install -y ros-noetic-navigation
  sudo apt-get install -y ros-noetic-robot-localization
  sudo apt-get install -y ros-noetic-robot-state-publisher
  

• gtsam (Georgia Tech Smoothing and Mapping library, recommended 4.0.3)

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

Install

Use the following commands to download and compile the package.

mkdir -p ~/limot_ws/src
cd ~/limot_ws/src
git clone https://github.com/tiev-tongji/LIMOT.git
cd ..
catkin_make -j

Run the package

1. Run the launch file:

source devel/setup.bash
roslaunch limot run_kitti.launch

2. Run your object detector, which subscribes to LiDAR scans and publishes the detection results with the formats:

   [timestamp, [type, x, y, z, l, w, h, yaw, score], ...,[type ,x, ..., score]].

3. Play existing bag files:

rosbag play your-bag.bag

Sample dataset

• Download the KITTI tracking dataset to test the functionality of the LIMOT. The dataset below are configured to run using the params_kitti.yaml:

  • KITTI tracking dataset: [Google Drive]

• Download the self-collected dataset to test the functionality of the LIMOT. The dataset below are configured to run using the params_hdl64.yaml:

  • Self-collected dataset: [Google Drive]

Paper

• LIMOT has been accepted by IEEE Robotics and Automation Letters.

@article{zhu2024limot,
  title={LIMOT: a tightly-coupled system for LiDAR-inertial odometry and multi-object tracking},
  author={Zhu, Zhongyang and Zhao, Junqiao and Huang, Kai and Tian, Xuebo and Lin, Jiaye and Ye, Chen},
  journal={IEEE Robotics and Automation Letters},
  year={2024},
  publisher={IEEE}
}

• Our previous work DL-SLOT has been accepted by the IEEE Transactions on Intelligent Vehicles.

@article{tian2023dl,
  title={DL-SLOT: Tightly-Coupled Dynamic LiDAR SLAM and 3D Object Tracking Based on Collaborative Graph Optimization},
  author={Tian, Xuebo and Zhu, Zhongyang and Zhao, Junqiao and Tian, Gengxuan and Ye, Chen},
  journal={IEEE Transactions on Intelligent Vehicles},
  year={2023},
  publisher={IEEE}
}

Acknowledgments

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