FAST-LIVO2 (ROS2 Port)

FAST-LIVO2: Fast, Direct LiDAR-Inertial-Visual Odometry

Important

ROS2 & Isaac Sim Proof-of-Concept: This repository is a ROS2 Humble adaptation based on the original FAST-LIVO2. All core algorithms and intellectual properties belong to the original authors from HKU MARS Lab. This port is intended solely for scientific research and education, ensuring full respect for the original work.

📢 News

• 🔓 2025-01-23: Code released!
• 🎉 2024-10-01: Accepted by T-RO '24!
• 🚀 2024-07-02: Conditionally accepted.

📬 Contact

For original project inquiries, contact zhengcr@connect.hku.hk.

1. Introduction

FAST-LIVO2 is an efficient and accurate LiDAR-inertial-visual fusion localization and mapping system.

Developer: Chunran Zheng 郑纯然

2. Prerequisites (ROS2 Humble)

2.1 OS and Middlewares

• Ubuntu 22.04 + ROS2 Humble
• Omniverse Isaac Sim (Optional, for simulation tests)

2.2 Dependencies

• PCL >= 1.12
• Eigen >= 3.4
• OpenCV >= 4.5

2.3 Sophus

Sophus Installation for the non-templated/double-only version.

git clone https://github.com/strasdat/Sophus.git
cd Sophus
git checkout a621ff
mkdir build && cd build && cmake ..
make
sudo make install

3. Build (colcon)

Clone the repository into your ROS2 workspace src folder:

cd ~/ros2_ws/src
# Clone this ROS2-ported version
git clone <this_repo_url>
cd ../
colcon build --packages-select fast_livo2 --symlink-install
source install/setup.bash

4. Run with Isaac Sim

1. Start the Simulator:

   python3 scripts/isaac_sim_drone_ros2_fast_livo.py

2. Launch Mapping:

   ros2 launch fast_livo2 mapping_isaac_sim.launch.py

5. Acknowledgments

This ROS2 version is deeply appreciative of the HKU MARS Lab's contribution to the community. Please cite their original paper if you find this work helpful:

@article{zheng2024fast,
  title={FAST-LIVO2: Fast, Direct LiDAR-Inertial-Visual Odometry},
  author={Zheng, Chunran and Liu, Jiarong and Zhu, Junkai and Zhang, Fu},
  journal={IEEE Transactions on Robotics},
  year={2024}
}

6. License

The original source code of this package is released under the GPLv2 license. For commercial use, please contact the original authors at HKU MARS Lab.