LIGO: Tightly Coupled LiDAR-Inertial-GNSS Odometry based on a Hierarchy Fusion Framework for Global Localization with Real-time Mapping
Code, paper, video are coming soon......
The codes of this repo are contributed by: Dongjiao He (贺东娇)
LIGO is a multi-sensor fusion framework that maximizes the complementary properties of both LiDAR and GNSS systems. This package achieves the following properties:
- Competitive accuracy in trajectory estimation across large-scale scenarios.
- Robustness to malfunctions of either GNSS or LiDAR sensors, enabling seamless handling of added or lost sensor signals during operation.
- High-output-frequency odometry.
- Capability of providing globally referenced pose estimations in both indoor and outdoor environments, suitable for ground vehicles and uncrewed aerial vehicles (UAVs).
- No requirement for GNSS observations to be obtained exactly at the beginning or end time of LiDAR scans.
- Robustness to large outliers and high noise levels in GNSS observations.
Performance on a sequence with severe LiDAR degeneracy
