Reproduce “Traj-LO: In Defense of LiDAR-Only Odometry Using an Effective Continuous-Time Trajectory” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Xin Zheng, Jianke Zhu
Researchers at Zhejiang University developed Traj-LO, a LiDAR-only odometry system that models sensor motion as a continuous-time trajectory. This approach achieves performance comparable to state-of-the-art LiDAR-inertial systems and demonstrates superior robustness, exhibiting minimal drift in extreme dynamic scenarios where IMUs typically saturate.
Key insights
- A continuous-time trajectory model, even a simple piecewise linear one, effectively mitigates motion distortion and enables robust LiDAR-only odometry by inherently preserving spatio-temporal consistency.
- Kinematic priors, typically provided by IMUs, can be effectively inferred from the smoothness of the LiDAR-derived continuous trajectory itself, allowing LiDAR-only systems to achieve competitive performance.
- LiDAR-only odometry can outperform tightly-coupled LiDAR-inertial systems in extreme dynamic scenarios where IMUs saturate due to high angular velocities or accelerations, demonstrating superior resilience.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.