Ship-Relative UAV Pose Estimation with 3D LiDAR
Open Access DepositedAccurate relative pose estimation between unmanned aerial vehicles (UAVs) and marinevesselsiscriticalforautonomouslandingandnavigationinchallengingmaritimeenvironments. Traditional approaches based on GPS or visual sensing are often impacted by satellite availability, security concerns, or harsh environmental conditions such as poor lighting and visibility. To overcome these limitations, this thesis presents a learning-based pipeline that leverages 3D LiDAR sensing to estimate the six-degree-of- freedom (6DoF) pose of a UAV relative to a known ship from a single point cloud scan. The proposed method adopts a Point Transformer-based architecture that extracts both local and global geometric features from sparse LiDAR data using self-attention and cross-attention mechanisms. It predicts the positions of ship keypoints in the sensor coordinate frame and aligns them with canonical CAD model keypoints using a closed-form algorithm, followed by a lightweight refinement via Generalized ICP. The system is trained entirely on simulated LiDAR scans and evaluated on real-world LiDAR data collected during UAV flights aboard a US Naval Academy research vessel on the Chesapeake Bay, MD.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.