Abstract
Accurate environmental perception is a key challenge for uncrewed surface vehicles in dynamic maritime environments, where waves, wakes, and variable visibility often reduce the reliability of conventional sensing systems. This article presents a light detection and ranging (LiDAR)-based perception framework that integrates wake-aware filtering and real-time adaptive clustering for maritime object detection and tracking. The framework consists of a multiframe spatiotemporal fusion module, an adaptive dynamic random sample consensus (AD-RANSAC) filter, and a range-tolerant adaptive spectral density-based spatial clustering of applications with noise (RTA-DBSCAN) clustering algorithm. AD-RANSAC combines a Gaussian mixture model with a physics-informed wake model to suppress structured and unstructured maritime noise, while RTA-DBSCAN adaptively adjusts the clustering radius according to sensor geometry and local point-cloud density. Extensive sea-trial results show that the proposed method achieves a true positive rate of 93.37%, reduces false positives, and maintains real-time performance, outperforming representative DBSCAN variants and several deep learning-based detectors. The results demonstrate the effectiveness of the proposed framework for robust maritime perception under varying sea states.
| Original language | English |
|---|---|
| Pages (from-to) | 15670-15686 |
| Number of pages | 17 |
| Journal | IEEE Transactions on Aerospace and Electronic Systems |
| Volume | 62 |
| Early online date | 19 May 2026 |
| DOIs | |
| Publication status | Published - 2026 |
Bibliographical note
Publisher Copyright:© 1965-2011 IEEE.
Funding
This work is supported by the Joint Training Fund Project of Hanjiang National Laboratory (No.HJLJ20230406), Basic Research Funds for central universities (3072024YY0401), the National Key Laboratory of Underwater Robot Technology Fund (No. JCKYS2022SXJQR-09), and a Special Program to Guide High-Level Scientific Research (No. 3072022QBZ0403).
Keywords
- Bayesian optimization
- LiDAR perception
- density-based spatial clustering of applications with noise (DBSCAN)
- random sample consensus (RANSAC)
- real-time clustering
- sensor fusion
- uncrewed surface vehicle (USV)
- wake modeling
Fingerprint
Dive into the research topics of 'Wake-Aware LiDAR Perception for Uncrewed Surface Vehicles via Physics-Informed Filtering and Adaptive Clustering'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver