Reproduce “HARBOR: Heading Analysis and Reconstruction from Behavioral Observation and Radar” 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
Joao P. A. Dantas, Paulo F. Silva Filho, Jelton A. Cunha, Gabriel Dietzsch
The HARBOR pipeline transforms a single Synthetic Aperture Radar (SAR) image into predictive vessel motion information by detecting vessels, estimating their heading based on local radar returns, and projecting future positions using statistically calibrated motion parameters derived offline from AIS data. This approach generates direction-aware probabilistic heatmaps, enabling short-term trajectory forecasting in data-denied environments without requiring real-time auxiliary data.
Key insights
- Separating motion parameter calibration (offline using AIS) from online SAR image analysis enables predictive capabilities without real-time reliance on cooperative signals or auxiliary data.
- Local radar return intensity patterns around vessel endpoints can be leveraged to infer bow/stern direction, providing a mechanism for heading estimation from static SAR images.
- Probabilistic heatmaps, generated by combining estimated heading with statistically derived speed and angular spread, effectively transform static SAR detections into dynamic, direction-aware short-term trajectory projections.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.