Rip Current Detection in Nearshore Areas through UAV Video Analysis with Almost Local-Isometric Embedding Techniques on Sphere
Anchen Sun, Kaiqi Yang
TL;DR
This work targets low-cost, real-time rip current detection using UAV video analysis. It fuses optical-flow-derived fluid velocities with offshore-direction estimates obtained via an almost local-isometric embedding on the sphere, and adds temporal data fusion to produce pixelwise rip-current likelihoods. The framework integrates sea-water and wave segmentation and demonstrates superior performance when using high-order regularization with Horn-Schunck flow, achieving alignment with expert annotations across multiple beaches and yielding robust segmentation and detection metrics. The approach promises scalable coastal monitoring by leveraging accessible UAV data and efficient processing for real-time rip-current alerts.
Abstract
Rip currents pose a significant danger to those who visit beaches, as they can swiftly pull swimmers away from shore. Detecting these currents currently relies on costly equipment and is challenging to implement on a larger scale. The advent of unmanned aerial vehicles (UAVs) and camera technology, however, has made monitoring near-shore regions more accessible and scalable. This paper proposes a new framework for detecting rip currents using video-based methods that leverage optical flow estimation, offshore direction calculation, earth camera projection with almost local-isometric embedding on the sphere, and temporal data fusion techniques. Through the analysis of videos from multiple beaches, including Palm Beach, Haulover, Ocean Reef Park, and South Beach, as well as YouTube footage, we demonstrate the efficacy of our approach, which aligns with human experts' annotations.