Efficient On-Board Processing of Oblique UAV Video for Rapid Flood Extent Mapping

TL;DR

This paper introduces Temporal Token Reuse (TTR), a novel adaptive inference framework that accelerates on-board processing of oblique UAV videos for rapid flood mapping by exploiting spatiotemporal redundancy, reducing latency without significant accuracy loss.

Contribution

The paper presents TTR, a new method for real-time oblique UAV video segmentation that leverages static region detection to reuse features and improve processing speed under SWaP constraints.

Findings

TTR reduces inference latency by 30% on embedded hardware.

TTR maintains segmentation accuracy with less than 0.5% mIoU degradation.

Validated on standard benchmarks and a new floodwater dataset.

Abstract

Effective disaster response relies on rapid disaster response, where oblique aerial video is the primary modality for initial scouting due to its ability to maximize spatial coverage and situational awareness in limited flight time. However, the on-board processing of high-resolution oblique streams is severely bottlenecked by the strict Size, Weight, and Power (SWaP) constraints of Unmanned Aerial Vehicles (UAVs). The computational density required to process these wide-field-of-view streams precludes low-latency inference on standard edge hardware. To address this, we propose Temporal Token Reuse (TTR), an adaptive inference framework capable of accelerating video segmentation on embedded devices. TTR exploits the intrinsic spatiotemporal redundancy of aerial video by formulating image patches as tokens; it utilizes a lightweight similarity metric to dynamically identify static regions and propagate their precomputed deep features, thereby bypassing redundant backbone computations. We validate the framework on standard benchmarks and a newly curated Oblique Floodwater Dataset designed for hydrological monitoring. Experimental results on edge-grade hardware demonstrate that TTR achieves a 30% reduction in inference latency with negligible degradation in segmentation accuracy (< 0.5% mIoU). These findings confirm that TTR effectively shifts the operational Pareto frontier, enabling high-fidelity, real-time oblique video understanding for time-critical remote sensing missions