Two-Stage Framework for Efficient UAV-Based Wildfire Video Analysis with Adaptive Compression and Fire Source Detection

TL;DR

This paper introduces a two-stage UAV framework for wildfire video analysis that combines clip filtering with adaptive compression and fire source detection, enabling efficient and real-time wildfire monitoring.

Contribution

The authors propose a novel lightweight two-stage framework that reduces computational load and improves fire detection accuracy on UAVs with limited resources.

Findings

Significant reduction in computational costs during video analysis.

High accuracy in fire source localization in real-time.

Effective clip filtering improves overall efficiency without sacrificing accuracy.

Abstract

Unmanned Aerial Vehicles (UAVs) have become increasingly important in disaster emergency response by facilitating aerial video analysis. Due to the limited computational resources available on UAVs, large models cannot be run efficiently for on-board analysis. To overcome this challenge, we propose a lightweight and efficient two-stage framework for wildfire monitoring and fire source detection on UAV platforms. Specifically, in Stage 1, we utilize a policy network to identify and discard redundant video clips, thereby reducing computational costs. We also introduce a station point mechanism that incorporates future frame information within the sequential policy network to improve prediction accuracy. This mechanism allows Stage 1 to operate in a near-real-time manner. In Stage 2, for frames classified as containing fire, we apply an improved YOLOv8 model to accurately localize the fire source in real-time on selected frames. We evaluate Stage 1 using the FLAME and HMDB51 datasets, and Stage 2 using the Fire & Smoke Detection Dataset. Experimental results show that our method significantly reduces computational costs while maintaining classification accuracy in Stage 1, and achieves high detection accuracy with real-time inference in Stage 2.