High-precision Wildfire Detection Algorithm Using an Enhanced You Only Look Once Model in a Jetson Xavier Environment

ABSTRACT: The You Only Look Once (YOLO) algorithm is employed to develop a system for detecting wildfire smoke and providing early warnings, with a deep neural network (DNN) as its underlying architecture. To achieve accurate real-time smoke detection, the system was optimized through experiments conducted under diverse conditions. It was then implemented in an embedded computing environment to enhance the efficiency of wildfire detection. The findings demonstrate the effectiveness of this DNN-based smoke detection system in real-world environments.