Research on Fire Detection Based on the Yolov9 Algorithm
Linhan Song, Yuhang Wu, Wenbo Zhang
Abstract
This paper explores fire detection technology using the YOLOv9 algorithm, highlighting its importance in mitigating the risks to life and property. It reviews traditional methods and discusses the evolution towards deep learning, focusing on YOLOv9 for its speed and accuracy in real-time detection. The algorithm’s principles and optimizations, such as Programmable Gradient Information (PGI) and Generic ELAN (GELAN), are detailed. The study includes dataset selection, preprocessing, and model implementation, showcasing YOLOv9’s effectiveness in detecting fires. The paper emphasizes practical applications like real-time monitoring and future trends in model optimization and application expansion, contributing to fire safety advancement.