A forest smoldering fire real-time identification method based on lightweight YOLO

Abstract

The present application relates to the technical field of forest fire early monitoring, and particularly provides an infrared smoldering fire identification method based on YOLO. The method is based on the YOLO algorithm, and a lightweight YOLO model with simple structure and good real-time performance is proposed. The model has small parameter quantity, fast inference speed and strong anti-interference ability, and can meet the requirements of the infrared smoldering fire identification scene of the unmanned aerial vehicle airborne edge device. Deep learning and machine vision algorithms are used to realize the identification and positioning of smoldering fire in the infrared thermal imaging image, improve the accuracy of smoldering fire identification, effectively distinguish the weak thermal characteristics of smoldering fire and the background thermal interference of the forest area, and realize the rapid detection of infrared smoldering fire in the complex environment of the forest area. The method can accurately identify smoldering fire in infrared thermal imaging, and is suitable for continuous monitoring operation of unmanned aerial vehicle forest patrol.

Images