Forest Fire Detection Method Based on Deep Learning
Wenjie Wang, Qifu Huang, Liu Haiping, Yanxiang Jia, Qing Chen
Abstract
Forest fire causes irreparable damage to human beings and ecological environment with the big concealment and the difficulty to fight. However, conventional fire warning technologies suffer from relatively low sensitivity and accuracy. It's of great importance to detect the forest fire accurately in the budding stage. Herein, we reported a technology to improve the forest fire early warning capability. We analyzed common fire detection methods, studied the forest fire detection in combination with the deep learning technology. The calculation efficiency was improved by introduction of the data enhancement and feature enhancement methods. The lightweight real-time fire detection technology is realized by combination training the deep learning YOLO model and conducting experiments. And the results show that the proposed methods have high accuracy and sensitivity in flame data set.