Academic Journal of Environment & Earth Science, 2022, 4(2); doi: 10.25236/AJEE.2022.040204.
Guochen Zhang1, Pei Yu1, Changle Wang1
1China Fire and Rescue Institute, Beijing, 102202, China
In order to realize the goal of "early, small and small" of forest fire and eliminate fire in the bud, a forest fire smoke identification and monitoring system based on deep learning is designed. Through real-time image processing and deep learning target recognition, the fast forest fire and smoke identification within the monitoring range is realized, and the information can be conveyed to the fire prevention headquarters in the first time after the fire occurs, so as to conduct rapid emergency response.
Deep learning; Forest fire detection; Smoke detection; Monitoring system
Guochen Zhang, Pei Yu, Changle Wang. Application of Deep Learning in Forest Fire Smoke Identification Monitoring System. Academic Journal of Environment & Earth Science (2022) Vol. 4 Issue 2: 14-17. https://doi.org/10.25236/AJEE.2022.040204.
[1] Yang Mingxiao.Linfire Smoke Recognition Based on Deep Learning [D].Beijing Forestry University, 2019.
[2] Zheng Huaibing, Zhai Jiyun.Forest fire smoke detection method based on video analysis [J].Journal of Nanjing University of Science and Technology, 2015,39 (06): 686-691.
[3] Shimmanshu Singh Himanshu Singh.Python Machine learning and image processing in actual combat [M].Tsinghua University Press, 2019.
[4] Chen Zhe et al.Image object detection technology and Application [M].People's Post and Telecommunications Publishing House, 2016.
[5] Redmon J, Divvala S, Girshick R, et al.You Only Look Once: Unified, Real-Time Object Detection [J].2015.
[6] Redmon J, Farhadi A.YOLO9000: Better, Faster, Stronger[J].2017:6517-6525.
[7] https://github.com/ttttt/yolov5.