Fire detection model of wildland-urban interface based on YOLOv5s

doi:10.16265/j.cnki.issn1003-3033.2023.06.1532

Abstract

Abstract:

In order to accurately monitor the fires at wildland-urban interfaces and locate their spatial distribution, a target detection model of wildland-urban interface fires based on the improved YOLOv5s network was proposed. Images of fires at the wildland-urban interfaces were collected, and object detection datasets were annotated with the image annotation tool. CA mechanism was introduced into the backbone network of YOLOv5s to enhance the orientation and location information perception of the model to accurately locate the fire point at the wildland-urban interface. Based on the evaluation indicators of accuracy, recall rate and average accuracy, training and testing were carried out on the self-built data set. The experimental results show that the overall performance of the improved YOLOv5s model is improved, and the average accuracy of building fires increases by 0.8% and forest fires by 1.3% in the detection of fire targets in the wildland-urban interface.

Key words: wildland-urban interface, YOLOv5s, fire detection, target detection

WANG Zhe, LI Xiang, YANG Dongliang, LIU Dan. Fire detection model of wildland-urban interface based on YOLOv5s[J]. China Safety Science Journal, 2023, 33(6): 152-158.

Figures/Tables 10

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Fig.6

Tab.1

Fig.7

Fig.8

Fig.9

References 11

[1]	宋英华, 张小莹, 吕伟. 城镇-森林交界域火灾灾害链网络模型构建及风险分析[J]. 中国安全生产科学技术, 2020, 16(5):122-128.
	SONG Yinghua, ZHANG Xiaoying, LYU Wei. Construction and risk analysis of fire disaster chain network model in wildland-urban interface[J]. China Safety Science and Technology, 2020, 16(5):122-128.
[2]	国务院河南平顶山“5.25”特别重大火灾事故调查组. 河南平顶山“5·25”特别重大火灾事故调查报告[R/OL]. (2015-10-14).https://www.mem.gov.cn/gk/sgcc/tbzdsgdcbg/2015/201510/t20151014_245219.shtml
[3]	张庆杰, 郑二功, 徐亮, 等. 森林防火无人机系统设计与林火识别算法研究[J]. 电子测量技术, 2017, 40(1): 145-150.
	ZHANG Qingjie, ZHENG Ergong, XU Liang, et al. Design of forest fire prevention UAV system and research on forest fire identification algorithm[J]. Electronic Measurement Technology, 2017, 40(1): 145-150.
[4]	牛磊. 激光技术在建筑物火灾检测的应用研究[J]. 激光杂志, 2016, 37(4): 29-32.
	NIU Lei. Research on the application of laser technology in building fire detection[J]. Laser Magazine, 2016, 37(4): 29-32.
[5]	李柯泉, 陈燕, 刘佳晨, 等. 基于深度学习的目标检测算法综述[J]. 计算机工程, 2022, 48(7):1-12. doi: 10.19678/j.issn.1000-3428.0062725
	LI Kequan, CHEN Yan, LIU Jiachen, et al. Overview of target detection algorithms based on deep learning[J]. Computer Engineering, 2022, 48(7):1-12. doi: 10.19678/j.issn.1000-3428.0062725
[6]	张全, 张伟, 杨先凤, 等. 融合YOLOv5-ResNet级联网络的烟火检测方法[J]. 安全与环境学报, 2023, 23(2):397-405.
	ZHANG Quan, ZHANG Wei, YANG Xianfeng, et al. Smoke and fire detection method integrating YOLOv5-Resn cascaded network[J]. Journal of Safety and Environment, 2023, 23(2):397-405.
[7]	栗俊杰, 毛鹏军, 淡文慧, 等. 基于YOLOv2-Tiny的无人机火灾检测与云台跟踪研究[J]. 消防科学与技术, 2022, 41(1): 108-112.
	LI Junjie, MAO Pengjun, DAN Wenhui, et al. Research on UAV fire detection and PTZ tracking based on YOLOv2-Tiny[J]. Fire Science and Technology, 2022, 41(1): 108-112.
[8]	拉毛措, 安见才让, 拉毛杰. 基于嵌入并行注意力机制的YOLOv3模型高原动物种类检测算法[J]. 微处理机, 2021, 42(2): 37-40.
	Lamaocuo, Anjiancairang, Lamaojie. Algorithm of YOLOv3 model for plateau animals category detection based on embedded parallel attention mechanism[J]. Microprocessor, 2021, 42(2): 37-40.
[9]	赵江平, 徐恒, 党悦悦. 基于改进Faster R-CNN的铁路客车螺栓检测研究[J]. 中国安全科学学报, 2021, 31(7):82-89. doi: 10.16265/j.cnki.issn 1003-3033.2021.07.012
	ZHAO Jiangping, XU Heng, DANG Yueyue. Research on railway passenger car bolt detection based on improved faster R-CNN[J]. China Safety Science Journal, 2021, 31 (7):82-89. doi: 10.16265/j.cnki.issn 1003-3033.2021.07.012
[10]	包晗, 康泉胜, 周明. 一种基于LVQ神经网络与图像处理的火焰识别算法[J]. 中国安全科学学报, 2011, 21(6): 60-64.
	BAO Han, KANG Quansheng, ZHOU Ming. A flame recognition algorithm based on LVQ neural network and image processing[J]. China Safety Science Journal, 2011, 21(6):60-64.
[11]	HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[C]. Computer Vision-ECCV 2014:13^th European Conference, 2014:346-361.

模型	类别	P	R	m
YOLOv5s- CA	全体	0.958	0.960	0.957
	建筑火灾	0.927	0.920	0.922
	森林火灾	0.989	0.990	0.993
YOLOv5s	全体	0.928	0.928	0.947
	建筑火灾	0.898	0.872	0.914
	森林火灾	0.957	0.985	0.980