摘要
针对人工监督施工人员是否符合安全规范存在漏检的问题,本文基于YOLOv5目标检测算法来自动检测安全帽和反光衣的穿戴,通过深度学习,模型精度达到75.8%,实现了在复杂的施工场景下对安全帽和反光衣等小目标的检测,较好解决了人工检测漏检的问题。
In response to the issue of manual supervision of construction workers'compliance with safety regulations and missed inspections,this paper uses the YOLOv5 object detection algorithm to automatically detect the wearing of safety helmets and reflective clothing.Through deep learning,the model accuracy reaches 75.8%,achieving the detection of small targets such as safety helmets and reflective clothing in complex construction scenarios,effectively solving the problem of manual detection missed inspections.
作者
陈冲
崔平
鲁璐
张新宇
周南
CHEN Chong;CUI Ping;LU Lu;ZHANG Xinyu;ZHOU Nan(School of Information Engineering,Xuzhou Institute of Technology,Xuzhou,Jiangsu 221018,China;Guangdong Wave Intelligent Computing Technology Co.,Ltd.,Guangzhou,Guangdong 510630,China;Traffic Police Detachment Science and Technology,Xuzhou Public Security Bureau,Xuzhou,Jiangsu 221000,China)
出处
《自动化应用》
2023年第13期21-23,27,共4页
Automation Application
基金
江苏省高等学校大学生创新创业训练计划重点项目(xcx2022183)
广东省重点研发计划(2021B0101400001)。