摘要
现场安全管控是电力安全生产的重要保障环节。随着智慧变电站的发展,现场安全管控由人工分析监控视频方式逐渐向智能管控方式转变。为此提出了一种人体姿态识别技术,通过OpenPose建立多人姿态识别模型,结合传统图像识别技术实现人员状态的全面感知,对作业人员着装、安全帽佩戴、安全带使用、作业中的危险及不规范行为进行实时管控及预警,减轻运检人员压力,提升变电站安全管控智慧化程度。
On-site safety management and control is an important guarantee for power safety production.With the development of smart substations,on-site safety management and control has gradually changed from manual analysis of surveillance video to intelligent management and control.Therefore,a human gesture recognition technology is proposed in this paper.A multi-person gesture recognition model is established through OpenPose.Combined with traditional image recognition technology,the comprehensive perception of personnel state is realized.The real-time control and early warning are carried on operators′ clothing,helmet wearing,safety belt use,hazards and non-standard behaviors in operation,the pressure of operation and inspection personnel are reduced and the intelligence of substation safety control are improved.
作者
葛东阳
GE Dongyang(Maintenance Branchof State Grid Beijing Electric Power Company,Beijing 100069,China)
出处
《电工技术》
2022年第13期132-134,共3页
Electric Engineering
关键词
人体姿态识别
智慧变电站
变电运检
智能管控
human body gesture recognition
smart substation
substation operation inspection
intelligent management and control