基于视频识别的直接作业环节安全管控

Safety Control of Operation based on Video Recognition

概述了石化行业现场直接作业环节的安全管控现状,分析了现场管控过程中存在的问题。考虑现场安全管理和监督监控的实际需要,设计了石化行业直接作业现场智能化管控系统,并深入研究作业现场违章行为图像识别的关键技术。现场应用表明,系统充分体现了日常管理、实时监控、智能报警、统计分析等方面的便利性和有效性,各类违章行为图像识别准确率达到80%以上,具有一定的推广价值。

This paper summarized the present situation of safety control of operation in petrochemical industry,and analyzed the problems existing in the process of on-site control.Considering the actual needs of on-site safety management and supervision and monitoring,the intelligent management and control system of direct operation site in petrochemical industry was designed,and the key technologies of image recognition of illegal behaviors in operation site were deeply studied.The field application showed that the system fully embodied the convenience and effectiveness of daily management,real-time monitoring,intelligent alarm,statistical analysis and other aspects,and the image recognition accuracy rate of all kinds of violations reached over 80%,which had certain promotion value.