基于智能视觉物联网的油库人员行为识别与监测系统

针对油库区域人员行为监测响应慢、时效低的问题,提出一种基于智能视觉物联网的油库人员行为识别与监测系统。首先建立智能视觉物联网监测系统架构,满足信号采集、传输、处理与反馈需求;然后提出一种视频语义分析模型,将人体行为识别与人脸识别进行协同分析,实现对油库作业人员行为的分析与监测。经实验验证,在自建的数据库中,系统的人脸识别准确率达96.5%,人体行为识别准确率达85%,说明该系统可有效减少因人员的行为失误造成的油库安全事故,在油库的安全管理中具有很高的应用价值。

Behavior recognition and monitoring system for oil depot personnel based on intelligent vision internet of things

The response of the existing personnel behavior monitoring system in oil depot area is slow and inefficient.In order to solve the problem,a behavior monitoring system for oil depot personnel based on intelligent visual internet of things is proposed in this paper.Firstly,the intelligent visual internet of things monitoring system is established to meet the requirements of signal acquisition,transmission,processing and feedback.Then a video semantic analysis model is proposed,which combines human behavior recognition with face recognition to analysis and monitoring of the behavior of oil depot operators.It has been verified by the self-built database.The face recognition accuracy rate of the system is 96.5%,and the human body behavior recognition accuracy rate is 85%.The system can effectively reduce the oil depot safety accidents which caused by behavior mistakes of the oil depot personnel.It has extremely high application value to the safety management of oil depots.