Detection And Tracking Of People And Recognition Of Abnormal Behavior In Substation Room

Substation rooms are high-voltage live operation,staff will enter the live working area by mistake to cause electric shock.when the maintenance personnel are still operating,but they send electricity artificially by mistake,resulting in electric shock.In order to ensure the personal safety and protect the equipment,video surveillance has been widely used in this field.But the video is still need to be identified by people,and it is ineffective.With the development of the technology,the informatization,intellectualization and unmanned of the power grid system has become a direction of development.This paper focuses on the detection and tracking of the staff in the substation room and the identification of abnormal behavior.Intelligent monitoring around the status of the staff in the substation room has achieved the following results:1.In order to realize the detection of substation staff,this paper presents a detection method based on background modeling,background subtraction and morphological synthesis algorithm.Firstly,ViBe(Visual background extractor)algorithm is used to model the detection method.Then detect the target.The expansion operator is used for morphological processing,and the rectangular frame is used for marking and moving object extraction.2,In order to achieve the tracking of staff in substation,this paper proposes a tracking algorithm for KCF.By comparing the Mean shift algorithm,Kalman filter tracking algorithm and KCF(Kernel Correlation Filter)tracking algorithm,the real-time performance and accuracy are compared to realize the real-time tracking of substation staff.3.In order to realize the behavior recognition of substation staff,this paper proposes a method based on HOG(Histogram of Orientation Gradient)+ SVM(support vector machine)to judge moving target walking,squatting,falling states.If the staff falls down and the situation is abnormal,the alarm should be given immediately to protect the safety of the staff.4.Based on the above contents,this paper uses GUI graphical interface to design a visual interactive computer-aided substation staff detection and abnormal behavior detection alarm monitoring system.This paper introduces the design idea,module,function and technology of the system,so as to realize better human-computer interaction mode.Based on the above research,the substation monitoring system can detect and track the staff,and conduct behavior judgment,high reliability,can replace the traditional record which also need manual day and night monitoring mode to achieve intelligent identification.