Research And Implementation Of Target Tracking Early-Warning System Based On Point Cloud And Image

With the continuous expansion of the scale of the State Grid,the number of construction operations of substations has also continued to increase.Since the equipment in the substation is all high-voltage equipment,it is easy to cause electric shock accidents when the worker or construction vehicle is too close to the high-voltage equipment.At present,in order to reduce the occurrence of construction accidents,it mainly relies on the staff to supervise the whole process of construction on site to ensure that the distance between worker or construction vehicle and high-voltage equipment is within a safe range.However,the efficiency of relying on the supervision of the staff is very low,and construction accidents will also occur due to the inattention of the supervisors.As our country promotes the intelligent transformation of the State Grid and the rapid development of 3D point cloud technology and artificial intelligence technology.It is of great significance to the development of the State Grid to develop an intelligent system to automatically monitor the construction site,and to realize automatic alarm when the operator or the operation vehicle is too close to the high-voltage equipment.The purpose of the target tracking early-warning system is to intelligently monitor the distance between moving objects and highvoltage equipment at the construction site of the substation.If the distance between the two is less than the set safety distance,it will automatically alarm.The system will display the 3D model of the substation,which is restored in proportion to the real world,and the user can mark the dangerous areas in the substation through the interface.First,the system will collect the image data and 3D point cloud data of the construction site in real time;then use the target detection and tracking model to complete the tracking of the moving objects on the construction site,and predict the subsequent trajectory of the moving objects;Finally,calculate the distance between the moving object and the dangerous area and judge whether to alarm or not.The system development adopts the B/S architecture,placing a large amount of computing logic in the back-end service,and the client requests the back-end service through the HTTP protocol to get the data for real-time scene rendering.This paper introduces the desigm and implementation of the system in detail.Firstly,the application scenarios and functional requirements of the system are analyzed,and the architecture design and functional module division of the system are completed;secondly,the relevant technical implementation schemes involved in each module are investigated,and the technical selection of each module of the system is completed;Then it introduces the operation process and detailed implementation scheme of each module,and completes the related development work;finally,through the complete test of the system,the deficiencies of the system are improved and the stable operation of the system is ensured.