Research On The Application Of Computer Vision-based Personnel Detection And Positioning Technology In Dangerous Workshops

The management of workshop staff is the main difficulty and key link in the workshop production activities.Because the activities of the workshop staff are subjective and uncertain.Standing at staff safety point of view,the manufacture of military products(eg,pyrotechnics)has strict requirements on the number of staff and the location of staff in different areas of the workshop.Excessive personnel density in local areas may cause catastrophic safety incidents.If we can effectively manage dangerous workshop staff,it can not only ensure safe production,but also improve production efficiency and improve product quality.We analyzes the needs of the staff intelligent control system,and builds a system framework structure and functional structure.The staff management method is divided into three stages: Real-Time Staff Detection,Multi-Camera Data Fusion,Multi-Regional Staff Positioning.Secondly,we expounds the scheme of arranging the omnidirectional surveillance visual sensor network in the workshop.The visual sensor is used to capture the scene of the workshop,the video data is obtained from the surveillance video,and then we analyze the video data:1)Real-Time Staff Detection: Using a 3-Stage CCNN model to detect the staff in the image and to obtain the location information of the staff in the image;2)Multi-Camera Data Fusion: Combining the detection results of multiple images to eliminate the problem of repeated recognition by people in the overlapping area of the video;3)Multi-Regional Staff Positioning: The detection information of the staff position in the image scene is mapped to the actual workshop scene,thereby obtaining real-time staff location information.Through the above research work,we can obtain the position information of the staff in the workshop in an all-round and full-time,and synchronously present it on Web visual windows.This work will greatly reduce human dependence in Safety supervision task.Finally,we designed and developed a monitoring and positioning system for workshop staff.At the same time,the successful implementation of the system explored a new path for building the digital twins model of the workshop staff.