Research On The Application Of VIBE Model In The Obstacle Detection Of Railway Level Crossing

With the continuous expansion of the railway network,it is inevitable to have more intersection with the existing road network.As most of the current level crossings are equipped with cameras,It is of great significance to ensure the safety of level crossing by using surveillance video data and realize real-time intelligent monitoring.It can not only reduce the cost brought by completely manual monitoring,but also can greatly avoid unforeseen consequences caused by manual negligence,which is beneficial to the safety and smooth crossing of railway crossing.Therefore,this paper implements an obstacle detection system for railway level crossing based on video surveillance.This paper first introduces the research background and significance of the subject,and analyzes the research status of the subject at home and abroad.Based on the situation that the railway level crossing has been installed with the monitoring camera,the video image monitoring system is used to design the railway level crossing obstacle detection system.The advantages and disadvantages of several commonly used algorithms for moving object detection are compared,and the principle of the algorithm and the characteristics of the algorithm are analyzed in depth.At the same time,by comparing two frames difference method,three frames difference method,mixed Gauss background model,coded model and VIBE background model,the moving target detection effect in real railway level crossing monitoring are compared.The running time,running memory and detection effect of these algorithms are compared and analyzed.This paper analysis of the characteristics of the actual scene of the railway level crossing,combined with the VIBE background model and optical flow based on the characteristics of the actual scene of the railway level crossing.The ONVIF protocol is used for the information acquisition,video capture and time correction of webcam.Combined with the QT framework and OpenCV,the obstacle detection system for railway level crossing is developed.