Design And Implementation Of Pedestrian Along The Railway Abnormal Behavior Recognition System

As an indispensable part of the modernization of railway construction,video monitoring system in the protection of the security and stability of the railway running plays a very important role.The railway departments increasingly attach importance to the management of the intelligent way,one of intelligent video surveillance is an important item.In recent years,the railway station of violence to get people to the increased attention given to its own security,at the same time these terrorist violence also poses challenges to the traditional monitoring system,if the system can early warning the abnormal events and provides railway security alarm to the Railway Police,it can reduce the harm of terrorist violence.Based on the requirement,we have carried out relevant research about intelligent monitoring for real-time monitoring and analysis on the video image,to detect a person's abnormal operation.After investigation and analysis about the video monitoring technology at home and abroad,put forward the function of this system demand analysis,and then a detailed requirements analysis,mainly divided into video acquisition,target tracking,target motion analysis,movement analysis and behavior analysis of five parts,on this basis,using MFC based on OpenCV structures,I have built the system framework,the system design stage clear system architecture,and design each function module in detail.There are four main modules,namely,the Video Management Module:The video image preprocessing,provide video images with high quality,reliable video image,and can achieve read and store video;User Management Module:to implement user login and according to different user permissions to management;Target Behavior Analysis Module:to acquire pretreatment of image,the image of target detection,on the basis of the target tracking and the target classification,then is to analyze the operation target,thus trajectory analysis;Abnormal Behavior Analysis Module:In the human abnormal behavior analysis,this thesis is divided into five parts,the main boundary cross the line detection method,the stranded hovered anomaly detection method,the remnant anomaly detection method,abnormal image storage and exception alerts.In the concrete implementation method,I have given the algorithm of the abnormal analysis method,and the correctness of the algorithm in system test period and the efficiency were evaluated,the practice proved that the algorithm used for this system satisfied our requirements.Currently in the strong support and active promotion of the system under the railway sector,the system has been gradually moving towards the use of the various departments.After a period of time using to draw the system operating normally,reducing the burden of safety supervision department staff,and greatly improve the safety and reliability guarantee.