Design And Realization Of Passenger Motoring System In Subway Based On Video

With the rapid development of the subway transport, it has become the primarymethod to solve traffic jam. Passenger flows security monitoring is an importantcontent of the daily operations of the subway station. However, the CCTVmonitoring system which is widely used in subway station currently can onlymonitor passively and is lack of active response capability which is difficult to meetthe automation control needs of station. To solve these problems, this thesis doessome research.First, this thesis describes the current study of passenger monitoring, andsummarizes the difficulties of this technical. Second, it introduces the technical ofimage processing and the knowledge of subway passengers. What’s more the changeof the traffic passengers features is mainly analyzed.Third, depending on the study of the characteristics of the subway passengerimage, the passenger detection method is designed. After using adaptive threshold toachieve the image binarization based on the multi-frame average method and extractthe target area, the target area pedestrian is extracted. Based on a template matchingmethod for pedestrian recognition, the algorithm is optimized by image scaletransformation method. Then, the passenger rail security level is designed based onthe core parameters passenger density.Finally, the details of the system software and hardware design are described.this thesis uses Visual C++and Matlab programming to achieve hybrid algorithmsand human-machine interface, and the experimental results verified the system iseffectiveness, timeliness and reliability.This system can achieve functions of video image analysis, passenger flowextraction and security level determination, and can provide the appropriate technicalsupport for rail stations’ daily operational management and security, which hasstrong applicability.