Techniques And System About Identification Of Large Scale Passenger Flow And Assessment Of Security Status Of Beijing Mass Transit Railway

With rapid development of urban rail transit and the network, the big scale operation has brought the impact of normalization "blowout" passenger flow to the network. Rail transit stations as the nodes of the network, passenger congestion occur frequently, limited resources and space between the supply and demand is more prominent. How to realize the accurate surveillance, recognition, data acquisition and security status assessment of passenger flow have become a key problem. It is also important to improve the operation and safety of rail transportation. Paper is based on theoretical research, made applied research as the goal, use computer vision and image technology to achieve better results in the research about recognition and safety assessment of passengers. This paper based on the National Natural Science Foundation of "The evolution of large-scale crowded pedestrian and dynamics model of complex transportation systems"(61004105), studies the safety problems of large passenger congestion and the application of urban rail transit stations.The paper first introduces the development of researches about monitoring, identification and safety assessment of rail passengers, and summarizes domestic and international research results, and clarifies the significance of this study. Then defines the large scale passenger flow, its classification and feature, uses the image processing technology to identify the large scale passenger flow, and then proves its legitimacy and accuracy through the statistics and research, proposes the recognition functions for the follow research.Based on these studies, paper introduces the technology about pedestrian identification and tracking, establishes a database for rail transit passenger flow through the analysis of its characteristics, then passenger flow data acquisition system is developed to achieve semi-automatic and automatic pedestrian data collection, this system also provides a basis data collection tool for the management and research of passenger flow. On this basis, paper Investigate and analyze different values of passenger flow characteristics, assess the state of passenger safety, propose the evaluation indicators (such as passenger group characteristics, individual characteristics of pedestrians, traffic conflicts) and assessment system, and add it into passenger data acquisition system to monitor and display passenger security status dynamically.Finally, paper uses the pedestrian video screened at Lama Temple channel as an example, calculates relevant indicators, assesses its security status, and then proves the rationality and availability of passenger flow data acquisition system on pedestrian identification, monitoring, evaluation and security management.The paper provides a wealth of ideas and a more theoretical basis for the future research of urban rail transit and provides a more efficient and reliable tool for the future security management in the rail transit.