复杂场景密集人群的流动规律研究-电子与通信工程专业论文.docx

上海交通大学硕士学位论文 上海交通大学硕士学位论文 万方数据 万方数据 RESEARCH ON CROWD FLOWING RULE IN COMPLEX SCENES ABSTRACT With the large-scale activities increasing gradually, crowd behavior analysis becomes more and more important. Also, with more data collected, mining useful information from multi-modal data becomes more popular, such as vector data, text data and video data. Hence, how to analyze crowd behavior in complex scenes from gigantic, multi-modal data becomes a worthy topic. Crowd behavior analysis in complex scenes can enhance the level of urban services, such as detecting human density and easing congestion; it can have applications in video surveillance system such as detecting criminal behavior and unexpected incidents; it can have value in crowd management in large activities such as analyzing crowd flow and confusion degree to prevent stampede, to decrease waiting time and to improve customer satisfaction. Taking Shanghai World Expo as an example, the Expo site recorded gigantic multi-modal data related to various information including tourist flow, ticketing, transportation, activity. The data stored as the form of vector, text, video reflects the real-time information in the most straightforward way. Common methods for multi-modal based crowd behavior analysis are numerical prediction method for tourists flow, feature extraction and event detection in computer vision, pattern analysis and graph clustering based mobile techniques. Difficulties of crowd behavior analysis are multi-modal data, complex scenes, multi-camera, real-time monitoring, high level semantic analysis, etc. To research on crowd behavior analysis in complex scenes while overcoming viewing angle limitations in video surveillance system and data collection privacy limitations in mobile techniques, we mainly focus on two problems: one is prediction for tourists flow distribution; the other one is crowd behavior analysis for different scenes III including simple scene like video surveillance system and comple