查看文章

To ensure the safety of people in massively crowded activities, crowd monitoring systems are very important. Presently, many researches focus on studying video for crowd monitoring, and realize the recognition of behavior patterns. However, most systems only implement a real-time display of the monitored scene to comprehensively model a place. Because the environment at crowd gatherings is complex and dynamic, an intelligent monitoring system is necessary. This paper proposes a limo-powered crowd monitoring system for modeling dynamic crowds. On one hand, limo-system realizes life modeling at the individual level, offering personalized cognition of each person. On the other hand, it uses data collected by multi-sensor devices and intelligent algorithms deployed on the edge cloud to dynamically perceive the environment. At the same time, considering the redundancy of monitoring data, the value of information can be recognized via information cognition components in the edge cloud. Finally, combining the results of environment awareness and life modeling, simulation experiments of crowd scenes are executed for multi-scenarios. Results show that the intelligent system can be effectively applied to crowd monitoring in comprehensive environments, with important significance for crowd perception and modeling.