阐述一款根据景区的监控视频自动识别游客并实时统计人流数据的系统.系统主要采用YOLO(You Only Look Once)框架识别游客,并根据游客的轨迹判断其进出景区的情况.同时采用实时处理框架Flink对游客进出数据按时间进行排序和动态统计,从而为景区管理人员提供及时准确的人流信息.
Design of Scenic Area Pedestrian Flow Monitoring System Based on YOLO and Flink
This paper describes a system that automatically identifies tourists based on surveillance videos of scenic spots and calculates real-time flow data.The system mainly uses the YOLO(You Only Look Once)framework to identify tourists and judge their entry and exit of the scenic area based on their trajectories.At the same time,it uses the real-time processing framework Flink to sort and dynamically count tourist entry and exit data by time,providing timely and accurate crowd information for scenicarea management personnel.
computer engineeringscenic data monitoringcrowd flow statisticsYOLOFlink
NETL
NSTL
万方数据
