航站楼旅客群体性事件预警监控系统架构及关键技术

Architecture and key technology of an early warning and monitoring system for mass events in terminal

针对航站楼旅客群体性事件的智能化预警问题,在民用机场航站楼内的现有视频监控系统的基础上,提出旅客群体性事件预警监控系统架构和关键技术。针对航站楼内的视频监控系统的视频数据进行预处理;基于空间聚类算法,从图像中提取人群聚集特征;针对目标监控区域的历史视频和突发事件数据进行统计分析,研究正常和非正常两种情况下的人群分布特征,并建立相应的预警规则库;根据预警规则,在人群聚集特征识别的基础上,对非正常目标密集人群进行识别。该系统可根据航站楼旅客群集性事件的时空特性,针对不同事件、不同时间、不同地点采用不同的规则进行预警,提高了预警的可靠性。

The architecture of an early warning and monitoring system for mass events in terminal was constructed on foundation of terminal video surveillance system in civil airports. The system preprocessed video data of the sur- veillance system and extracted the accumulation features of crowds on spatial clustering algorithm. History videos and emergency data of monitored area were analyzed statistically, besides, crowd distribution trends in both normal and abnormal situations were studied. Based on above work, an early warning rules library was built. The system was able to recognize abnormal dense crowds with the help of previous rules and crowds feature recognitions. Final- ly, different crowd events in different time and different locations were pre-warned under different rules, based on characteristics of time and space of crowd events in terminal.