Crowd Monitoring System Using Unmanned Aerial Vehicle （UAV）
作者机构：Department of Civil Engineering Umm Al-Qura University Makkah 21955 Saudi Arabia Department of Geography Yarmouk University Irbid 21163 Jordan
出 版 物：《Journal of Civil Engineering and Architecture》 (土木工程与建筑（英文版）)
年 卷 期：2017年第11卷第11期
页 面：1014-1024页
学科分类：08[工学] 0813[工学-建筑学] 0814[工学-土木工程]
主 题：UAV, crowd monitoring, crowd density, geo-referencing, mapping.
摘 要：Understanding and dealing with safety aspects of crowd dynamics in mass gatherings of people related to sports, religiousand cultural activities is very important, specifically with respect to crowd risk analysis and crowd safety. Historical trends from theKingdom of Saudi Arabia hosting millions of pilgrims each year during the Hajj and Omrah seasons suggest that stampedes in massgatherings occur frequently and highlight the importance of studying and dealing with the crowd dynamics more scientifically. In thisregard, efficient monitoring and other safe crowd management techniques have been used to minimize the risks associated with suchmass gathering. An example of these techniques is real-time monitoring of crowd using a UAV （Unmanned Aerial Vehicle）; thistechnique is becoming increasingly popular with the objective to save human lives, preserve environment, protect property, keep thepeace, and uphold governmental authority. In this paper, a crowd monitoring system for pedestrians has been proposed and tested. Thesystem has deployed crowd monitoring technique using real-time images taken by UAVs; the collected data was investigated, andcrowd density was estimated using image segmentation procedures. A color-based segmentation method has been employed to detect,identify and map crowd density under different camera positions and orientations. Furthermore, the associated anomalies/outlierswhich may lead to non-classification of features have been eliminated using image enhancement tools. The paper presents a crowdmonitoring system for pedestrians that can contribute to an area of research still in its infancy. The proposed system is a valuable tool interms of facilitating timely decisions, based on highly accurate information. The results show that the used image segmentationtechnique has the capability of mapping the crowd density with an accuracy level up to 80%.
维普期刊数据库