Extraction of Traffic Parameters at Urban Intersections Using the Low-Altitude Unmanned Aerial Vehicle Image

With the rapid development of civilian unmanned aerial vehicle (UAV), real-time traffic state-monitoring based on UAV has attracted traffic researchers' attention. UAV holds great value for traffic information collection because of its high mobility, wide field of view, freedom from the limitation of road networks, and low cost. There are many potential applications of UAV video in traffic monitoring of urban intersections. In this study, taking the urban intersection as the research object, the queue length, and OD matrix are proposed based on the vehicle detection of Viola-Jones, the support vector machine (SVM) and the tracking method of location prediction along with color similarity. The estimation results can correctly assess traffic congestion state and fully understand the source, flow, and evolution of traffic in the urban intersections.