Autonomous Aerial Surveillance of Road Traffic

摘要

This paper deals with a vision system applied to traffic surveillance mounted on airplane reduced model namely unmanned aerial vehicle (UAV). The main objective of this paper is to show how the vehicle recognition is performed using spatiotemporal grouping of primitives formulated as a normalized cuts problem. A verification step based on Dempster Shafer theory is also proposed in order to recognize the vehicles. The UAV used in this application is described in the second section. The images have been collected flying the airplane manually. The UAV will be able to follow a path predefined in future work, it is under development.