DETECTION AND TRACKING OF VEHICLES IN LOW FRAMERATE AERIAL IMAGE SEQUENCES *

Traffic monitoring requires mobile and flexible systems that are able to extract densely sampled spatial and temporal traffic data in large areas in near-real time. Video-based systems mounted on aerial platforms meet these requirements, however, at the expense of a limited field of view. To overcome this limitation of video cameras, we are currently developing a system for automatic derivation of traffic flow data which is designed for commercial medium format cameras with a resolution of 25-40 cm and a rather low frame rate of only 1-3 Hz. In addition, the frame rate is not assumed to be constant over time. Novel camera systems as for instance DLR’s 3K camera image a scene with “bursts”, thereby each burst consisting of several frames. After a time gap of few seconds for readout, the next burst starts etc. This kind of imaging results in an along-track overlap of 90 % (and more) during bursts and less than 50% between bursts. These peculiarities need to be considered in the design of an airborne traffic monitoring system. We tested the system with data of several flight campaigns, for which also ground-truth data in form of car tracks is available. The evaluation of the results shows the applicability and the potentials of this approach