Drone video analytics (Main Roads WA, 2019/20)

Drone Analytics

The objectives of this project are to explore new techniques for automated camera calibration and undertake video analytics from drone footage at signalised intersections and roundabouts. Automated camera calibration is necessary to make the system operational. The feasibility of a few new techniques will be explored. Video analytics of signalised intersections will generate results for vehicle counts, average delay, saturation flow, queue length, back-of-queue arrival rate, PCU, blocked turning pocket detection, out of lane detection, gap acceptance at give-ways and preliminary auto-calibration. For roundabouts, video analytics will generate results of vehicle counts, trajectory/path, missed opportunity to enter and gap acceptance.