Abstract: A framework for automatic capturing of traffic flow parameters from drone videos was developed using state-of-the-art techniques in computer vision. This framework would include extracting critical traffic parameters (i.e. traffic volume, density, queue length, and headway timings) from complex traffic scenarios, such as signalised intersections, with high accuracy and high-speed processing.
UAVs or drones hold definite advantages over the traditional traffic sensors as they are well known for easy manoeuvring, low cost, wider field of view and no disturbance on traffic which translates into a safer and quicker data collection strategy. Another advantage of UAV is its top-view perspective, which naturally avoids occlusions in traditional traffic surveillance videos.
Along with the prosperity of UAVs, vehicle detection through computer vision facilitates real-time traffic detection and tracking applications, and capable of withstanding climate condition changes as well as complex traffic scenes.
Authors: Bahaa alddin Mansour
