Nikos Papanikolopoulos, Professor, Computer Science and Engineering
An important component of a real-time traffic control system is the acquisition, processing, and interpretation of the available sensory information regarding the traffic flow. We propose the use of computer -vision techniques in conjunction with a vision sensor in order to collect data about ramps. This data can be used in a variety of applications (e.g., design of ramps). We will use state-of-the-art hardware and imaging methods to design a ramp data collection system. The experimental verification of our approach will be performed with real images and field-data. Our system is different from other commercially available traffic vision systems since it can not only detect vehicles but can also track and classify them. The proposed imaging techniques are not limited to ramps. They have a general applicability. We plan to compute parameters such as ramp queue length, ramp queue occupancy, vehicle velocity, ramp volume, ramp queue wait, classes of vehicles, etc. Finally, we will use detectors to gather ground-truth data.