Nikos Papanikolopoulos, Professor, Computer Science and Engineering
This project works toward the development and deployment of a flexible, portable, and reliable intersection-monitoring and data-collection system, based on vehicle-tracking methodologies implemented on one or more video cameras. Research activities included computation of a variety of traffic and behavior data such as turning-vehicle counts, vehicle trajectories, vehicle classes, delays, lane changes, gap acceptance behavior by turning vehicles, speed variations, safety-related data, and other types of intersection data as needed by traffic engineers. The proposed system could be used for accessing the effectiveness of existing signal timing plans and operational methods and the level of intersection safety. The researchers' approach included the development of a user-friendly interface and the employment of existing camera hardware. This report presents solutions to the problem of reliable target detection and tracking in unconstrained outdoor environments as they pertain to vision-based data collection at traffic intersections. The researchers also envisioned the proposed system to be used as the main tool for before/after study of the effectiveness of intersection and local area improvement treatments.