Testing and Feasibility of VIPS for Traffic Detectionby Alypios E. Chatziioanou,
Stephen L. M. Hockaday,
Carl A. McCarley,
Edward C. Sullivan,
Abstract: The California Department of Transportation (Caltrans) has asked the California Polytechnic State University (Cal Poly) at San Luis Obispo to investigate the effectiveness of existing traffic Video Image Processing Systems (VIPS); as well as the theoretical, hardware and/or financial limitations of VIPS technology. The project is due to be completed in June, 1991. Initial results show that VIPS are both feasible and in some cases well developed as a traffic detection technology. Three 'turnkey' (available for commercial sale) and five 'prototype' VIPS detectors from around the world were identified and tested at Cal Poly. The testing procedure was aimed at evaluating the effectiveness of these products under conditions that are likely to be experienced on the congested California highway system. Variables such as camera positioning, traffic volumes, weather, and lighting conditions were examined. Approaches employed by the different systems vary from performing vehicle tracking throughout the video image to analyzing only specific regions of the picture for vehicle presence. Some of the advantages of each approach which became apparent in the tests will give guidance for future VIPS development. The technology as a whole seems to be extremely effective for several categories of detection problems, and the majority of current limitations are likely to be overcome by further development.
Subject Headings: Traffic signals | Imaging techniques | Feasibility studies | Traffic management | Traffic surveys | Vehicles | Cameras | Signal processing | Tracking | North America | California | United States
Services: Buy this book/Buy this article
Return to search