Closed-loop traffic-signal control, freeway-control, and congestion-management systems rely on traffic monitoring and detection systems. The majority of monitoring and detection systems depend on inductive loops. Video-image--based detection systems are gaining popularity because cameras can collect data on several traffic lanes and can be installed quite rapidly on an adjacent pole or bridge. Although users believe the basic hardware architecture of video systems to be well defined, room for improvement is evident in performance and capability of video-detection algorithms. New algorithms are being developed; however, due to technology limitations these algorithms are assessed at a macroscopic level by comparing counts obtained by loop detectors with the image-based detection system. This paper addresses the development of a multimedia database-management system for studying the microscopic performance of video-detection algorithms. The process of acquiring the proper data, a definition for an appropriate database schema, and the peripheral database clients necessary to assess detection algorithms are described.


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00711534
  • Record Type: Publication
  • Files: TRIS, ATRI
  • Created Date: Sep 13 1995 12:00AM