TRAFFIC ANALYSIS USING LOW-COST IMAGE PROCESSING: A REAL-TIME SYSTEM FOR TEMPORARY TRAFFIC CENSUSES

In the first chapter of the report, a brief introduction is given on image processing and the components of an image processing system. In chapter two, research on image processing applied to traffic monitoring and control is outlined, including: (1) the traffic research image processing system (trip) developed in the early 198o's for vehicle counting, automatic speed measurement and vehicle tracking computation and (2) the cellular logic image processor (clip) system designed to manipulate digital images at high speed. The latest research is concerned with counting and tracking moving vehicles in real time, utilizing two basic techniques: (a) image segmentation and (b) vehicle tracking. A new computerised system, traffic analysis using low cost image processing (tulip) is described. The system utilizes an IBM personal computer acting as host to the fg-100 image processor. Vehicle count, speed and headway data from video pictures of a traffic scene is collected. Five different segmentation methods can be utilized on tulip: (i) recursive frame differencing; (II) reference frame differencing; (III) reference value differencing; (IV) signed reference value differencing; and (v) pixel pass filter. These methods are reviewed and compared, and results obtained discussed.

  • Availability:
  • Corporate Authors:

    Newcastle upon Tyne University, England

    Transport Operations Research Group, Claremont Road
    Newcastle NE1 7RU, Tyne and Wear,   England 
  • Authors:
    • Rourke, A
    • Bell, MGH
  • Publication Date: 1989-5

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00498740
  • Record Type: Publication
  • Source Agency: Transport and Road Research Laboratory (TRRL)
  • Files: ITRD, TRIS
  • Created Date: Sep 30 1990 12:00AM