Efficacy of Video Incident Detection System for Advanced Traffic Management Under Non-Adherence of Lane Discipline

This paper reports on a study of the effectiveness of a video incident detection system (VIDS) used for advanced traffic management in India. The authors note that India is characterized by traffic conditions where there is non-adherence to lane discipline. In this study, video image processing using Citilog software was applied on one urban section for traffic incident detection and measurement of traffic parameters like traffic flow. The main function of the media video processing system is the automatic incident detection and the recording of video sequences showing the different phases of an accident from its causes to its consequences. In addition, the system is able to provide the operator with various types of traffic easements and the system diagnostic indicators, including alarms upon detection of a vehicle stopped on traffic lane in fluid traffic conditions, a vehicle stopped on traffic lane in congested traffic conditions, general slow down on traffic lanes, a slow-moving vehicle, a fast-moving vehicle, or a vehicle traveling against the flow of traffic. The authors discuss the equipment needed, the set-up and installation parameters, and the reliability of the VIDS (including the frequency of false alarms). They found an accuracy level of up to 85% in the field tests. The authors support the use of Intelligent Transportation Systems (ITS) as an option for innovative and cost effective solutions for achieving maximum capacity out of the existing highway facilities, thereby enhance traffic safety. They conclude that VIDS can be successfully implemented under Indian traffic conditions where there is nonadherence of lane discipline for advanced traffic management system.


  • English

Media Info

  • Media Type: CD-ROM
  • Features: Appendices; Figures; Photos; References; Tables;
  • Pagination: 21p
  • Monograph Title: ITS Connections: Saving Time. Saving Lives

Subject/Index Terms

Filing Info

  • Accession Number: 01142028
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Oct 5 2009 7:07AM