IMPROVED ILLUMINATION ASSESSMENT FOR VISION-BASED TRAFFIC MONITORING

This paper presents methods for determining parameter values or algorithms for vision systems that must operate autonomously over varying environmental conditions. It looks at measures for determining levels of visibility, contrast, and shadow presentation. The methods are applied in the domain of traffic monitoring systems. Performance data for 34 representative installations are provided including an improved algorithm for determining whether a scene is well-lit or not.

• Supplemental Notes:
  • Publication Date: 1998 Published By: IEEE Computer Society Press, Los Alamitos CA
• Corporate Authors:

  University of Sussex. School of Cognitive and Computing Sciences

  ,    
• Authors:
  • Wixson, L
  • Hanna, K
  • Mishra, D
• Conference:
  • IEEE Workshop on Visual Surveillance (1998 : Bombay, India)
• Publication Date: 1998

Language

• English

Media Info

• Pagination: p. 34-41

Subject/Index Terms

• TRT Terms: Automatic vehicle monitoring; Computer algorithms; Computer vision

Filing Info

• Accession Number: 00776089
• Record Type: Publication
• Source Agency: UC Berkeley Transportation Library
• Files: PATH
• Created Date: Nov 17 1999 12:00AM