COLLISION-AVOIDANCE SYSTEM BASED ON OPTICAL FLOW

The authors propose a collision-avoidance system which warns the driver of impending collision. It is based on the parallel computation of the image- motion velocity vector field over a large part of the image plane. This image-motion field is then used as the input to a neural network which computes two properties of the movement of the vehicle relative to the environment.

• Supplemental Notes:
  • Publication Date: 1992 Published By: Institute of Electrical and Electronics Engineers, New York
• Corporate Authors:

  Princeton University

  Department of Civil Engineering and Operations Research
  Princeton, NJ  United States  08544

  Rockwell International Science Center

  ,    

  Carnegie Mellon University

  School of Computer Science
  Pittsburgh, PA  United States  15213

  University of Pennsylvania, Philadelphia

  Department of Computer and Information Science
  Philadelphia, PA  United States  19104

  Regie nationale des usines Renault. Direction de la recherche

  ,    

  Crain Communications, Incorporated

  1155 Gratiot Avenue
  Detroit, MI  United States  48207-2997

  Nihon Denso Kabushiki Kaisha

  ,    

  Ecole nationale superieure des mines de Paris

  ,    

  Daimler-Benz A.G.

  Mercedesstrade 136
  Stuttgart-Unterturk,   Germany 

  Matsushita Denki Sangyo

  ,    

  Nissan Jidosha Kabushiki Kaisha

  ,    

  Honda Gijutsu Kenkyujo

  ,    

  Daihatsu Jidosha Kabushiki Kaisha

  ,    

  Istituto per la Ricerca Scientifica e Tecnologica (Trento, Italy)

  ,    
• Authors:
  • Hatsopoulos, Nicholas
  • Anderson, James A
• Conference:
  • Intelligent Vehicles '92 Symposium (1992 : Detroit, Mich.). Proceedings of the Intelligent Vehicles '92 Symposium
• Publication Date: 1992

Language

• English

Media Info

• Pagination: p. 79-84

Subject/Index Terms

• TRT Terms: Computer vision; Crash avoidance systems; Neural networks; Parallel processing
• Subject Areas: Data and Information Technology;

Filing Info

• Accession Number: 00784760
• Record Type: Publication
• Source Agency: UC Berkeley Transportation Library
• Files: PATH
• Created Date: Nov 17 2000 12:00AM