A VISION-BASED ENVIRONMENT RECOGNITION SYSTEM FOR AUTOMATED VEHICLES

This paper describes an on-board vision system for fully-automated highway vehicles. It is required that the fully-automated highway vehicles achieve safe and smooth driving, cooperating with adjacent automated/non-automated vehicles. The vision system recognizes lane lines and leading vehicles, which is essential to lane keeping and vehicle following. Moreover, to ensure safe lane changing, an approaching vehicle from behind is implemented on the vision system. Aimed at accomplishing these tasks in video-rate, the system is provided with four types of specific image processing modules, a central processing unit mode, and a video I/O module. Several experiments have been done on lane line recognition using a template-matching module, preceding vehicle recognition using edge extraction/labeling module, obstacle recognition on the road using stereo vision module, and approaching vehicle recognition using optical flow module. The results obtained by the algorithms implemented on the vision system are presented. The detection of the approaching vehicle, in particular, has proved to be effective to safe lane-changing.

  • Supplemental Notes:
    • Five volumes of papers and one volume of abstracts comprise the published set of conference materials.
  • Corporate Authors:

    VERTIS

    TORANOMOM 34 MORI BUILDING 1-25-5
    TORANOMON, MINATOKU, TOKYO 105  Japan 
  • Authors:
    • Usami, M
    • Suzuki, T
    • Nishida, M
    • Ohta, M
  • Conference:
  • Publication Date: 1995-11

Language

  • English

Media Info

  • Pagination: p. 1263

Subject/Index Terms

Filing Info

  • Accession Number: 00722071
  • Record Type: Publication
  • Report/Paper Numbers: Volume 3
  • Files: TRIS
  • Created Date: Jun 27 1996 12:00AM