• DocumentCode
    793980
  • Title

    VISTA: visual interpretation system for technical applications -architecture and use

  • Author

    Paul, D. ; Hattich, W. ; Nill, W. ; Tatari, S. ; Winkler, G.

  • Author_Institution
    Fraunhofer Inst. of Inf. & Data Process., Karlsruhe, West Germany
  • Volume
    10
  • Issue
    3
  • fYear
    1988
  • fDate
    5/1/1988 12:00:00 AM
  • Firstpage
    399
  • Lastpage
    407
  • Abstract
    The present state of the construction of a vision system for the fast interpretation of complicated industrial scenes is reported. The system is being built at the Fraunhofer Institute for Information and Data Process in Karlsruhe, West Germany. The scenes for which the system is intended are also presented. They form the basis for specifying the requirements for the system called VISTA. Its principal architecture as a bus-oriented multiprocessor system is discussed. Some special hardware modules of VISTA are described in more detail, and its basic software concept is introduced. As an application, detection and classification of defects that appear in automatic visual inspection of hard wood are reviewed. Using special modules for the iconic and the symbolic stage, VISTA is capable of real-time processing and evaluation of endless images, image sequences, images of various formats, and multisensorial data
  • Keywords
    computer architecture; computer vision; computerised pattern recognition; inspection; real-time systems; VISTA; automatic visual inspection; bus-oriented multiprocessor system; computer vision; computerised pattern recognition; image sequences; real-time processing; visual interpretation system; wood inspection; Application software; Computer architecture; Data processing; Electrical equipment industry; Image sequences; Industrial control; Inspection; Layout; Machine vision; Manufacturing industries;
  • fLanguage
    English
  • Journal_Title
    Pattern Analysis and Machine Intelligence, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0162-8828
  • Type

    jour

  • DOI
    10.1109/34.3904
  • Filename
    3904