Object recognition based on pattern features

  • Weijing Zhang
  • Anca Ralescu
Part of the Lecture Notes in Computer Science book series (LNCS, volume 945)


We propose an approach to object recognition based on low level image pattern features. This approach is intended to avoid creating complex high level object model. Our goal is to achieve “quick and (possibly) rough” object recognition from images. An object is defined by a collection of patterns and relations among them. A pattern is a visual clue to an object, such as color, edges texture, etc. Operations like edge detection and region segmentation are performed in a predefined area, which is a small window over which the pattern is defined. Relations among patterns are used to provide constraints on the search space for patterns and for an intelligent pattern matching. Fuzzy variables are used to express patterns in order to build the system's tolerance to imprecision. Sample data is summarized into fuzzy sets. Pattern matching is realized by fuzzy reasoning. The overall evaluation of the matching can be provided by aggregations (i.e. fuzzy integral).


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    M. Sugeno, W. Zhang: An approach to scene understanding using fuzzy case based reasoning, Proc. Sino-Japan Joint Meeting on Fuzzy Sets and Systems, 1990.Google Scholar
  2. [2]
    W. Zhang, M. Sugeno: A Fuzzy Approach to Scene Understanding, Proc. FUZZ-IEEE'93, 564–569, San Francisco, Mar. 28–Apr. 1, 1993.Google Scholar
  3. [3]
    R. A. Brooks: Intelligence without Representation, Artificial Intelligence 47, 139/159, 1991.Google Scholar
  4. [4]
    L. A. Zadeh: Fuzzy logic and soft computing, LIFE lecture, February 1994.Google Scholar
  5. [5]
    Y. Shirai: Three-Dimensional Computer Vision, Springer-Verlag, 1987.Google Scholar
  6. [6]
    D. Marr: Vision, Freeman, 1982.Google Scholar
  7. [7]
    M. Mukunoki, M. Minoh, K. Ikeda: Pixel based object labeling method for out-door scenes, Proc. Asian Conference on Computer Vision'93 (1993–11).Google Scholar
  8. [8]
    J. Gasos, A. Ralescu: Using Environment Information for Guiding Object Recognition. to appear in Proc. IPMU'94.Google Scholar
  9. [9]
    Q. Luong: Color in Computer Vision, Handbook of Pattern Recognition and Computer Vision, 311/368, World Scientific Publishing Company.Google Scholar
  10. [10]
    F. Perez and C. Koch: Toward Color Image Segmentation in Analog VLSI: Algorithm and Hardware, Int'l J. Computer Vision, 12(1), 17/42 (1994).Google Scholar
  11. [11]
    A. Ralescu and J. F. Baldwin: Concept Learning from Examples with Applications to a Vision Learning System, The Third Alvey Vision Conference, Cambridge England, September 15–17 1987.Google Scholar
  12. [12]
    K. Miyajima, T. Norita, A. Ralescu: Management of Uncertainty in Top-down, Fuzzy logic-based Image Understanding of Natural Objects, LIFE TR-3E0011-E, 1992.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1995

Authors and Affiliations

  • Weijing Zhang
    • 1
  • Anca Ralescu
    • 1
  1. 1.The Laboratory for International Fuzzy Engineering ResearchYokohamaJapan

Personalised recommendations