KBED: A Knowledge-Based Edge Detection System

  • Cyril Boucher
  • Christian Daul
  • Pierre Graebling
  • Ernest Hirsch
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 978)


The importance of applying different kinds of knowledge in image processing has been recognized for a long time. But a common belief is that reasoning and use of explicit knowledge can only be useful in the domain of high level processing and especially for image interpretation. As a result, usually, for the earlier steps of processing, emphasis is on filtering operations for edge detection and localization. But even the most elegant and powerful of these techniques does not give satisfactory results: the higher the accuracy to achieve in e.g. edge localization, the more unsatisfactory appear results of edge detectors. In order to fill this gap in the whole processing chain of vision applications, our contribution describes a knowledge-based approach for this kind of processing.

We are convinced that early processing can benefit from problem solving techniques and from symbolic reasoning. A primary goal of our work is thus to design a knowledge-based system that manages to improve the edge detection and localization in images of simple quasi-polyhedral manufactured parts. An implementation of such a knowledge-based system for early vision is to be described in this contribution.


Knowledge-based Systems and Engineering Applications Blackboard Architecture Vision and Image Processing Edge Detection 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    V. Berzins. ”Accuracy of Laplacian edge detector”. Comput. Vision, Graphics, and Image Process., (27), 195–210 (1984).Google Scholar
  2. 2.
    C. Boucher and E. Hirsch. ”A blackboard expert system for improving edge detection”. Technical Report, Université Louis Pasteur, Strasbourg/France (1995).Google Scholar
  3. 3.
    E. Bourennane, M. Paindavoine and F. Truchetet ”Amélioration du filtre Canny Deriche pour la détection des contours sous forme de rampe”. Traitement du signal, 10(4), 297–310 (1993).Google Scholar
  4. 4.
    J. F. Canny. ”A computational approach to edge detection”. IEEE Trans. PAMI-8(6), 679–698 (1986).Google Scholar
  5. 5.
    D. Crevier. ”Expert systems as design aids for artificial vision systems: a survey”. SPIE Intelligent Robots and Computer Vision, 2055(12), 84–96 (1993).Google Scholar
  6. 6.
    D. Forsyth and A. Zisserman. ”Mutual illumination”. In Proc. IEEE Conf. Comput. Vision Pattern Recog., 466–473 (1989).Google Scholar
  7. 7.
    D. Forsyth and A. Zisserman. ”Reflections on shading”. IEEE Trans. PAMI-13(7), 671–679 (1991).Google Scholar
  8. 8.
    Y. Lu and R. C. Jain. ”Reasoning about edges in scale space”. IEEE Trans. PAMI-14(4), 450–467 (1992).Google Scholar
  9. 9.
    E. Micheli, B. Caprile, P. Ottonello, and V. Torre. ”Localization and noise in edge detection”. IEEE Trans. PAMI-11(10), 1106–1116 (1989).Google Scholar
  10. 10.
    V. Lesser N. Carver. ”Evolution of blackboard control architectures”. Expert Systems with Applications, 7, 1–30 (1994).Google Scholar
  11. 11.
    A. M. Nazif and D. Levine. ”Low level image segmentation: an expert system”. IEEE Trans. PAMI-6(5), 555–577 (1984).Google Scholar
  12. 12.
    H. P. Nii. ”Blackboard systems at the architecture level”. Expert Systems with Applications, 7, 43–54 (1994).Google Scholar
  13. 13.
    P. Paillou. ”Une architecture parallèle hétérogène pour la vision par ordinateur: application à la comparaison d'images réelles et conceptuelles pour la mesure dimensionnelle.”. PhD thesis, Université Louis Pasteur, Strasbourg/France (1992).Google Scholar
  14. 14.
    P. Perona and J. Malik. ”Detecting and localizing edges composed of steps, peaks and roofs”. In 3rd Int. Conf. on Computer Vision ICCV'90, 52–57, Osaka/Japan, december 1990.Google Scholar
  15. 15.
    T. Morgan and R. S. Engelmore, editors. Blackboard Systems, chapter 1, pages 1–22. Addison-Wesley, 1988.Google Scholar
  16. 16.
    J. Shen and S. Castan. ”An optimal linear operator for step edge detection”. Comput. Vision, Graphics, and Image Process., 54(2), 112–133 (1992).Google Scholar
  17. 17.
    F. Ulupinar and G. Medioni. ”Refining edges detected by LoG operator”. Comput. Vision, Graphics, and Image Process., (51), 275–298 (1990).Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1995

Authors and Affiliations

  • Cyril Boucher
    • 1
    • 2
  • Christian Daul
    • 1
    • 2
  • Pierre Graebling
    • 1
    • 2
  • Ernest Hirsch
    • 1
    • 2
  1. 1.Laboratoire des Sciences de l'ImageDe l'Informatique et de la Télédétection (URA CNRS 1871)IllkirchFrance
  2. 2.Ecole Nationale Supérieure de Physique de StrasbourgUniversité Louis PasteurIllkirchFrance

Personalised recommendations