A multiresolution shape description algorithm

  • Gabriella Sanniti di Baja
  • Edouard Thiel
Part of the Lecture Notes in Computer Science book series (LNCS, volume 719)


A multiresolution shape description algorithm is presented, which is adequate to describe patterns perceived as the superposition of elongated regions. The weighted skeleton of the pattern is partitioned into a number of subsets, each corresponding to a pattern subset having simple shape, by means of a polygonal approximation. Different levels of description are possible, depending on the tolerance adopted during the approximation process. The computational cost of the algorithm is rather modest. A compact representation of the pattern is obtained, that takes simultaneously into account the representations of the pattern at the different levels.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    H. Blum, R. N. Nagel: Shape description using weighted symmetric axis features. Pattern Recognition 10, 167–180 (1978)Google Scholar
  2. 2.
    A. Montanvert: Medial line — graph representation and shape description. Proc. 8th Int. Conf. on Pattern Recognition. Paris, 430–432 (1986)Google Scholar
  3. 3.
    E. Thiel, A. Montanvert: Shape splitting from medial lines using the 3–4 chamfer distance. In: C. Arcelli et al. (eds.): Visual Form Analysis and Recognition. New York: Plenum 1992, pp. 537–546Google Scholar
  4. 4.
    A. Rosenfeld, J. L. Pfaltz: Sequential operations in digital picture processing. J. ACM 13, 471–494 (1966)Google Scholar
  5. 5.
    G. Borgefors: Distance Transformation in Digital Images. Comput. Vision Graphics Image Process 34, 344–371 (1986)Google Scholar
  6. 6.
    E. Thiel, A. Montanvert: Chamfer masks — discrete distance functions, geometrical properties and optimization. Proc. 11th Int. Conf. on Pattern Recognition. The Hague, 244–247 (1992)Google Scholar
  7. 7.
    G. Sanniti di Baja: Well-shaped, stable and reversible skeletons from the (3, 4)-distance transform. Journal of Visual Communication and Image Representation. Accepted for publication (1992)Google Scholar
  8. 8.
    C. Arcelli, R. Colucci, G. Sanniti di Baja: On the description of digital strips. Proc. Int. Conf. on Artificial Intelligence Applications and Neural Networks. Zurich, 193–196, (1990)Google Scholar
  9. 9.
    L. Dorst: Pseudo-Euclidean skeletons. Proc. 8th Int. Conf. on Pattern Recognition. Paris, 286–288 (1986)Google Scholar
  10. 10.
    C. Arcelli, M. Frucci: Reversible skeletonization by (5, 7, 11)-erosion. In: C. Arcelli et al. (eds.): Visual Form Analysis and Recognition. New York: Plenum 1992, pp. 21–28Google Scholar
  11. 11.
    J. M. Chassery, A. Montanvert: Géométrie discrète en analyse d'images. Paris: Editions Hermès 1992Google Scholar
  12. 12.
    Y. Xia: Skeletonization via the realization of the fire front propagation and extinction in digital binary shapes. IEEE Trans. Patt. Anal. Mach. Intell 11, 1076–1086 (1989)Google Scholar
  13. 13.
    H. Blum: A transformation for extracting new descriptors of shape. In: W. Wathen-Dunn (ed.): Models for the Perception of Speech and Visual Form. Cambridge: M. I. T. Press 1967, pp. 362–380Google Scholar
  14. 14.
    C. Arcelli, G. Sanniti di Baja: Finding local maxima in a pseudo Euclidean distance transform. Comput. Vision Graphics Image Process. 43, 361–367 (1988)Google Scholar
  15. 15.
    T. Pavlidis: Structural pattern recognition. New York: Springer Verlag 1977Google Scholar
  16. 16.
    C. Arcelli: Pattern thinning by contour tracing. Comput. Graphics Image Process. 17, 130–144 (1981)Google Scholar
  17. 17.
    A. R Dill, M. D. Levine, P. B. Noble: Multiple resolution skeletons. IEEE Trans. Patt. Anal. Mach. Intell. 9, 495–504 (1987)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1993

Authors and Affiliations

  • Gabriella Sanniti di Baja
    • 1
  • Edouard Thiel
    • 2
  1. 1.Istituto di CiberneticaCNRNaplesItaly
  2. 2.Equipe TIMC-IMAG, CERMO BP 53 XGrenoble cxFrance

Personalised recommendations