Picture Labeling and Shape Descriptors for Machine Vision

  • C. Arcelli
  • G. Sanniti di Baja
Part of the International Centre for Mechanical Sciences book series (CISM, volume 307)


Picture labeling can be used to give structure to the region of the digital plane occupied by a single-valued figure. Different labeling methods can be adopted, depending on the features one desires to extract. The distance transformation provides an adequate picture labeling when shape is of interest. In this paper, figures digitized on the square grid as well as on the hexagonal grid are taken into account. Distance transforms approximating at different extent the Euclidean distance transform are discussed, and criteria to identify simple shape features, as the local maxima and the layers, are introduced. More powerful shape descriptors can also be extracted, e.g., the skeleton. The structure of the distance transform allows the detection of all the skeletal pixels by using only local operations, and with low computational effort. The conditions to detect the skeletal pixels on the distance transform computed according to the city block distance function are given. Finally, it is shown how the distance transform constitutes a useful starting point to construct the Voronoi Diagram of a set of sources, not necessarily consisting of single pixels.


Local Maximum Voronoi Diagram Edge Point Hexagonal Grid Distance Transformation 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Rosenfeld, A. and J.L. Pfaltz, Sequential operations in digital picture processing, J.ACM, 13 (1976), 471–494.Google Scholar
  2. 2.
    Hilditch, J. and D. Rutovitz, Chromosome recognition, Ann. New York Acad. Sci., 157 (1969), 339–364.CrossRefGoogle Scholar
  3. 3.
    Borgefors, G., Distance transformations in arbitrary dimensions, CVGIP, 27 (1984), 321–345.Google Scholar
  4. 4.
    Borgefors, G., Distance transformations in digital images, CVGIP, 34 (1986), 344–371.Google Scholar
  5. 5.
    Borgefors, G., Distance transformations in hexagonal grids, Image Analysis and Processing 2, Cantoni, Di Gesu’, Levialdi Eds., Plenum Press (1988), 213–220.Google Scholar
  6. 6.
    Arcelli, C. and G. Sanniti di Baja, Weighted distance transforms: a characterization, Image Analysis and Processing 2, Cantoni, Di Gesu, Levialdi Eds., Plenum Press (1988), 205–212.Google Scholar
  7. 7.
    Arcelli, C. and G. Sanniti di Baja, A width-independent fast thinning algorithm, IEEE Trans. PAMI, 7 (1985), 463–474.CrossRefGoogle Scholar
  8. 8.
    Dorst, L., Pseudo-Euclidean skeletons, Proc. 8-th ICPR, Paris, France, (1986), 286–288.Google Scholar
  9. 9.
    Arcelli, C. and G. Sanniti di Baja, A speeded up thinning algorithm, Proc. ISELDECS 87, Kharagpur, India, (1987), vol.2, 378–380.Google Scholar
  10. 10.
    Borgefors G. and G. Sanniti di Baja, Skeletonizing the distance transform on the hexagonal grid, Proc. 9-th ICPR, Rome, Italy, 1988 (in press).Google Scholar
  11. 11.
    Mase, K., J.I. Toriwaki and T. Fukumura, Modified digital Voronoi diagram and its application to image processing, Trans. IECE of Japan, J 64-D (11), (1981), 1029–1036.Google Scholar
  12. 12.
    Arcelli, C. and G. Sanniti di Baja, Computing Voronoi diagrams in digital pictures, Pattern Recognition Letters, 4, (1986), 383–389.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Wien 1989

Authors and Affiliations

  • C. Arcelli
    • 1
  • G. Sanniti di Baja
    • 1
  1. 1.Istituto di Cibernetica del CNRArco Felice, (NA)Italy

Personalised recommendations