Scale space: Its natural operators and differential invariants

  • Bart M. ter Haar Romeny
  • Luc M. J. Florack
  • Jan J. Koenderink
  • Max A. Viergever
5. Segmentation: Multi-Scale, Surfaces And Topology
Part of the Lecture Notes in Computer Science book series (LNCS, volume 511)


Why and how one should study a scale-space is prescribed by the universal physical law of scale invariance, expressed by the so-called Pi-theorem. The fact that any image is a physical observable with an inner and outer scale bound, necessarily gives rise to a ‘scale-space representation’, in which a given image is represented by a one-dimensional family of images representing that image on various levels of inner spatial scale. An early vision system is completely ignorant of the geometry of its input. Its primary task is to establish this geometry at any available scale. The absence of geometrical knowledge poses additional constraints on the construction of a scale-space, notably linearity, spatial shift invariance and isotropy, thereby defining a complete hierarchical family of scaled partial differential operators: the Gaussian kernel (the lowest order, rescaling operator) and its linear partial derivatives. They enable local image analysis in a robust way, while at the same time capturing global features through the extra scale degree of freedom. The operations of scaling and differentiation cannot be separated. This framework permits us to construct in a systematic way multiscale, orthogonal differential invariants, i.e. true image descriptors that exhibit manifest invariance with respect to a change of cartesian coordinates. The scale-space operators closely resemble the receptive field profiles in the mammalian front-end visual system.


Scale-space gaussian kernel gaussian derivatives differential invariants 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Babaud, J., Witkin, A. P., Baudin, M., and Duda, R. O. Uniqueness of the gaussian kernel for scale-space filtering. IEEE Trans. Pattern Anal. Machine Intell. 8, 1 (1986), 26–33.Google Scholar
  2. Brunnström, K., Eklundh, O., and Lindeberg, T. On scale and resolution in the analysis of local image structure. In First European Conf. on Computer Vision, ECCV 90 (Antibes, Fr., 1990), pp. 3–13.Google Scholar
  3. Burt, P., Hong, T. H., and Rosenfeld, A. Segmentation and estimation of image region properties through cooperative hierarchical computation. IEEE Tr. SMC 11 (1981), 802–825.Google Scholar
  4. Canny, J. A computational approach to edge detection. IEEE Trans. Pattern Anal. Machine Intell. 8, 6 (1987), 679–698.Google Scholar
  5. Clark, J. J. Authenticating edges produced by zero-crossing algorithms. IEEE Trans. Pattern Anal. Machine Intell. 11, 1 (1989), 43–57.Google Scholar
  6. Cooper, N. G., and West, G. B., Eds. Particle physics. Cambridge University Press, Los Alamos National Laboratory, 1988, ch. Scale and Dimension, pp. 4–21.Google Scholar
  7. Danielsson, P. E., and Seger, O. Generalized and separable sobel operators. In Machine Vision for Three Dimensional Scenes, H. Freeman, Ed. Academic Press Inc., London, 1990.Google Scholar
  8. De Micheli, E., Caprile, B., Ottonello, P., and Torre, V. Localization and noise in edge detection. IEEE Trans. Pattern Anal. Machine Intell. 10, 11 (1989), 1106–1117.Google Scholar
  9. Hildreth, E. C. Implementation of a theory of edge detection. A. i. memo 579, Massachusetts Inst. Technol., Cambridge, 1980.Google Scholar
  10. Hubel, D. H., and Wiesel, T. N. Receptive fields, binocular interaction, and functional architecture in the cat's visual cortex. J. Physiol. 160 (1962), 106–154.Google Scholar
  11. Jain, A. K. Fundamentals of digital image processing. Prentice Hall, Englewood Cliffs, NJ, 1989, ch. 9.Google Scholar
  12. Kanatani, K. Group-theoretical methods in image understanding, vol. 20 of Series in Information Sciences. Springer-Verlag, 1990.Google Scholar
  13. Kay, D. C. Tensor calculus. Schaum's Outline Series. McGraw-Hill Book Campany, New York, 1988.Google Scholar
  14. Koenderink, J. J. The structure of images. Biol. Cybern. 50 (1984), 363–370.Google Scholar
  15. Koenderink, J. J., and van Doorn, A. J. Representation of local geometry in the visual system. Biol. Cybern. 55 (1987), 367–375.Google Scholar
  16. Koenderink, J. J., and van Doorn, A. J. Receptive field families. Biol. Cybern. 63 (1990), 291–298.Google Scholar
  17. Korn, A. F. Toward a symbolic representation of intensity changes in images. IEEE Trans. Pattern Anal. Machine Intell. 10, 5 (1988), 610–625.Google Scholar
  18. Lindeberg, T. Scale-space for discrete signals. IEEE Trans. Pattern Anal. Machine Intell. 12, 3 (1990), 234–245.Google Scholar
  19. Mallat, S. G. A theory for multiresolution signal decomposition: the wavelet representation. IEEE Trans. Pattern Anal. Machine Intell. 11, 7 (1989), 674–694.Google Scholar
  20. Marr, D. C., and Hildreth, E. C. Theory of edge detection. Proc. Roy. Soc. London B 207 (1980), 187–217.Google Scholar
  21. Olver, P. J. Applications of Lie groups to differential equations, vol. 107 of Graduate texts in mathematics. Springer-Verlag, 1986.Google Scholar
  22. Poston, T., and Steward, I. Catastrophe theory and its applications. Pitman, London, 1978.Google Scholar
  23. Spivak, M. A comprehensive introduction to differential geometry, Publish or Perish Inc., Berkeley, 1970.Google Scholar
  24. ter Haar Romeny, B. M., and Florack, L. M. J. A multiscale geometric model of human vision. In Perception of visual information, B. Hendee and P. N. T. Wells, Eds. Springer-Verlag, Berlin, 1991. In press.Google Scholar
  25. Torre, V., and Poggio, T. A. On edge detection. IEEE Trans. Pattern Anal. Machine Intell. 8, 2 (1986), 147–163.Google Scholar
  26. Witkin, A. P. Scale space filtering. In Proc. IJCAI (Karlsruhe, W. Germany, 1983), pp. 1019–1023.Google Scholar
  27. Young, R. A. The gaussian derivative model for machine vision: visual cortex simulation. J. Opt. Soc. Amer. (July 1986).Google Scholar
  28. Young, R. A. Simulation of human retinal function with the gaussian derivative model. In Proc. IEEE CVPR (Miami, Fla., 1986), pp. 564–569.Google Scholar
  29. Yuille, A. L., and Poggio, T. A. Scaling theorems for zero-crossings. IEEE Trans. Pattern Anal. Machine Intell. 8 (1986), 15–25.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1991

Authors and Affiliations

  • Bart M. ter Haar Romeny
    • 1
  • Luc M. J. Florack
    • 1
  • Jan J. Koenderink
    • 1
  • Max A. Viergever
    • 1
  1. 1.Computer Vision Research GroupUtrecht University HospitalUtrechtThe Netherlands

Personalised recommendations