Journal of Mathematical Imaging and Vision

, Volume 3, Issue 4, pp 349–376

Discrete derivative approximations with scale-space properties: A basis for low-level feature extraction

  • Tony Lindeberg
Article

Abstract

This article shows how discrete derivative approximations can be defined so thatscale-space properties hold exactly also in the discrete domain. Starting from a set of natural requirements on the first processing stages of a visual system,the visual front end, it gives an axiomatic derivation of how a multiscale representation of derivative approximations can be constructed from a discrete signal, so that it possesses analgebraic structure similar to that possessed by the derivatives of the traditional scale-space representation in the continuous domain. A family of kernels is derived that constitutediscrete analogues to the continuous Gaussian derivatives.

The representation has theoretical advantages over other discretizations of the scale-space theory in the sense that operators that commute before discretizationcommute after discretization. Some computational implications of this are that derivative approximations can be computeddirectly from smoothed data and that this will giveexactly the same result as convolution with the corresponding derivative approximation kernel. Moreover, a number ofnormalization conditions are automatically satisfied.

The proposed methodology leads to a scheme of computations of multiscale low-level feature extraction that is conceptually very simple and consists of four basic steps: (i)large support convolution smoothing, (ii)small support difference computations, (iii)point operations for computing differential geometric entities, and (iv)nearest-neighbour operations for feature detection.

Applications demonstrate how the proposed scheme can be used for edge detection and junction detection based on derivatives up to order three.

Key words

scale-space visual front end smoothing Gaussian filtering Gaussian derivative discrete approximation edge detection junction detection multiscale representation computer vision digital signal processing 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    M. Abramowitz and I.A. Stegun,Handbook of Mathematical Functions, Applied Mathematics Series, vol. 55, National Bureau of Standards, Gaithersburg, MD, 1964.Google Scholar
  2. 2.
    J. Babaud, A.P. Witkin, M. Baudin and R.O. Duda, “Uniqueness of the Gaussian kernel for scale-space filtering,”IEEE Trans. Patt. Anal. Mach. Intell. vol. 8, pp. 26–33, 1986.Google Scholar
  3. 3.
    J. Blom, “Topological and geometrical aspects of image structure,”Ph.D. thesis, Department of Medical Physics, University of Utrecht, Utrecht, The Netherlands, 1992.Google Scholar
  4. 4.
    K., Brunnström, T.P. Lindeberg and J.-O. Eklundh, “Active detection and classification of junctions by foveation with a head—eye system guided by the scale-space primal sketch,” inProc. 2nd European Conference on Computer Vision, Santa Margherita Ligure, Italy, May 19–22, 1992, pp. 701–709.Google Scholar
  5. 5.
    P.J. Burt and E.H. Adelson, “The Laplacian pyramid as a compact image code,”IEEE Trans. Commun. vol. 9, pp. 532–540, 1983.Google Scholar
  6. 6.
    J. Canny, “A computational approach to edge detection,”IEEE Trans. Patt. Anal. Mach. Intell. vol. 8, pp. 679–698, 1986.Google Scholar
  7. 7.
    J.L. Crowley and R.M. Stern “Fast computation of the difference of low pass transform,”IEEE Trans. Patt. Anal. Mach. Intell. vol. 6, pp. 212–222, 1984.Google Scholar
  8. 8.
    G. Dahlquist, Å. Björk, and N. Anderson,Numerical Methods, Prentice-Hall: London, 1974.Google Scholar
  9. 9.
    R. Deriche, “Using Canny's criteria to derive a recursively implemented optimal edge detector,”Int. J. Comput. Vis., vol. 5, pp. 167–187, 1987.Google Scholar
  10. 10.
    L.M.J. Florack, B.M. ter Haar Romeny, J.J. Koenderink and M.A. Viergever, “General intensity transformations and second order invariants,” inProc. 7th Scandinavian Conference on Image Analysis, Aalborg, Denmark, Aug. 13–16, 1991, pp. 338–345.Google Scholar
  11. 11.
    L.M.J. Florack, B.M. ter Haar Romeny, J.J. Koenderink, M.A. Viergever, “Scale and the differential structure of images,”Image Vis. Comput. vol. 10, pp. 376–388, 1991.Google Scholar
  12. 12.
    W. Freeman, and E. Adelson, “The design and use of steerable filters for image analysis and wavelet decomposition,” inProc. 3rd International Conference on Computer Vision, Osaka, Japan, 1990, pp. 406–415.Google Scholar
  13. 13.
    R.M. Haralick, “Digital step edges from zero-crossings of second directional derivatives,”IEEE Trans. Patt. Anal. Mach. Intell., vol. 6, 1984.Google Scholar
  14. 14.
    L. Kitchen and R. Rosenfeld, “Gray-level corner detection,”Patt. Recogn. Lett., vol. 1, pp. 95–102, 1982.Google Scholar
  15. 15.
    J.J. Koenderink, “The structure of images,”Biol. Cybernet. vol. 50, pp. 363–370, 1984.Google Scholar
  16. 16.
    J.J. Koenderink and A.J. van Doorn, “Representation of local geometry in the visual system,”Biol. Cybernet. vol. 55, pp. 367–375, 1987.Google Scholar
  17. 17.
    J.J. Koenderink and W. Richards, “Two-dimensional curvature operators,”J. Opt. Soc. Am., vol. 5, pp. 1136–1141, 1988.Google Scholar
  18. 18.
    J.J. Koenderink and A.J. van Doorn, “Receptive field families,”Biol. Cybernet., vol. 63, pp. 291–297, 1990.Google Scholar
  19. 19.
    A.F. Korn, “Toward a symbolic representation of intensity changes in images,”IEEE Trans. Patt. Anal. Mach. Intell., vol. 10, pp. 610–625, 1988.Google Scholar
  20. 20.
    T. Lindeberg, “Scale-space for discrete signals,”IEEE Trans. Patt. Anal. Mach. Intell., vol. 12, pp. 234–254, 1990.Google Scholar
  21. 21.
    T. Lindeberg,Discrete scale-space theory and the scale-space primal sketch, Ph.D. thesis, Royal Institute of Technology, Stockholm; revised and extended version is to appear as bookScale-Space Theory in Early Vision in the Kluwer International Series in Engineering and Computer Science.Google Scholar
  22. 22.
    T. Lindeberg and J.O. Eklundh, “On the computation of a scale-space primal sketch,”J. Vis. Commun. Image Repro. vol. 2, pp. 55–78, 1991.Google Scholar
  23. 23.
    T. Lindeberg “Scale-space behaviour of local extrema and blobs,”J. Math. Imag. Vis. vol. 1, pp. 65–99, 1992.Google Scholar
  24. 24.
    T. Lindeberg “Scale-spaceforN-dimensional discrete signals,” inProc. Shape in Picture, Driebergen, The Netherlands, Sept. 7–11, 1992. Ying et al. (eds.) NATO ASI Series F, Springer-Verlag (in press).Google Scholar
  25. 25.
    T. Lindeberg, “Scale-space behaviour and invariance properties of differential singularities,” inProc. Shape in Picture, Driebergen, The Netherlands, Sept. 7–11, 1992. Ying et al. (eds.) NATO ASI Series F, Springer-Verlag (in press).Google Scholar
  26. 26.
    T. Lindeberg, “On scale selection for differential operators,” inProc. 8th Scandinavian Conference on Image Analysis Tromsö, Norway, May 1992, pp. 857–866.Google Scholar
  27. 27.
    T. Lindeberg and J. Gårding, “Shape from texture from a multi-scale perspective,” inProc. 4th International Conference on Computer Vision, Berlin, May 11–14, 1993, pp. 683–691; An extended version is available as a technical report from the Royal Institute of Technology, Stockholm.Google Scholar
  28. 28.
    D. Marr and E. Hildreth “Theory of edge detection,”Proc. Roy. Soc. London, Ser. B, vol. 207, pp. 187–217, 1980.Google Scholar
  29. 29.
    P. Meer and I. Weiss, “Smoothed differentiation filters for images,”J. Vis. Commun. Image Repro., vol. 3, pp. 58–72, 1992.Google Scholar
  30. 30.
    N. Nordström, “Biased anisotropic diffusion — a unified regularization and diffusion approach to edge detection,”Proc. 2nd Conference on Computer Vision, Antibes, France, 1990, pp. 18–27.Google Scholar
  31. 31.
    P. Perona and J. Malik, “Scale space and edge detection using anisotropic diffusion,” inProc. IEEE Workshop on Computer Vision, Miami, FL, 1987, pp. 16–22.Google Scholar
  32. 32.
    P. Perona, “Steerable-scalable kernels for edge detection and junction analysis,” inProc. 2nd European Conference on Computer Vision, Santa Margherita Ligure, Italy, May 19–22, 1992, pp. 3–18.Google Scholar
  33. 33.
    T. Vieville and O.D. Faugeras, “Robust and fast computation of unbiased intensity derivatives in images,” inProc. 2nd European Conference on Computer Vision, Santa Margherita Ligure, Italy, May 19–22, 1992, pp. 203–211.Google Scholar
  34. 34.
    A.P. Witkin, “Scale-space filtering,” inProc. 8th International Joint Conference on Artificial Intelligence, Karlsruhe, West Germany, Aug. 8–12, 1983, pp. 1019–1022.Google Scholar
  35. 35.
    A. Yuille and T. Poggio, “Scaling theorems for zerocrossings,”IEEE Trans. Patt. Anal. Mach. Intell., vol. 9, pp. 15–25, 1986.Google Scholar
  36. 36.
    W. Zhang and F. Bergholm, “An extension of Marr's signature based edge classification,” inProc. 7th Scandinavian Conference on Image Analysis, Aalborg, Denmark, Aug. 13–16, 1991, pp. 435–443.Google Scholar

Copyright information

© Kluwer Academic Publishers 1993

Authors and Affiliations

  • Tony Lindeberg
    • 1
  1. 1.Department of Numerical Analysis and Computing Science, Royal Institute of TechnologyComputational Vision and Active Perception LaboratoryStockholmSweden

Personalised recommendations