Scale and segmentation of grey-level images using maximum gradient paths

  • L D Griffin
  • A C F Colchester
  • G P Robinson
5. Segmentation: Multi-Scale, Surfaces And Topology
Part of the Lecture Notes in Computer Science book series (LNCS, volume 511)


We present a technique for the construction of multi-scale representations of grey-level images. Unlike conventional representations the scales are discrete as opposed to continuous and their level is solely determined by the data. The technique is based upon connecting singular points in the image with maximum gradient paths. We also describe two segmentation methods which use the maximum gradient paths generated during the construction of the multi-scale representation. In both segmentation techniques the paths are used to determine significant ridges and troughs. The first technique operates directly on the image, while the second technique uses the magnitude of the image derivative.


morphology ridge trough saddle-point grey-level skeleton multi-scale representation 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Bailes DR (1990) The use of the grey-level SAT to find the salient cavities in echo-cardiograms. In: Proc. Brit. Machine Vison Conf. 1990, Sheffield University Press, pp151–156.Google Scholar
  2. Bischof WF and Caelli TM (1988) Parsing scale-space and spatial stability analysis. Comp. Vision Graph. Image Process. 42:192–205.Google Scholar
  3. Blum H (1973) Biological shape and visual science (part 1). Internat. J. Theo. Biol. 38:205–287.Google Scholar
  4. Brelstaff JG, Ibison MC and Elliott PJ (1990) Edge-region integration for segmentation of MR images. In: Proc. Brit. Machine Vision Conf. 1990, Sheffield University Press, pp139–144.Google Scholar
  5. Burns JB, Hanson AR and Riseman EM (1986) Extracting straight lines. IEEE Trans. Patt. Anal. and Machine Intell. PAMI-8:425–455.Google Scholar
  6. Caelli TM, Brettel H, Rentschler I and Hilz R (1983) Discrimination thresholds in the two-dimensional spatial frequency domain. Vision Res. 23:129–133.Google Scholar
  7. Canny J (1986) A computational approach to edge detection. IEEE Trans. Patt. Anal. and Machine Intell. PAMI-8:679–698.Google Scholar
  8. Colchester ACF (1990) Network representation of 2D and 3D images. In: 3D Imaging in Medicine. Hoehne KH, Fuchs H, Pizer SM (eds), Springer-Verlag, Berlin, pp45–62.Google Scholar
  9. Colchester ACF, Ritchings RT and Kodikara ND (1988) A method for multi-scale representation of data sets based on maximum gradient profiles: initial results on angiographic images. In: Proc. NATO ASI meeting on The Formation, Handling and Evaluation of Medical Images, Sep. 1988. Todd-Pokropek A and Viergever M (eds).Google Scholar
  10. Colchester ACF, Ritchings RT and Kodikara ND (1990) A new approach to image segmentation using maximum gradient profiles orthogonal to edges. Image and Vision Computing. 8:211–217.Google Scholar
  11. Delaunay B (1943) Sur la sphere vide. Bull. Acad. Sci. USSR(VII), Classe Sci. Nat. pp793–800.Google Scholar
  12. Fu KS and Mui JK (1981) A survey on image segmentation. Patt. Recog. 3:3–16.Google Scholar
  13. Gauch JM and Pizer SM (1988) Image description via the multi-resolution intensity axis of symmetry. In: Proc. of 2nd ICCV 1988, pp269–274.Google Scholar
  14. Hawkes DJ, Hill DLG, Lehmann ED, Robinson GP, Maisey MN, Colchester ACF (1990) Preliminary work on the interpretation of SPECT images with the aid of registered MR images and an MR derived 3D neuro-anatomical atlas. In: 3D Imaging in Medicine, Hoehne KH, Fuchs H, Pizer SM (eds), Springer-Verlag, Berlin pp241–252.Google Scholar
  15. Hummel RA (1986) Representations based on zero-crossings in scale-space. In: Proc. IEEE Conf. on Comp. Vision and Patt. Recog. pp204–209.Google Scholar
  16. Koenderink JJ, van Doorn AJ (1984) The structure of images. Trans. Biol. Cyb. 50:363–370.Google Scholar
  17. Koenderink JJ (1990) Solid Shape. MIT Press, Cambridge MA.Google Scholar
  18. Kovasznay LS and Joseph HM (1955) Image processing. In: Proc. IRE 43:560–570.Google Scholar
  19. Lee DT and Schachter BJ (1980) Two algorithms for constructing a Delaunay triangulation. Internat. J. Comp. Info. Sci. 9:219–242.Google Scholar
  20. Marr DC and Hildreth E (1980) Theory of edge detection. Proc. Roy. Soc. B B-207:187–217.Google Scholar
  21. Morgenthaler DC and Rosenfeld A (1981) Multidimensional edge detection by hypersurface fitting. IEEE Trans. Patt. Anal. Machine Intell. PAMI-3:482–486.Google Scholar
  22. Nazif AM and Levine MD (1984) Low level image segmentation: an expert system. IEEE Trans. Patt. Anal. and Machine Intell. PAMI-6:555–577.Google Scholar
  23. Rosin PL, Colchester ACF and Hawkes DJ (1990) Early visual representations using regions defined by maximum gradient profiles between singular points. In: Information Processing in Medical Imaging, Ortendahl D and Llacer J (eds), Wiley-Liss, New York. pp369–388.Google Scholar
  24. Rosenfeld A and Thurston M (1972) Edge and curve detection for visual scene analysis. IEEE Trans. Comput. 21:562–569.Google Scholar
  25. Voronoi G (1908) Nouvelles applications des parametres a la theorie des formes quadratiques. Deuxieme Memoire: Recherches sur la paralleloedres. Deuexieme reiene angnew. Math. 134, pp198–287.Google Scholar
  26. Witkin AP (1983) Scale-space filtering. In: Proc. 7th Internat. Joint Conf. Art. Intell. pp1019–1022.Google Scholar
  27. Yuille AL and Poggio TA (1986) Scaling theorems for zero-crossings. IEEE Trans. Pat. Anal. Machine Intell. PAMI-8:15–25.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1991

Authors and Affiliations

  • L D Griffin
    • 1
  • A C F Colchester
    • 1
  • G P Robinson
    • 1
  1. 1.Department of NeurologyUMDS, Guy's HospitalLondonEngland

Personalised recommendations