Advertisement

A neural network for image smooothing and segmentation

  • Herbert Jahn
Shape Representation and Image Segmentation
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1451)

Abstract

Older work of the author on image smoothing and segmentation with graph based parallel-sequential processing structures gave encouraging results but was not able to segment images with very strong noise. Therefore, a modified method was developed. Instead of the used “hard” Pixel Adjacency Graph (PAG) which was used formerly, a “soft” or fuzzy PAG is defined via a degree of adjacency of 4-neighbored pixels. Furthermore, the averaging over 4-neighbors is applied recursively using a nonlinear weighting function which is closely connected with the degree of adjacency and which guarantees efficient noise reduction, edge preserving, and adaptation. The discrete nonlinear dynamic equation system describing the averaging process can be implemented with a Discrete Time Cellular Neural Network (CNN). Its stable states are the smoothed images. Then the soft PAG describing the edge strength' and the hard PAG defining the segments can be calculated. The method now can cope with strong noise. Some results demonstrate its smoothing and segmenting capability.

Keywords

image preprocessing segmentation edge preserving smoothing edge detection graphs neural networks 

References

  1. 1.
    H.C. Nothdurft, “Common properties of visual segmentation”, in: “Higher-order processing in the visual system” (Eds. G.R. Bock, J.A. Goode), J. Wiley & Sons, Chichester, 1994Google Scholar
  2. 2.
    H. Jahn, “Image segmentation with a layered graph network”, SPIE Proceedings, Vol. 2662, “Nonlinear Image Processing VII”, pp. 217–228, 1996Google Scholar
  3. 3.
    H. Jahn, “A graph structure for image segmentation”, SPIE Proceedings, Vol. 3026, “Nonlinear Image Processing VIII”, pp. 198–208, 1997Google Scholar
  4. 4.
    H. Jahn, “A graph structure for grey value and texture segmentation”, GbR'97: Workshop on Graph based Representations, Lyon 1997, Computing Suppl. 12, Springer-Verlag, in pressGoogle Scholar
  5. 5.
    J.M. Jolion, and A. Montanvert, “The Adaptive Pyramid: A Framework for 2D Image Analysis”, CVGIP: Image Understanding 55, pp. 339–348, 1992CrossRefGoogle Scholar
  6. 6.
    W.G. Kropatsch, and S.B. Yacoub, “A revision of pyramid segmentation”, Proc. of ICPR'96, pp. 477–481, 1996Google Scholar
  7. 7.
    P. Meer, “Stochastic Image Pyramids”, CVGIP 45, pp. 269–294, 1989Google Scholar
  8. 8.
    P. Nacken, “Image Segmentation by Connectivity Preserving Relinking in Hierarchical Graph Structures”, Pattern Recognition 28, pp. 907–920, 1995CrossRefGoogle Scholar
  9. 9.
    L. O. Chua and L. Yang, “Cellular neural networks: Theory”, IEEE Trans. on Circuits and Systems 35, pp. 1257–1272, 1988CrossRefGoogle Scholar
  10. 10.
    T. Roska and J. Vandevalle (Eds.), Cellular Neural Networks,: J. Wiley&Sons, West Sussex, 1993Google Scholar
  11. 11.
    P. Perona, T. Shiota, and J. Malik, “Anisotropic diffusion”, in: B. ter Haar Romeny (Ed.), Geometry-Driven Diffusion in Computer Vision, pp. 73–92, Kluwer Academic Publishers, Dordrecht, 1994Google Scholar
  12. 12.
    I. Bajla, “Geometry-driven diffusion — an alternative approach to image filtering/segmentation in diagnostic imaging”, 6th Int. Workshop on Digital Image Processing and Computer Graphics (DIP'97), Vienna, 1997, SPIE Proceedings 3346, pp. 56–71Google Scholar
  13. 13.
    I. Bajla and I. Hollander, “Pixel dissimilarities for local controlling of the conductance in geometry-driven diffusion”, 6th Int. Workshop on Digital Image Processing and Computer Graphics (DIP'97), Vienna, 1997, SPIE Proceedings 3346, pp. 120–128Google Scholar
  14. 14.
    H. Jahn, “Dot pattern clustering using a cellular neural network”, 6th Int. Workshop on Digital Image Processing and Computer Graphics (DIP'97), Vienna, 1997, SPIE Proceedings 3346, pp. 298–307Google Scholar
  15. 15.
    T. Pavlidis, Structural Pattern Recognition, Springer-Verlag,Berlin, 1977Google Scholar
  16. 16.
    T. Lindeberg and B. M. ter Haar Romeny; “Linear Scale Space l: Basic Theory”, in: B. ter Haar Romeny (Ed.), Geometry-Driven Diffusion in Computer Vision, pp. 1–38, Kluwer Academic Publishers, Dordrecht 1994Google Scholar
  17. 17.
    H. Jahn, “Eine Methode zur Clusterbildung in metrischen Rdumen”, Bild & Ton 39, pp. 362–370, 1986Google Scholar
  18. 18.
    J. Zhang and D. Wang, Image segmentation by multigrid Markov random field optimization and perceptual considerations; J. Electronic Imaging 7, pp. 52–60, 1998CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • Herbert Jahn
    • 1
  1. 1.Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR)Institut für WeltraumsensorikBerlinGermany

Personalised recommendations