Advertisement

A Hypergraph Reduction Algorithm for Joint Segmentation and Classification of Satellite Image Content

  • Alain Bretto
  • Aurélien Ducournau
  • Soufiane Rital
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6419)

Abstract

In this paper, we introduce a novel hypergraph reduction algorithm, and we evaluate it in an innovative method for joint segmentation and classification of satellite image content. It operates in 3 steps. First, we compute an Image Neighborhood Hypergraph representation (INH). Second, we reduce the INH model and we exploit a morphism from INH to Reduced INH (RINH) to generate superpixels. Then, we perform a superpixels supervised classification according to their features. Our approach is very fast and can deal with great sized images. Its reliability has been tested on several satellite images with comparison to single pixelwise classification.

Keywords

Hypergraph Superpixel Hypergraph Reduction Satellite Image Supervised Classification 

References

  1. 1.
    Agarwal, S., Lim, J., Zelnik-Manor, L., Perona, P., Kriegman, D., Belongie, S.: Beyond pairwise clustering. In: CVPR 2005, vol. 2, pp. 838–845 (June 2005)Google Scholar
  2. 2.
    Aykanat, C., Cambazoglu, B.B., Uçar, B.: Multi-level direct k-way hypergraph partitioning with multiple constraints and fixed vertices. J. Parallel Distrib. Comput. 68(5), 609–625 (2008)CrossRefzbMATHGoogle Scholar
  3. 3.
    Boser, B.E., Guyon, I.M., Vapnik, V.N.: A training algorithm for optimal margin classifiers. In: COLT 1992, pp. 144–152. ACM, New York (1992)Google Scholar
  4. 4.
    Bretto, A.: Introduction to Hypergraph Theory and its Applications to Image Processing. Mongraphy in: Advances in Imaging and Electron Physics, vol. 131, pp. 1–64. Academic Press, London (2004)Google Scholar
  5. 5.
    Cour, T., Benezit, F., Shi, J.: Spectral segmentation with multiscale graph decomposition. In: CVPR, vol. 2, pp. 1124–1131 (June 2005)Google Scholar
  6. 6.
    Ding, L., Yilmaz, A.: Interactive image segmentation using probabilistic hypergraphs. Pattern Recognition 43(5), 1863–1873 (2010)CrossRefzbMATHGoogle Scholar
  7. 7.
    Felzenszwalb, P.F., Huttenlocher, D.P.: Efficient graph-based image segmentation. Intl. Journal of Computer Vision 59(2), 167–181 (2004)CrossRefGoogle Scholar
  8. 8.
    Hanbury, A.: How do superpixels affect image segmentation? In: Ruiz-Shulcloper, J., Kropatsch, W.G. (eds.) CIARP 2008. LNCS, vol. 5197, pp. 178–186. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  9. 9.
    Karypis, G., Aggarwal, R., Kumar, V., Shekhar, S.: Multilevel hypergraph partitioning: applications in vlsi domain. IEEE Trans. Very Large Scale Integr. Syst. 7(1), 69–79 (1999)CrossRefGoogle Scholar
  10. 10.
    Levinshtein, A., Stere, A., Kutulakos, K.N., Fleet, D.J., Dickinson, S.J., Siddiqi, K.: Turbopixels: Fast superpixels using geometric flows. IEEE Trans. Pattern Anal. Mach. Intell. 31(12), 2290–2297 (2009)CrossRefGoogle Scholar
  11. 11.
    Martin, D., Fowlkes, C., Tal, D., Malik, J.: A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics. In: ICCV, vol. 2, pp. 416–423 (July 2001)Google Scholar
  12. 12.
    Ren, X., Malik, J.: Learning a classification model for segmentation. In: ICCV, pp. 10–17 (2003)Google Scholar
  13. 13.
    Rital, S.: Hypergraph cuts & unsupervised representation for image segmentation. Fundamenta Informaticae Journal 96(1-2), 153–179 (2009)MathSciNetGoogle Scholar
  14. 14.
    Torsello, A., Escolano, F., Brun, L. (eds.): GbRPR 2009. LNCS, vol. 5534, pp. 42–51. Springer, Heidelberg (2009)CrossRefGoogle Scholar
  15. 15.
    Wong, A.K.C., Lu, S.W., Rioux, M.: Recognition and shape synthesis of 3-d objects based on attributed hypergraphs. IEEE Trans. Pattern Anal. Mach. Intell. 11(3), 279–290 (1989)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Alain Bretto
    • 1
    • 2
  • Aurélien Ducournau
    • 3
  • Soufiane Rital
    • 2
  1. 1.GREYCCaenFrance
  2. 2.Telecom ParisTech, TSIParisFrance
  3. 3.ENISE-DIPISaint-EtienneFrance

Personalised recommendations