Advertisement

Numerical Aspects of Spectral Segmentation on Polygonal Grids

  • Anna Matsekh
  • Alexei Skurikhin
  • Lakshman Prasad
  • Edward Rosten
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7133)

Abstract

We present an implementation of the Normalized Cuts method for the solution of the image segmentation problem on polygonal grids. We show that in the presence of rounding errors the eigenvector corresponding to the k-th smallest eigenvalue of the generalized graph Laplacian is likely to contain more than k nodal domains. It follows that the Fiedler vector alone is not always suitable for graph partitioning, while the eigenvector subspace, corresponding to just a few of the lowest eigenvalues, contains sufficient information needed for obtaining meaningful segmentation. At the same time, the eigenvector corresponding to the trivial solution often carries nontrivial information about the nodal domains in the image and can be used as an initial guess for the Krylov subspace eigensolver. We show that proposed algorithm performs favorably when compared to the Multiscale Normalized Cuts and Segmentation by Weighted Aggregation.

Keywords

image segmentation spectral graph partitioning symmetric eigenvalue problem generalized graph Laplacian 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Fiedler, M.: Algebraic Connectivity of Graphs. Czechoslovak Math. J. 23(98), 298–305 (1973)MathSciNetzbMATHGoogle Scholar
  2. 2.
    Biyikoǧlu, T., Leydold, J., Stadler, P.F.: Laplacian Eigenvectors of Graphs. Perron-Frobenius and Fraber-Krahn Type Theorems, vol. 1915. Springer, Heidelberg (2007)Google Scholar
  3. 3.
    Dhillon, I.S., Guan, Y., Kulis, B.: Weighted Graph Cuts without Eigenvectors: a Multilevel Approach. IEEE Transactions on Pattern Analysis and Machine Intelligence 29(11), 1944–1957 (2007)CrossRefGoogle Scholar
  4. 4.
    Yu, S.X., Shi, J.: Multiclass Spectral Clustering. In: International Conference on Computer Vision, Nice, France, pp. 11–17 (2003)Google Scholar
  5. 5.
    Yu, S.X.: Segmentation using multiscale cues. In: IEEE Conference on Computer Vision and Pattern Recognition, pp. 70–77 (2004)Google Scholar
  6. 6.
    Cour, T., Benezit, F., Shi, J.: Spectral Segmentation with Multiscale Graph Decomposition. In: IEEE International Conference on Computer Vision and Pattern Recognition, CVPR (2005), http://www.seas.upenn.edu/~timothee/software/ncut_multiscale/ncut_multiscale.html
  7. 7.
    Sharon, E., Galun, M., Sharon, D., Basri, R., Brandt, A.: Hierarchy and Adaptivity in Segmenting Visual Scenes. Nature 442, 810–813 (2006), http://www.wisdom.weizmann.ac.il/~swa/index.html CrossRefGoogle Scholar
  8. 8.
    Malik, J., Belongie, S., Leung, T., Shi, J.B.: Contour and Texture Analysis for Image Segmentation. International Journal of Computer Vision 43(1), 7–27 (2001)CrossRefzbMATHGoogle Scholar
  9. 9.
    Cullum, J.K., Willoughby, R.A.: Lanczos algorithms for large symmetric eigenvalue computations, vol. 1. Society for Industrial and Applied Mathematics, Philadelphia (2002)CrossRefzbMATHGoogle Scholar
  10. 10.
    Matsekh, A.M., Shurina, E.P.: On Computing Spectral Decomposition of Symmetric Matrices and Singular Value Decomposition of Unsymmetric Matrices with Guaranteed Accuracy. Optoelectronics, Instrumentation and Data Processing 43(2), 159–169 (2007)CrossRefGoogle Scholar
  11. 11.
    Prasad, L., Skourikhine, A.N.: Vectorized Image Segmentation via Trixel Agglomeration. Pattern Recognition 39, 501–514 (2006)CrossRefzbMATHGoogle Scholar
  12. 12.
    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: Proc. 8th Int’l Conf. Computer Vision, vol. 2, pp. 416–423 (2001)Google Scholar
  13. 13.
    Shi, J.B., Malik, J.: Normalized Cuts and Image Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence 22(8), 888–905 (2000)CrossRefGoogle Scholar
  14. 14.
    Godunov, S.K., Antonov, A.G., Kiriljuk, O.P., Kostin, V.I.: Guaranteed accuracy in numerical linear algebra. Kluwer Academic Publishers Group, Dordrecht (1993); ISBN 0-7923-2352-1. Translated and revised from the 1988 Russian originalGoogle Scholar
  15. 15.
    Matsekh, A.M.: The Godunov-Inverse Iteration: a Fast and Accurate Solution to the Symmetric Tridiagonal Eigenvalue Problem. Applied Numerical Mathematics 54(2), 208–221 (2005)MathSciNetCrossRefzbMATHGoogle Scholar
  16. 16.
    Demmel, J.: Lecture Notes S267. U.C. Berkeley (1996), http://www.cs.berkeley.edu/~demmel/cs267
  17. 17.
    Comaniciu, D., Meer, P.: Mean Shift: A Robust Approach toward Feature Space Analysis. IEEE Trans. Pattern Analysis Machine Intell., 603–619 (2002)Google Scholar
  18. 18.
    Sturm, U., Weidner, U.: Further investigations on segmentation quality assessment for remote sensing applications. In: ISPRS Hannover Workshop 2009 High-Resolution Earth Imaging for Geospatial Information, Hannover, Germany (2009)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Anna Matsekh
    • 1
  • Alexei Skurikhin
    • 1
  • Lakshman Prasad
    • 1
  • Edward Rosten
    • 1
  1. 1.Space and Remote Sensing Sciences Group, Los Alamos National Laboratory, Engineering DepartmentUniversity of CambridgeUK

Personalised recommendations