Indexing Spatially Sensitive Distance Measures Using Multi-resolution Lower Bounds

  • Vebjorn Ljosa
  • Arnab Bhattacharya
  • Ambuj K. Singh
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3896)


Comparison of images requires a distance metric that is sensitive to the spatial location of objects and features. Such sensitive distance measures can, however, be computationally infeasible due to the high dimensionality of feature spaces coupled with the need to model the spatial structure of the images.

We present a novel multi-resolution approach to indexing spatially sensitive distance measures. We derive practical lower bounds for the earth mover’s distance (EMD). Multiple levels of lower bounds, one for each resolution of the index structure, are incorporated into algorithms for answering range queries and k-NN queries, both by sequential scan and using an M-tree index structure. Experiments show that using the lower bounds reduces the running time of similarity queries by a factor of up to 36 compared to a sequential scan without lower bounds. Computing separately for each dimension of the feature vector yields a speedup of ~14. By combining the two techniques, similarity queries can be answered more than 500 times faster.


Lower Bound Feature Vector Image Retrieval Linear Programming Problem Index Structure 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Swedlow, J.R., Goldberg, I., Brauner, E., Sorger, P.K.: Informatics and quantitative analysis in biological imaging. Science 300, 100–102 (2003)CrossRefGoogle Scholar
  2. 2.
    Manjunath, B.S., Salembier, P., Sikora, T. (eds.): Introduction to MPEG 7: Multimedia Content Description Language. Wiley (2002)Google Scholar
  3. 3.
    Mahalanobis, P.: On the generalised distance in statistics. In: Proc. Nat. Inst. Sci., India, vol. 12, pp. 49–55 (1936)Google Scholar
  4. 4.
    Lewis, G.P., Guerin, C.J., Anderson, D.H.: Rapid changes in the expression of glial cell proteins caused by experimental retinal detachment. Am. J. of Ophtalmol 118, 368–376 (1994)Google Scholar
  5. 5.
    Rubner, Y., Tomasi, C., Guibas, L.J.: The earth mover’s distance as a metric for image retrieval. International Journal of Computer Vision 40, 99–121 (2000)zbMATHCrossRefGoogle Scholar
  6. 6.
    Werman, M., Peleg, S., Rosenfeld, A.: A distance metric for multi-dimensional histograms. Computer, Vision, Graphics, and Image Proc 32, 328–336 (1985)CrossRefGoogle Scholar
  7. 7.
    Peleg, S., Werman, M., Rom, H.: A unified approach to the change of resolution: Space and gray-level. IEEE Trans. PAMI 11, 739–742 (1989)Google Scholar
  8. 8.
    Levina, E., Bickel, P.: The earth mover’s distance is the Mallows distance: Some insights from statistics. Proc. ICCV 2, 251–256 (2001)Google Scholar
  9. 9.
    Stricker, M.A., Orengo, M.: Similarity of color images. In: Niblack, C.W., Jain, R.C. (eds.) Storage and Retrieval for Image and Video Databases III. Vol. 2420 of Proceedings of SPIE, pp. 381–392 (1995)Google Scholar
  10. 10.
    Grauman, K., Darrell, T.: Fast contour matching using approximate earth mover’s distance. In: Proc. CVPR (2004)Google Scholar
  11. 11.
    Lazebnik, S., Schmid, C., Ponce, J.: Sparse texture representation using affineinvariant neighborhoods. In: Proc. CVPR (2003)Google Scholar
  12. 12.
    Typke, R., Veltkamp, R., Wiering, F.: Searching notated polyphonic music using transportation distances. In: Proc. Int. Conf. Multimedia, pp. 128–135 (2004)Google Scholar
  13. 13.
    Demirci, M.F., Shokoufandeh, A., Dickinson, S., Keselman, Y., Bretzner, L.: Many-to-many feature matching using spherical coding of directed graphs. In: Pajdla, T., Matas, J(G.) (eds.) ECCV 2004. LNCS, vol. 3021, pp. 322–335. Springer, Heidelberg (2004)CrossRefGoogle Scholar
  14. 14.
    Lavin, Y., Batra, R., Hesselink, L.: Feature comparisons of vector fields using earth mover’s distance. In: Proc. of the Conference on Visualization, pp. 103–109 (1998)Google Scholar
  15. 15.
    Hillier, F.S., Lieberman, G.J.: Introduction to Mathematical Programming, 1st edn. McGraw-Hill, New York (1990)Google Scholar
  16. 16.
    Klee, V., Minty, G.: How good is the simplex algorithm. In: Shisha, O. (ed.) Inequalities, vol. III, pp. 159–175. Academic Press, New York (1972)Google Scholar
  17. 17.
    Weber, R., Schek, H.J., Blott, S.: A quantitative analysis and performance study for similarity-search methods in high-dimensional spaces. In: Proc. VLDB, pp. 194–205 (1998)Google Scholar
  18. 18.
    Ciaccia, P., Patella, M., Zezula, P.: M-tree: An efficient access method for similarity search in metric spaces. In: Proc. VLDB, pp. 426–435 (1997)Google Scholar
  19. 19.
    Ciaccia, P., Patella, M.: Bulk loading the M-tree. In: Proc. ADC (1998)Google Scholar
  20. 20.
    Indyk, P., Thaper, N.: Fast image retrieval via embeddings. In: Proc. Internat. Workshop on Statistical and Computational Theories of Vision (2003)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Vebjorn Ljosa
    • 1
  • Arnab Bhattacharya
    • 1
  • Ambuj K. Singh
    • 1
  1. 1.University of CaliforniaSanta BarbaraUSA

Personalised recommendations