Towards Hypothesis Testing and Lossy Minimum Description Length: A Unified Segmentation Framework

  • Mingyang Jiang
  • Chunxiao Li
  • Jufu Feng
  • Liwei Wang
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6494)


We propose a novel algorithm for unsupervised segmentation of images based on statistical hypothesis testing. We model the distribution of the image texture features as a mixture of Gaussian distributions so that multi-normal population hypothesis test is used as a similarity measure between region features. Our algorithm iteratively merges adjacent regions that are “most similar”, until all pairs of adjacent regions are sufficiently “dissimilar”. Standing on a higher level, we give a hypothesis testing segmentation framework (HT), which allows different definitions of merging criterion and termination condition. Further more, we derive an interesting connection between HT framework and previous lossy minimum description length (LMDL) segmentation. We prove that under specific merging criterion and termination condition, LMDL can be unified as a special case under HT framework. This theoretical result also gives novel insights and improvements on LMDL based algorithms. We conduct experiments on the Berkeley Segmentation Dataset, and our algorithm achieves superior results compared to other popular methods including LMDL based algorithms.


Image Segmentation Natural Image Good Segmentation Optimal Segmentation Segmentation Framework 
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.
    Comaniciu, D., Meer, P.: Mean shift: A robust approach toward feature space analysis. PAMI 24, 603–619 (2002)CrossRefGoogle Scholar
  2. 2.
    Shi, J., Malik, J.: Normalized cuts and image segmentation. PAMI 22, 888–905 (2000)CrossRefGoogle Scholar
  3. 3.
    Kim, T.H., Lee, K.M., Lee, S.U.: Learning full pairwise affinities for spectral segmentation. In: CVPR (2010)Google Scholar
  4. 4.
    Felzenszwalb, P.F., Huttenlocher, D.P.: Efficient graph-based image segmentation. IJCV 59, 167–181 (2004)CrossRefGoogle Scholar
  5. 5.
    Zhu, S.C., Tu, Z.W.: Image segmentation by data-driven markov chain monte carlo. PAMI II, 131–138 (2002)Google Scholar
  6. 6.
    Arbelaez, P., Maire, M., Fowlkes, C., Malik, J.: From contours to regions: An empirical evaluation. In: CVPR, pp. 2294–2301 (2009)Google Scholar
  7. 7.
    Ma, Y., Derksen, H., Hong, W., Wright, J.: Segmentation of multivariate mixed data via lossy data coding and compression. PAMI 29, 1546–1562 (2007)CrossRefGoogle Scholar
  8. 8.
    Yang, A.Y., Wright, J., Ma, Y., Sastry, S.S.: Unsupervised segmentation of natural images via lossy data compression. CVIU 110, 212–225 (2008)Google Scholar
  9. 9.
    Rao, S.R., Mobahi, H., Yang, A.Y., Sastry, S.S., Ma, Y.: Natural image segmentation with adaptive texture and boundary encoding. In: Zha, H., Taniguchi, R.-i., Maybank, S. (eds.) ACCV 2009. LNCS, vol. 5994, pp. 135–146. Springer, Heidelberg (2010)CrossRefGoogle Scholar
  10. 10.
    Mobahi, H., Rao, S.R., Yang, A.Y., Sastry, S.S., Ma, Y.: Segmentation of natural images by texture and boundary compression. In: arXiv:1006.3679v1 (2010)Google Scholar
  11. 11.
    Haris, K., Efstratiadis, S.N., Maglaveras, N., Katsaggelos, A.K.: Hybrid image segmentation using watersheds and fast region merging. IEEE Transactions on Image Processing 7, 1684–1699 (1998)CrossRefGoogle Scholar
  12. 12.
    Hsieh, H.K.: On asymptotic optimality of likelihood ratio tests for multivariate normal distributions. The Annals of Statistics 7, 592–598 (1979)MathSciNetCrossRefzbMATHGoogle Scholar
  13. 13.
    Perng, S.K., Littell, R.C.: A test of equality of two normal population means and variances. Journal of the American Statistical Association 71, 968–971 (1976)MathSciNetCrossRefzbMATHGoogle Scholar
  14. 14.
    Dowson, D., Landau, B.: The frechet distance between multivariate normal distributions. Journal of Multivariate Analysis 12, 450–455 (1982)MathSciNetCrossRefzbMATHGoogle Scholar
  15. 15.
    Mori, G.: Guiding model search using segmentation. In: ICCV (2005)Google Scholar
  16. 16.
    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 (2001)Google Scholar
  17. 17.
    Rand, W.M.: Objective criteria for the evaluation of clustering methods. American Statistical Association Journal 66, 846–850 (1971)CrossRefGoogle Scholar
  18. 18.
    Meila, M.: Comparing clusterings: An axiomatic view. In: ICML (2005)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Mingyang Jiang
    • 1
    • 2
  • Chunxiao Li
    • 1
    • 2
  • Jufu Feng
    • 1
    • 2
  • Liwei Wang
    • 1
    • 2
  1. 1.Key Laboratory of Machine PerceptionPeking UniversityBeijingP.R. China
  2. 2.MOE, Department of Machine Intelligence, School of Electronics Engineering and Computer SciencePeking UniversityBeijingP.R. China

Personalised recommendations