The Moments of the Mixel Distribution and Its Application to Statistical Image Classification

  • Asanobu Kitamoto
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1876)

Abstract

The mixel is a heterogeneous pixel that contains multiple constituents within a single pixel, and the statistical properties of a population of mixels can be characterized by the mixel distribution. Practically this model has a drawback that it cannot be represented in closed form, and prohibitive numerical computation is required for mixture density estimation problem. Our discovery however shows that the “moments” of the mixel distribution can be derived in closed form, and this solution brings about significant reduction of computation cost for mixture density estimation after slightly modifying a typical algorithm. We then show the experimental result on satellite imagery, and find out that the modified algorithm runs more than 20 times faster than our previous method, but suffers little deterioration in classification performance.

Keywords

Probability Distribution Function Satellite Imagery Mixture Density Area Proportion Remote Sensing Image 
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.

References

  1. [1]
    Tabatabai, A.J. and Mitchell, R. Edge Location to Subpixel Values in Digital Imagery. IEEE Trans. Patt. Anal. Mach. Intell., Vol. 6, No. 2, pp. 188–201, 1984.CrossRefGoogle Scholar
  2. [2]
    Santago, P. and Gage, H.D. Quantification of MR Brain Images by Mixture Density and Partial Volume Modeling. IEEE Trans. Med. Img., Vol. 12, No. 3, pp. 566–574, 1993.CrossRefGoogle Scholar
  3. [3]
    Choi, H.S., Haynor, D.R., and Kim, Y. Partial Volume Tissue Classification of Multichannel Magnetic Resonance Images — A Mixel Model. IEEE Trans. Med. Img., Vol. 10, No. 3, pp. 395–408, 1991.CrossRefGoogle Scholar
  4. [4]
    Wang, F. Fuzzy Supervised Classification of Remote Sensing Images. IEEE Trans. Geo. Remote Sens., Vol. 28, No. 2, pp. 194–201, 1990.CrossRefGoogle Scholar
  5. [5]
    Kent, J.T. and Mardia, K.V. Spatial Classification Using Fuzzy Membership Models. IEEE Trans. Patt. Anal. Mach. Intell., Vol. 10, No. 5, pp. 659–671, 1988.MATHCrossRefGoogle Scholar
  6. [6]
    Settle, J.J. and Drake, N.A. Linear Mixing and the Estimation of Ground Cover Proportions. Int. J. Remote Sensing, Vol. 14, No. 6, pp. 1159–1177, 1993.CrossRefGoogle Scholar
  7. [7]
    Foody, G.M. Relating the Land-Cover Composition of Mixed Pixels to Artificial Neural Network Classification Output. Photo. Eng. Rem. Sens., Vol. 62, No. 5, pp. 491–499, 1996.Google Scholar
  8. [8]
    Kitamoto, A. and Takagi, M. Image Classification Using Probabilistic Models that Reflect the Internal Structure of Mixels. Patt. Anal. Appl., Vol. 2, No. 2, pp. 31–43, 1999.MATHCrossRefGoogle Scholar
  9. [9]
    Kitamoto, A. and Takagi, M. Image Classification Using a Stochastic Model that Reflects the Internal Structure of Mixels. In Amin, A., Dori, D., Pudil, P., and Freeman, H., editors, Advances in Pattern Recognition, Vol. 1451 of Lecture Notes in Computer Science, pp. 630–639. Springer, 1998.CrossRefGoogle Scholar
  10. [10]
    Kitamoto, A. and Takagi, M. Area Proportion Distribution-Relationship with the Internal Structure of Mixels and its Application to Image Classification. Syst. Comp. Japan, Vol. 31, No. 5, pp. 57–76, 2000.CrossRefGoogle Scholar
  11. [11]
    Redner, R.A. and Walker, H.F. Mixture Densities, Maximum Likelihood and the EM algorithm. SIAM Review, Vol. 26, No. 2, pp. 195–239, 1984.MATHCrossRefMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2000

Authors and Affiliations

  • Asanobu Kitamoto
    • 1
  1. 1.National Institute of Informatics (NII)TokyoJapan

Personalised recommendations