Journal of the Korean Institute of Intelligent Systems (한국지능시스템학회논문지)

Korean Institute of Intelligent Systems (한국지능시스템학회)
권호 표지
DOI QR코드

DOI QR Code

Incremental EM algorithm with multiresolution kd-trees and cluster validation and its application to image segmentation

다중해상도 kd-트리와 클러스터 유효성을 이용한 점증적 EM 알고리즘과 이의 영상 분할에의 적용

  • Lee, Kyoung-Mi (Department of Computer Science, Duksung Women's University)
  • 이경미 (덕성여자대학교 컴퓨터학과)
  • Received : 2015.08.13
  • Accepted : 2015.10.30
  • Published : 2015.12.25
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we propose a new multiresolutional and dynamic approach of the EM algorithm. EM is a very popular and powerful clustering algorithm. EM, however, has problems that indexes multiresolution data and requires a priori information on a proper number of clusters in many applications, To solve such problems, the proposed EM algorithm can impose a multiresolution kd-tree structure in the E-step and allocates a cluster based on sequential data. To validate clusters, we use a merge criteria for cluster merging. We demonstrate the proposed EM algorithm outperforms for texture image segmentation.

본 논문은 효율적인 영상 분할을 수행하기 위한 다중해상도와 동적인 성질을 가지고 있는 새로운 EM 알고리즘을 제안한다. EM 알고리즘은 가장 많이 사용되고 성능이 우수한 클러스터링 방법이다. 그러나, 기존의 EM 알고리즘은 다중해상도 데이터 처리에 대한 문제점과 클러스터 개수에 대한 사전 지식 요구라는 단점을 가지고 있다. 본 논문에서는 이러한 문제점을 해결하기 위해서 E-단계에 다중해상도 kd-트리를 적용함으로써 다중해상도 데이터 처리 문제를 해결하였고, 순차적 데이터에 따라 클러스터를 할당할 수 있데 하였다. 클러스터의 유효성을 검사하기 위해서, 클러스터 병합 원칙을 이용한다. 본 논문에서는 제안하는 알고리즘을 텍스쳐 영상 분할에 적용하였고, 우수한 성능을 보였다.

Keywords

References

  1. A.P. Dempster, N.M. Laird and D.B. Rubin, "Maximum likelihood from incomplete data via the EM algorithm (with discussion)," Journal of the royal statistical society B, vol. 39, no. 1, pp. 1-38, 1977.
  2. S.J. Nowlan, "Soft competitive adaptation: neural network learning algorithms based on fitting statistical mixtures," PhD thesis, School of Computer Science, Carnegie Mellon University, 1991.
  3. J. Xu, G. Ye, Y. Wang, G. Herman, B. Zhang and J. Yang, "Incremental EM for probabilistic latent semantic analysis on human action recognition," Proceedings of the IEEE conference on advanced video and signal based surveillance, 2009.
  4. S.-K. Ng, G.J. McLachlan and A.H. Lee, "An incremental EM-based learning approach for online prediction of hospital resource utilization," Artificial Intelligence in Medicine, vol. 36, pp. 257-267, 2006. https://doi.org/10.1016/j.artmed.2005.07.003
  5. K.-M. Lee, "Elliptical clustering with incremental growth and its application to skin color region segmentation," Journal of the Korean Information Science Society, vol. 31, no. 9, pp. 1161-1170, 2004.
  6. S.-S. Kim and J.-H. Kang, "Improved expectation and maximization via a new method for initial values," Journal of the Korean Institute of Intelligent Systems, vol. 13, no. 4, pp. 416-426, 2003. https://doi.org/10.5391/JKIIS.2003.13.4.416
  7. L. Xu, M.I. Jordan, and G.E. Hinton, "An alternative model for mixtures of experts," Advances in neural information processing systems, vol. 7, pp. 633-640, 1995.
  8. A.W. Moore, "Very fast EM-based mixture model clustering using multiresolution kd-tree," Advances in neural information processing systems, vol. 11, pp. 543-549, 1999.
  9. S.-K. Ng and G.J. McLachlan, "Speeding up the EM algorithm for mixture model-based segmentation of magnetic resonance images," Pattern recognition, vol. 37, no. 8, pp. 1573-1589, 2004. https://doi.org/10.1016/j.patcog.2004.02.012
  10. J. Platt, "A resource-allocating network for function interpolation," Neural computation, vol. 3, no. 2, pp. 213-225, 1991. https://doi.org/10.1162/neco.1991.3.2.213
  11. S.-B. Roh, T.-C. Ahn, Y.-S. Baek and Y.-S. Kim, "Space partition using context fuzzy c-Means algorithm for image segmentation," Journal of the Korean Institute of Intelligent Systems, vol. 20, no. 3, pp. 368-374, 2010. https://doi.org/10.5391/JKIIS.2010.20.3.368
  12. P. Brodatz, Textures: A photographic album for artists and designers, Dover publications, NY, 1966.
  13. J.G. Daugman, "An information- theoretic view of analogue representation in striate cortex," Computational neuroscience, In E.L. Schwartz (Eds.), Cambridge, MA: MIT Press, pp. 403-424, 1990.
  14. T. Randen and J.H. Husoy, "Filtering for texture classification: a comparative study," IEEE transactions on pattern analysis and machine intelligence, vol. 21, no. 4, pp. 291-310, 1999. https://doi.org/10.1109/34.761261