Journal of the Institute of Electronics and Information Engineers (전자공학회논문지)

The Institute of Electronics and Information Engineers (대한전자공학회)
권호 표지
DOI QR코드

DOI QR Code

Image Thresholding Based on Within-Class Standard Deviation

클래스 내 표준편차 기반의 문턱치 처리에 의한 영상분할

  • Sung, Jung-Min (School of Electronics Engineering, Kyungpook national university) ;
  • Ha, Ho-Gun (School of Electronics Engineering, Kyungpook national university) ;
  • Choi, Bong-Yeol (School of Electronics Engineering, Kyungpook national university)
  • Received : 2013.03.12
  • Published : 2013.07.25
Download PDF
⟨ Previous Next ⟩

Abstract

The within-class variance of Otsu's method is moderate but improper in expressing class statistical distributions. Otsu's method uses a variance to represent the distribution of each class. The variance utilizes a distance square from the mean to a data. This process is not proper in denoting a real class statistical distribution because of the distance square. In this paper, to express more exact class statistical distributions, the within-class standard deviation as a criterion for threshold selection is proposed and then the optimal threshold is determined by minimizing it. In order to have validity, it is shown through the experimental results that the proposed method was more superior to the counterparts.

영상분할에 사용되는 문턱치 처리 방법들 중 Otsu 방법은 클래스 내 분산(within-class variance)을 이용하여 최적의 문턱치를 자동으로 추정한다. 이때, Otsu 방법은 각 클래스(class)의 통계적 분포를 표현함에 있어 분산을 사용하며, 이러한 분산은 평균으로부터 해당 자료까지의 거리 제곱으로 표현된다. 그 결과, Otsu 방법의 최적 문턱치는 분산의 크기에 큰 영향을 받으며, 분산들 중 크기가 큰 쪽으로 편향되는 문제점을 보인다. 이에 본 논문은 분산을 표준편차로 변경함으로써 이러한 현상을 감소시켰으며, 보다 정확한 문턱치를 추정할 수 있었다. 본 논문은 기존의 클래스 분산(class variance)을 클래스 표준편차(class standard deviation)로 대체하였으며, 문턱치 선택 기준으로서 클래스 내 표준편차(within-class standard deviation)을 제안하였다. 타당성을 검증하기 위해 두 개의 정규분포 히스토그램(histogram) 및 음영이 있는 영상들에 대해 모의실험을 수행하였으며, 제안된 방법을 Otsu 방법 및 기존의 방법들과 비교하였다. 또한, 객관적 성능평가(Misclassification Error)를 통해 제안된 방법의 우수성을 확인하였다.

Keywords

References

  1. S. Chen and D. Li, "Image binarization focusing on objects," Neurocomputin, vol. 69, pp. 2411-2415, Oct. 2006. https://doi.org/10.1016/j.neucom.2006.02.014
  2. Rafael C. Gonzalez and Richard E. Woods, Digital Image Processing 3rd edition, Prentice Hall, 2002.
  3. M. Sezgin and B. Sankur, "Survey over image thresholding techniques and quantitative performance evaluation," J. Electron. Imaging, vol. 13(1), pp. 146-165, Jan. 2004. https://doi.org/10.1117/1.1631315
  4. Y. Solihin and C. G. Leedham, "Integral ratio: a new class of global threshoding techniques for handwriting images," IEEE Trans. Pattern Anal. Mach. Intell., vol. 21(8), pp. 761-768, Aug. 1990.
  5. N. Sang, H. Li, W. Peng, and T. Zhang, "Knowledge-based adaptive thresholding segmentation of digital subtraction angiography images," Image Vision Comput., vol. 25(8), pp. 1263-1270, Aug. 2007. https://doi.org/10.1016/j.imavis.2006.07.026
  6. P. L. Rosin and E. Ioannidis, "Evaluation of global image thresholding for change detection," Pattern Recognition Lett., vol. 24(14), pp. 2345-2356, Oct. 2003. https://doi.org/10.1016/S0167-8655(03)00060-6
  7. Judith M. S. Prewitt and Mortimer L. Mendelsohn, "The analysis of cell images," Ann. N.Y. Acad. Sci., vol. 128, pp. 1035-1053, Jan. 1966.
  8. Weszka J.S. and Rosenfeld A., "Histogram modification for threshold selection," IEEE Trans. Systems Man Cybernet, vol. 9, pp. 38-52, Jan. 1979. https://doi.org/10.1109/TSMC.1979.4310072
  9. Rosenfeld A. and Torre P.D.L., "Histogram concavity analysis as an aid in threshold selection," IEEE Trans. Systems Man Cybernet, vol. 13, pp. 231-235, Apr. 1983..
  10. Wu V. and Manmatah R., "Document image clean-up and binarization," IN: Proc. SPIE'98 Document Recognition, vol. 5, pp. 263-273, Apr. 1998.
  11. Otsu N., "A threshold selection method from gray-level histograms," IEEE Trans. Systems Man Cybernet, vo. 9, pp. 62-66, Jan. 1979. https://doi.org/10.1109/TSMC.1979.4310076
  12. J.N. Kapur, P.K. Sahoo, and A. K. C. Wong, "A new method for grey-level picture thresholding using the entropy of the histogram," Comput. Graphics Vision Image Process, vol. 29, pp. 273-285, Mar. 1985. https://doi.org/10.1016/0734-189X(85)90125-2
  13. J. Kittler and J. Illingworth, "Minimum error thresholding," Pattern Recognition, vol. 19, pp. 41-47, 1986. https://doi.org/10.1016/0031-3203(86)90030-0
  14. C. H. Li and C. K. Lee, "Minimum cross entropy thresholding," Pattern Recognition, vol. 26, No. 4, pp. 617-625, Apr. 1993. https://doi.org/10.1016/0031-3203(93)90115-D
  15. Z. Hou, Q. Hu, and W. L. Nowinski, "On minimum variance thresholding," Pattern Recognition Lett., vol. 27, pp. 1732-1743, Oct. 2006. https://doi.org/10.1016/j.patrec.2006.04.012
  16. Zuoyong Li, Yong Cheng, Chuancai Liu, and Cairong Zhao, "Minimum standard deviation difference- based thresholding," Internation Conference on Measuring Technology and Mechatronics Automation, vol. 2, pp. 664-667, Changsha, China, Mar. 2010.
  17. Cheng H. D. and Chen Y., "Fuzzy partition of two-dimensional histogram and its application to thresholding," Pattern Recognition, vol. 32, pp. 825-843, May 1999. https://doi.org/10.1016/S0031-3203(98)00080-6
  18. Wang Q., Chi Z., and Zhao R., "Image thresholding by maximizing the index of non-fuzziness of the 2-D gray-scale histogram," Comput. Image and Vision Understanding, vol. 85, pp. 100-116, Feb. 2002. https://doi.org/10.1006/cviu.2001.0955
  19. Xu X., Xu S., Jin L., and Song E., "Characteristic analysis of Otsu threshold and its applications," Pattern Recognition Lett., vol. 32(7), pp. 956-961, May 2011. https://doi.org/10.1016/j.patrec.2011.01.021
  20. Lloyd N. Trefethen and David Bau, Numerical linear algebra, siam, pp. 25-97, 1997.
  21. 이철학, 김상운, "A Multi-thresholding Approach Improved with Otsu's Method", 대한전자공학회논문지-CI, vol. 43, no. 5, pp. 29-37, 2006. 9.