DOI QR코드

DOI QR Code

Image Segmentation Using Anisotropic Diffusion Based on Diagonal Pixels

대각선 방향 픽셀에 기반한 이방성 확산을 이용한 영상 분할

  • Published : 2007.02.28

Abstract

Anisotropic diffusion is one of the widely used techniques in the field of image segmentation. In the conventional anisotropic diffusion [1]-[6], usually 4-neighborhood directions are used: north, south, west and east, except the image diagonal directions, which results in the loss of image details and causes false contours. Existing methods for image segmentation using conventional anisotroplc diffusion can't preserve contour information, or noises with high gradients become more salient as the umber of times of the diffusion increases, resulting in over-segmentation when applied to watershed. In this paper, to overcome the shortcoming of the conventional anisotropic diffusion method, a new arusotropic diffusion method based on diagonal edges is proposed. And a method of watershed segmentation is applied to the proposed method. Experimental results show that the process time of the proposed method including diagonal edges over conventional methods can be up to 2 times faster and the Circle image quality improvement can be better up to $0.45{\sim}2.33(dB)$. Experiments also show that images are segmented very effectively without over segmentation.

이방성 확산은 영상 분할 분야에서 광범위하게 사용되는 방식이다. 기존의 전통적인 이방성 확산 [1]-[6]에서는 이미지의 대각선 방향을 고려하지 않고 4 방향(동, 서, 남, 북)을 주로 이용하였다. 전통적인 이방성 확산(Diffusion)을 이용한 영상 분할은 확산이 반복될수록 윤곽선 정보를 적절히 유지 못하거나 잡음을 제거하지 못함으로써 웨터쉐드(Watershed) 알고리즘을 적용하는 경우 과다 분할을 피할 수 없다는 단점을 갖고 있다. 본 논문에서는 전통적인 이방성 확산의 이러한 단점을 보완하기 위하여 대각선 방향에 기반한 새로운 이방성 확산을 제안하고, 워터쉐드를 이용한 영상 분할 방법을 적용하였다. 실험 결과 본 논문에서 제안한 대각선 방향을 포함한 이방성 확산을 적용할 경우 기존의 방법과 비교하여 약 2배의 속도 향상을 가져왔으며, Circle 이미지의 경우 약 $0.45{\sim}2.33(dB)$정도 성능 향상된 화질을 보였다. 또한 기존의 방법보다 과다 분할이 줄어들고 영상이 매우 효과적으로 분할됨을 확인하였다.

Keywords

References

  1. A P. Witkin,"Scale-Space Filtering," Proc. 8th Int. Joint Conf. Art. Intelligence, Vol.2, pp.1019-1022,1983.
  2. P. Perona and J. Malik,"Scale- Space and Edge Deκrlion Using Anisotropic Diffusion," IEEE.Trans. Patt. Anal. and :Machine Intell., VoL12,No.7, pp.629-639, 1990. https://doi.org/10.1109/34.56205
  3. M J. Black, G. Sapiro, D. H Marimont, and D. Hegger, "Robust Anisotropic Diffusion," IEEE Trans. hnage Processing, Vol.7, No.3, pp.421-432, :Mar. 1998. https://doi.org/10.1109/83.661192
  4. F. Voci, S. Eiho, N. Sugirnoto, and H Sekiguchi,''Estirnating the Gradient Threshold in the Perona-Mailk Equation," IEEE Signal Processing Magazine, pp.39-46, May 2004.
  5. J. Canny, "A Computational Approach to Edge Detection," IEEE Trans. Pattern Anal. Machine Intell,Vol.PAMI-8, pp.679-698, 1986. https://doi.org/10.1109/TPAMI.1986.4767851
  6. H Y. Kim, "An Anisotropic Diffusion Meaningful Scale Pmαneter," Personal Comrnunication. 2005.
  7. R Lu, 'Novel Anisotropic Diffusion Algorithrn Based on PID Control Law Together with Stopping Mechanisrn," Proceedings of the 2005 IEEE Engineering in Medicine and Biology 27th Annual Conference Shanghai, China, pp.1-4, Sept. 2005.
  8. M. Sonka, V. Hlavac, and R Boyle, Image processing,analysis, and machine vision, 2nd ed. PWS publishing, 1998
  9. H. K Hahn and H. O. Peitgen, "IWT Interactive Watershed Transform A hierarchical method for efficient interactive and automated segmentation of multidimensional grayscale image," Proceeding of SPIE, Vol. 5032,pp.643-653, Feb. 2003.

Cited by

  1. Image Edge Detector Based on Analog Correlator and Neighbor Pixels vol.13, pp.10, 2013, https://doi.org/10.5392/JKCA.2013.13.10.054