DOI QR코드

DOI QR Code

고속 블록 정합 움직임 추정 기법 기반의 향상된 십자 다이아몬드 탐색

ENHANCED CROSS-DIAMOND SEARCH BASED FAST BLOCK MATCHING NOTION ESTIMATION ALGORITHM

  • 김정준 (한양대학교 전자통신컴퓨터공학과) ;
  • 전광길 (한양대학교 전자통신컴퓨터공학과) ;
  • 정제창 (한양대학교 전자통신컴퓨터공학과)
  • Kim, Jung-Jun (Department of Electronics and Computer Engineering, Hanyang University) ;
  • Jeon, Gwang-Gil (Department of Electronics and Computer Engineering, Hanyang University) ;
  • Jeong, Je-Chang (Department of Electronics and Computer Engineering, Hanyang University)
  • 발행 : 2007.09.29

초록

본 논문에서는 새로운 고속 움직임 추정 알고리듬을 제공한다. 이 알고리듬의 이름은 향상된 십자 다이아몬드 탐색이며, 이것은 다이아몬드 탐색을 기초로 한다. 비록 다이아몬드 알고리듬이 가장 널리 알려진 고속 탐색 알고리듬이지만 몇몇의 시퀀스에 대해서 객관적이고 주관적인 화질이 떨어지고 불필요한 후보 블록에 대한 블록 정합을 수행함을 알 수 있다. 그래서 고속 움직임 추정을 하기 위해서 우리가 제안한 향상된 십자 다이아몬드 알고리듬에서는 첫 번째 단계에서 작은 십자형 탐색을 하고 그 다음 단계에서는 다이아몬드 탐색을 사용한다. 실험결과 향상된 십자 다이아몬드 탐색은 많이 알려진 다른 고속 움직임 추정 알고리듬보다 화질을 향상시킬 뿐만 아니라 탐색 속도 또한 향상됨을 알 수 있다.

A new fast motion estimation algorithm is presented in this paper. The algorithm, named Enhanced Cross-Diamond Search (ECDS), is based on the Diamond Search (DS) algorithm. The DS algorithm, even though faster than the most well-known algorithms, was found not to be very robust in terms of objective and subjective qualities for several sequences and the algorithm searches unnecessary candidate blocks. We propose a novel ECDS algorithm using a small cross search as the initial step, and large/small DS patterns as subsequent steps for fast block motion estimation. Experimental results show that the ECDS is much more robust, provides a faster searching speed, and smaller distortions than other popular fast block-matching algorithms.

키워드

참고문헌

  1. J. D. Robbins and A. N. Netravali, 'Recursive motion compensation: a review,' Image Sequence Processing And Dynamic Sequence Analysis, pp76-103, Springer Verlag, 1983
  2. J. R. Jain and A. K. Jain, 'Displacement measurement and its application in inter frame image coding,' IEEE Trans. Commun., vol. COM-29, pp. 1799-1808, Dec. 1984 https://doi.org/10.1109/TCOM.1981.1094950
  3. K. P. Hom and B. G. Schunck, 'Determining Optical flow,' Artificial Intelligence, vol. 17, pp. 185-203, 1981 https://doi.org/10.1016/0004-3702(81)90024-2
  4. G. Sorwar, M. Mnshed, and L. Dooley, 'Block-based true motion estimation nsing distance dependent thresholded search,' in Proc. ISCA Compo. Appl In Indns. And Eng., pp. 45-48, 2001
  5. T. Koga, K. Iinnma, A. Hirano, Y. Iijirna, and T. Ishignro, 'Motion compensated interframe coding for video conferencing,' in Proc, Nat. Telecommon. Conf, New Orleans, LA. Nov.-Dec. 1981, pp. G5.3.1-G5.3.5
  6. R. Li, B. Zeng, and M. L. Lion, 'A new three-step search algorithm for block motion estimation,' IEEE Trans. Circnits Syst. Video Technology, vol 4, pp. 438-443, Ang. 1994 https://doi.org/10.1109/76.313138
  7. L. M. Po and W. C. Ma, 'A novel four-step search algorithm for fast block motion estimation,' IEEE Trans. Circnit Syst. Video Technology, vol 6, pp.313-317, Jnne 1996 https://doi.org/10.1109/76.499840
  8. C. H. Chenng, and L. M. Po, 'A novel cross-diamond search algorithm for fast block motion estimation,' IEEE Trans. Circnits Syst. Video Technology, vol 12, no. 12, Dec 2002
  9. J. B. Xn, L. M. Po, and C. K. Chenng, 'A new prediction model search algorithm for fast block motion estimation,' in Proc. ICIP, 1997, vol3, pp.610-613
  10. B. Lin and Zaccarin, 'A new fast algorithms for the estimation of block motion vectors,' IEEE Trans. CSVT, vol.3, no.2, pp. 148-157, April 1993
  11. C. W. Lam, L. M. Po, and C. H. Chenng, 'A new cross-diamond search algorithm for fast block matching motion estimation,' in Proc. Nenral Network and Signal Processing 2003, Dec. 2003
  12. J. Y. Tham, S. Ranganath, M. Ranganath, and A. A Kassim, 'A novel nnrestricted center-biased diamond search algorithm for block motion estimation,' IEEE Trans. Circnits Syst. Video Technology, vol. 8, no. 4, Ang. 1998
  13. C. Zhn, X. Lin, and L. P. Chan, 'Hexagon-based search pattern for fast block motion estimation,' IEEE Trans. Circnits Syst. Video Technology, vol. 12, no. 5, pp. 349-355, May 2002 https://doi.org/10.1109/TCSVT.2002.1003474
  14. L. K. Lin and E. Feig, 'A block-based gradient descent search algorithm for block motion estimation in video coding,' IEEE Trans. Circnits Syst. Video Technology, vol. 6, pp. 313-317, June 1996 https://doi.org/10.1109/76.499840