Spatio-temporal Mode Selection Methods of Fast H.264 Using Multiple Reference Frames 


Vol. 33,  No. 3, pp. 247-254, Mar.  2008


PDF
  Abstract

H.264 provides a good coding efficiency compared with existing video coding standards, H.263, MPEG-4, based on the use of multiple reference frame for variable block size motion estimation, quarter-pixel motion estimation and compensation, 4x4 integer DCT, rate-distortion optimization, and etc. However, many modules used to increase its performance also require H.264 to have increased complexity so that fast algorithms are to be implemented as practical approach. In this paper, among many approaches, fast mode decision algorithm by skipping variable block size motion estimation and spatial-predictive coding, which occupies most encoder complexity, is proposed. This approach takes advantages of temporal and spatial properties of fast mode selection techniques. Experimental results demonstrate that the proposed approach can save encoding time up to 65% compared with the H.264 standard while maintaining the visual perspectives.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

J. Kwon, M. Kang, C. Ryu, "Spatio-temporal Mode Selection Methods of Fast H.264 Using Multiple Reference Frames," The Journal of Korean Institute of Communications and Information Sciences, vol. 33, no. 3, pp. 247-254, 2008. DOI: .

[ACM Style]

Jae-Hyun Kwon, Min-Jung Kang, and Chul Ryu. 2008. Spatio-temporal Mode Selection Methods of Fast H.264 Using Multiple Reference Frames. The Journal of Korean Institute of Communications and Information Sciences, 33, 3, (2008), 247-254. DOI: .

[KICS Style]

Jae-Hyun Kwon, Min-Jung Kang, Chul Ryu, "Spatio-temporal Mode Selection Methods of Fast H.264 Using Multiple Reference Frames," The Journal of Korean Institute of Communications and Information Sciences, vol. 33, no. 3, pp. 247-254, 3. 2008.