Ultra low-complexity fast variable block size motion estimation algorithm in H.264/AVC

Yang Song, Zhenyu Liu, Takeshi Ikenaga, Satoshi Goto

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

Variable block size motion estimation (VBSME) is the most computation consuming part in the newest H.264/AVC video coding standard. To reduce computation, an ultra low-complexity fast VBSME algorithm is proposed in this paper. Different to previous fast algorithms which separately pressed the 7 block modes, in the presented algorithm, all the block modes are simultaneously calculated for most of the search points, and thus the SADs can be reused by all blocks. Moreover, a low-pass filter based pixel decimation is also introduced to decrease the matching cost. Experiments show that the proposed fast VBSME algorithm can reduce the computation to about 0.2% of fast full search (FFS) with robust image quality.

Original languageEnglish
Title of host publicationProceedings of the 2007 IEEE International Conference on Multimedia and Expo, ICME 2007
Pages376-379
Number of pages4
Publication statusPublished - 2007
EventIEEE International Conference onMultimedia and Expo, ICME 2007 - Beijing
Duration: 2007 Jul 22007 Jul 5

Other

OtherIEEE International Conference onMultimedia and Expo, ICME 2007
CityBeijing
Period07/7/207/7/5

Fingerprint

Motion estimation
Low pass filters
Image coding
Image quality
Pixels
Costs
Experiments

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Software

Cite this

Song, Y., Liu, Z., Ikenaga, T., & Goto, S. (2007). Ultra low-complexity fast variable block size motion estimation algorithm in H.264/AVC. In Proceedings of the 2007 IEEE International Conference on Multimedia and Expo, ICME 2007 (pp. 376-379). [4284665]

Ultra low-complexity fast variable block size motion estimation algorithm in H.264/AVC. / Song, Yang; Liu, Zhenyu; Ikenaga, Takeshi; Goto, Satoshi.

Proceedings of the 2007 IEEE International Conference on Multimedia and Expo, ICME 2007. 2007. p. 376-379 4284665.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Song, Y, Liu, Z, Ikenaga, T & Goto, S 2007, Ultra low-complexity fast variable block size motion estimation algorithm in H.264/AVC. in Proceedings of the 2007 IEEE International Conference on Multimedia and Expo, ICME 2007., 4284665, pp. 376-379, IEEE International Conference onMultimedia and Expo, ICME 2007, Beijing, 07/7/2.
Song Y, Liu Z, Ikenaga T, Goto S. Ultra low-complexity fast variable block size motion estimation algorithm in H.264/AVC. In Proceedings of the 2007 IEEE International Conference on Multimedia and Expo, ICME 2007. 2007. p. 376-379. 4284665
Song, Yang ; Liu, Zhenyu ; Ikenaga, Takeshi ; Goto, Satoshi. / Ultra low-complexity fast variable block size motion estimation algorithm in H.264/AVC. Proceedings of the 2007 IEEE International Conference on Multimedia and Expo, ICME 2007. 2007. pp. 376-379
@inproceedings{5fd4875e18b345bc9d1f85a6d6debda1,
title = "Ultra low-complexity fast variable block size motion estimation algorithm in H.264/AVC",
abstract = "Variable block size motion estimation (VBSME) is the most computation consuming part in the newest H.264/AVC video coding standard. To reduce computation, an ultra low-complexity fast VBSME algorithm is proposed in this paper. Different to previous fast algorithms which separately pressed the 7 block modes, in the presented algorithm, all the block modes are simultaneously calculated for most of the search points, and thus the SADs can be reused by all blocks. Moreover, a low-pass filter based pixel decimation is also introduced to decrease the matching cost. Experiments show that the proposed fast VBSME algorithm can reduce the computation to about 0.2{\%} of fast full search (FFS) with robust image quality.",
author = "Yang Song and Zhenyu Liu and Takeshi Ikenaga and Satoshi Goto",
year = "2007",
language = "English",
isbn = "1424410177",
pages = "376--379",
booktitle = "Proceedings of the 2007 IEEE International Conference on Multimedia and Expo, ICME 2007",

}

TY - GEN

T1 - Ultra low-complexity fast variable block size motion estimation algorithm in H.264/AVC

AU - Song, Yang

AU - Liu, Zhenyu

AU - Ikenaga, Takeshi

AU - Goto, Satoshi

PY - 2007

Y1 - 2007

N2 - Variable block size motion estimation (VBSME) is the most computation consuming part in the newest H.264/AVC video coding standard. To reduce computation, an ultra low-complexity fast VBSME algorithm is proposed in this paper. Different to previous fast algorithms which separately pressed the 7 block modes, in the presented algorithm, all the block modes are simultaneously calculated for most of the search points, and thus the SADs can be reused by all blocks. Moreover, a low-pass filter based pixel decimation is also introduced to decrease the matching cost. Experiments show that the proposed fast VBSME algorithm can reduce the computation to about 0.2% of fast full search (FFS) with robust image quality.

AB - Variable block size motion estimation (VBSME) is the most computation consuming part in the newest H.264/AVC video coding standard. To reduce computation, an ultra low-complexity fast VBSME algorithm is proposed in this paper. Different to previous fast algorithms which separately pressed the 7 block modes, in the presented algorithm, all the block modes are simultaneously calculated for most of the search points, and thus the SADs can be reused by all blocks. Moreover, a low-pass filter based pixel decimation is also introduced to decrease the matching cost. Experiments show that the proposed fast VBSME algorithm can reduce the computation to about 0.2% of fast full search (FFS) with robust image quality.

UR - http://www.scopus.com/inward/record.url?scp=46449102160&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=46449102160&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:46449102160

SN - 1424410177

SN - 9781424410170

SP - 376

EP - 379

BT - Proceedings of the 2007 IEEE International Conference on Multimedia and Expo, ICME 2007

ER -