DocumentCode :
1903960
Title :
A Fast Adaptive Motion Estimation Algorithm
Author :
Li, Ziyin ; Yang, Qi
Author_Institution :
Coll. of Opt. & Electron. Technol., China Jiliang Univ., Hangzhou, China
Volume :
3
fYear :
2012
fDate :
23-25 March 2012
Firstpage :
656
Lastpage :
660
Abstract :
UMHexagonS is a fast integer-pixel motion estimation algorithm adopted in H.264/AVC reference software, while in the real time applications, it still has many shortcomings, such as excessive search points and slow search speed. So, further improvement and optimization are needed for this algorithm. In this paper, a fast adaptive motion estimation algorithm is proposed based on the analysis of UMHexagonS algorithm. Many new technologies are applied in the novel algorithm, such as the dynamic search window, motion type adaptive search strategies, directional cross-shaped search, adaptive rectangle-diamond search, adaptive multi-levels octagon regional search and adaptive hexagon search. Experimental results show that the novel algorithm can reduce 10%-30% time of motion estimation while keeping the video quality almost unchanged. In addition, both the scene-adaptive ability and real-time performance are strongly strengthened.
Keywords :
motion estimation; optimisation; search problems; video coding; H.264-AVC reference software; UMHexagonS algorithm; adaptive hexagon search; adaptive multilevel octagon regional search; adaptive rectangle diamond search; directional cross-shaped search; dynamic search window; fast adaptive motion estimation this algorithm; integer pixel motion estimation algorithm; motion type adaptive search strategies; scene adaptive ability; video quality; Algorithm design and analysis; Encoding; Heuristic algorithms; Motion estimation; Prediction algorithms; Signal processing algorithms; Vectors; H.264; Video coding; adaptive search; motion estimation;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Science and Electronics Engineering (ICCSEE), 2012 International Conference on
Conference_Location :
Hangzhou
Print_ISBN :
978-1-4673-0689-8
Type :
conf
DOI :
10.1109/ICCSEE.2012.95
Filename :
6188259
Link To Document :
بازگشت