Title :
Fast Inter Mode Selection Algorithm for Motion Estimation in MPEG-4 AVC/JVT/H.264
Author :
Pei-Jun Lee ; Ming-Long Lin
Author_Institution :
Dept. of Electr. Eng., Nat. Chi-Nan Univ., Taiwan
Abstract :
The newest video coding standard H.264/AVC utilizes flexible block sizes and multiple reference frames to do motion estimation. Totally, there are seven possible types of block sizes to be selected in motion estimation. The exhaustive cost calculation for possible modes is required to find the motion vector. In order to reduce the computational complexity, this paper proposes a fast algorithm to select the block mode. The algorithm uses fuzzy inference with motion vector tightness and the variance of image intensity to determine whether it is necessary to do motion estimation with smaller size block mode. From the simulation result, the proposed fast algorithm can save at least 30% encoding time without sacrificing the picture quality and the bite rate efficiency.
Keywords :
data compression; fuzzy logic; motion estimation; video coding; AVC-JVT-H.264; MPEG-4; advanced video coding standard; fuzzy inference; mode selection algorithm; motion estimation; motion vector; Automatic voltage control; Computational complexity; Computational modeling; Costs; Encoding; Inference algorithms; Lagrangian functions; MPEG 4 Standard; Motion estimation; Video coding; Fuzzy Logic; H.264/AVC; MPEG-4; Mode decision; Motion Estimation;
Conference_Titel :
Image Processing, 2006 IEEE International Conference on
Conference_Location :
Atlanta, GA
Print_ISBN :
1-4244-0480-0
DOI :
10.1109/ICIP.2006.312588