Title :
An Efficient and Fast Block Size Decision Exploiting Boundary Information on Inner Block for H.264/AVC Intra Prediction
Author :
Kim, Sungmin ; Lee, Takgi ; Chung, Kidong
Author_Institution :
Kidong Chung Dept. of Comput. Eng., Pusan Nat. Univ, Pusan
Abstract :
H.264/AVC, the newest video coding standard, offers an intra prediction with various block sizes and various prediction modes for luma component and chroma component. Due to this new feature, H.264/AVC offers a considerably higher improvement in coding efficiency compared to other compression standards such as MPEG-2, MPEG-4, H.263. However, the computational complexity of H.264/AVC is drastically increased because of the various modes. In this paper, we propose an efficient and fast algorithm to decide block size between 16*16 block and 4*4 block for the computational complexity reduction in the intra prediction of H.264/AVC. The proposed algorithm uses boundary difference values of sub- block, which has 8*8 size, in a macroblock for intra prediction. The boundary differences of inner center block represent the moving patterns of macroblock. Thus, if the value of the differences is smaller than threshold value, this macroblock is selected by 16*16 intra prediction mode. Otherwise, this is selected by 4*4 intra prediction mode. The boundary pixel values´ differences of the inner center 8*8 block in 16*16 macroblock are correlated to the smoothness of macroblock. The experimental results show that the proposed algorithm saved on the average 22.9% encoding time without any significant PSNR losses.
Keywords :
code standards; computational complexity; prediction theory; video coding; H.264-AVC intra prediction; block size decision; chroma component; computational complexity reduction; inner block boundary information; luma component; macroblock; video coding standard; Automatic voltage control;
Conference_Titel :
Image Analysis and Interpretation, 2008. SSIAI 2008. IEEE Southwest Symposium on
Conference_Location :
Santa Fe, NM
Print_ISBN :
978-1-4244-2296-8
Electronic_ISBN :
978-1-4244-2297-5
DOI :
10.1109/SSIAI.2008.4512295