> (' \pJava Excel API v2.6 Ba==h\:#8X@"1Arial1Arial1Arial1Arial + ) , * `e)DC, title[*]contributor[author]contributor[advisor]keywords[*]date[issued] publisher citationsidentifier[uri]identifier[doi]abstractrelation[journal]relation[volume]relation[no]relation[page](-\ \T 0\ H.264/AVC 8T0X ĳ ҜT)X Qǩ;
Complexity Reduction in an H.264/AVC Encoder based on Rate Distortion Optimization and its Application to Transcoding;t=2010-08\ՑYPxhttps://repository.hanyang.ac.kr/handle/20.500.11754/141058;
http://hanyang.dcollection.net/common/orgView/200000415292;QH.264/AVC ISO/IECX Moving Picture Experts Group (MPEG) ITU-T Video Coding Experts Group (VCEG)X ٳ <\ \X D$ 8T| \ m \t. H.264/AVC 0tX D$ 8T \ @ 8T (D и. tǃ@ xǸ| !, 䲑\ l0X ] 0 Ʉ 0 , 8p D tǩ\ Ʉ , 0 T 0, ]D \ D0, 0 8t 8@ @ \ 0X ȩ 0x t.
t\ 䲑\ 0X \X 8T piD >0 t H.264/AVC 8T0 (-\ \T 0 D \. (-\ \T \X 8T ܴX pi@ @ (-\ D ɔ ܴX pi<\ . X̹ (-\ @ 8T Ĭ <\ 8T0 ι@ ŰɷD DՔ\ \. t\ lp@ 8T0X l iX J. 0| 8T (@ Xt (-\ XՔ ŰɷD ¤0 \ ) \ l DՔX. (-\ XՔ 8T0X ĳ| ¤0 t |8 8pX ) ҜTTX x0 ܴ X ĳ ) \ l D .
8p D tǩ\ Ʉ t 8T 췥X t Ʉ . D] 8p D \ Ʉ )t \ X ̹D XՔ ) @ Tt 8T 1D д̹ l ĳ Ʉ X 0| \. t\ ĳ 8| tհX0 t H 8p )@ Ʉ 8p ֬X l\ ]X ĳX 1D tǩ\. t\ Ĭ| tǩX H )@ 4` ̹\ 8T (X @ @ 8T | .
D$X 1D t0 t ҜT)@ ι@ QǩƄ|| . ҈ 0tX D$ 8T \<\0 @ 8T (D H.264/AVC\X ҜT)X ɔ1t ֬ p . t\ ҜT) 8T X\ (-\ \T ) 0\ H.264/AVC 8T0X ĳ ҜTT l iX J. 0|, |8 ҜTT 8<\0X | tǩ\ x0 ܴ )t H. H\ )@ % DX Ʉ 0 Xt 8p ĳ @ Q ) 0\. t\ D t 8T (X t ҜTTX ĳ| | .; H.264/AVC is the latest video coding standard which was developed by the joint video team of ISO/IEC Motion Picture Experts Group (MPEG) and ITU-T Video Coding Experts Group (VCEG). It has significantly better coding performance than other coding standards due to the adoption of several new techniques including directional prediction of intra-coded blocks, variable block size motion estimation, multi-reference frame motion compensation, integer transform, an in-loop de-blocking filter, and context adaptive variable length code.
To find the most efficient coding mode combination, the H.264/AVC encoder generally uses a rate distortion optimization (RDO) technique. In the RDO process, the best coding mode combination is one that produces the least rate distortion (RD) cost. However, since the RD cost is calculated only after a sequence of encoding operations, it results in extremely high computational complexity in the encoder. This process is not suitable for real-time encoding. Therefore, the algorithms for reducing the complexity of H.264/AVC caused by the RDO process are required with maintaining the coding efficiency. To reduce the complexity of the H.264/AVC encoder caused by RDO, this dissertation focuses on the multiple reference frame selection problem and inter-mode decision problem in transcoding to H.264/AVC.
When using the multiple reference frame motion compensation (MRF-MC) method, the best picture in several pre-coded frames is used to perform motion estimation. Although MRF-MC supports better inter-prediction performance than those of methods that use just one previous frame, the required computational complexity increases with the number of searched frames. To solve this problem, the proposed multiple reference frame selection algorithm is based on the correlation between the temporal complexity of the reference region for a motion estimation process and the current macro block. The proposed algorithm uses this property to achieve a noticeable coding time reduction with a negligible drop in coding efficiency.
Transcoding for universal multimedia access is used in many applications. In particular, the transc<oding to H.264/AVC, which is improved in terms of coding efficiency, is required. In the transcoding process, the complexity of the H.264/AVC encoder part with RDO is not suitable for real-time implementation. To alleviate the complexity, information from the decoding part in transcoder is used. The proposed algorithm is based on reference region complexity information, which is indicated by the motion vector in the input bitstream, and the adaptive threshold selection method. By using this algorithm, the transcoder complexity is alleviated without significant loss of the RD performance.&HQsj&)KLng#0#
!*+
dMbP?_*+%" ,,??U
>@
!"#$%&Root EntryWorkbook.SummaryInformation(DocumentSummaryInformation8