Improved video compression technology and the emerging high efficiency video coding standard

Abstract

Today, H.264/AVC is the state-of-the-art video coding standard. Especially after the 2004 development of its High Profile (HP), it has become one of the primary formats in high definition video content delivery. Recently, a joint Call for Proposals (CfP) on video compression technology has been issued by ISO/IEC MPEG and ITU-T VCEG, targeting at the next generation of video compression standards with substantially higher compression capability than H.264/AVC. As a response to this CfP, Fraunhofer HHI proposed a newly developed video coding scheme which achieves bit rate savings of around 30% when compared to H.264/AVC HP. This paper describes the proposed video coding scheme and discusses its innovative features

Review of Proposed High Efficiency Video Coding (HEVC) Standard

“…-Wide-range variable block-size prediction, with block sizes ranging from 64x64 down to 8x8 pixels and the size of prediction blocks can be adaptively chosen by using recursive quad-tree partitioning [8]. -The block sizes used for DCT based residual coding is adapted to the characteristics of the residual signal by using a nested, so-called residual quadtree (RQT) for partitioning of a given prediction block [9].…”

“…In HEVC each frame is divided into the basic processing unit scheme called Largest Coding Units (LCUs) [8]. LCUs can be recursively split into smaller Coding tree Units (CUs) using a generic quadtree segmentation structure (a nested quadtree structure) that indicates the subdivision of the CU for the purpose of prediction and residual coding.…”

“…However, they both come with a performance penalty since prediction dependencies are broken across boundaries and the statistics used in entropy coding have to be initialized for every slice/tile. To avoid these problems, Wavefront Parallel Processing (WPP) is supported in HEVC [8] [14]. Wavefront processing is a way to achieve parallel encoding and decoding without breaking prediction dependencies and using as much context as possible in entropy encoding.…”

“…Pixel values to be coded are predicted from already coded and reconstructed adjacent pixel values, and only small prediction errors are coded and the bit rate reduction depends on the correlation coefficient. In HEVC, for generating the prediction signal at subsample positions, one of the proposals [8] is used an interpolation method based on families of so-called maximal-order-minimal-support (MOMS) basis functions. This requires, however, an additional pre-filtering step on the reference picture before the actual interpolation.…”

“…This requires, however, an additional pre-filtering step on the reference picture before the actual interpolation. More details on the subsample interpolation scheme can be found in [8] [11].…”