Graphics Reference
In-Depth Information
Chapter
6
by Budagavi, Fuldseth, and Bjøntegaard describes the transform
and quantization related aspects of HEVC. Ultimately, no matter how effective a
prediction scheme is applied, there is generally a remaining unpredictable signal
that needs to be represented, and HEVC has greater flexibility and adaptivity in
its transform and quantization design than ever before, and it also includes some
additional coding modes in which the transform stage and sometimes also the
quantization stage are skipped altogether.
Chapter
7
by Norkin et al. discusses the in-loop filtering in HEVC, which
includes processing elements not found in older video coding designs. As with its
AVC predecessor, HEVC contains an in-loop deblocking filter—which has been
simplified and made more parallel-friendly for HEVC. Moreover, HEVC introduces
a new filtering stage called the sample-adaptive offset (SAO) filter, which can
provide both an objective and subjective improvement in video quality.
Chapter
8
by Sze and Marpe covers the entropy coding design in HEVC, through
which all of the decisions are communicated as efficiently as possible. HEVC
builds on the prior concepts of context-based arithmetic coding (CABAC) for this
purpose—pushing ever closer to the inherent entropy limit of efficiency while
minimizing the necessary processing requirements, enabling the use of parallel
processing, and limiting worst-case behavior.
Chapter
9
by Suzuki et al. covers the compression performance of the
design—investigating this crucial capability in multiple ways for various example
applications—and including both objective and subjective performance testing. It
shows the major advance of HEVC relative to its predecessors. It also shows that
the compression improvement cuts across a very broad range of applications, rather
than having only narrow benefits for particular uses.
Chapter
10
by Tikekar et al. describes hardware architecture design for HEVC
decoding. Decoders are likely to vastly outnumber encoders, and minimizing their
cost and power consumption is crucial to widespread use.
Chapter
11
by Tsai, Tsai, and Chen describes hardware architecture design
for HEVC encoding. While the requirements for making a decoder are relatively
clear—i.e., to properly decode the video according to the semantics of the syntax
of the standard—encoders present the open-ended challenge of determining how to
search the vast range of possible indications that may be carried by the syntax and
select the decisions that will enable good compression performance while keeping
within the limits of practical implementation.
We are proud to provide the community with this timely and valuable information
collected together into one volume, and we hope it will help spread an understanding
of the HEVC standard and of video coding design in general. We expect this topic
to facilitate the development and widespread deployment of HEVC products and of
video-enabled devices and services in general.
