Information Technology Reference
In-Depth Information
1)
Binarisation
2)
Context Modeling
3)
Adaptive arithmetic coding
Fig. 18
Entropy Coding Block Diagram
The purpose of the first stage is to provide a bit stream with easily analyzable
statistics that can be encoded using arithmetic coding, which can adapt to those
statistics, thus reflecting any local statistical features.
5.4.1 Binarization
Binarization is the process of transforming the multi-valued coefficient symbols
into bits. The resulting bit stream can then be arithmetically coded. Dirac uses an
interleaved exp-Golomb binarisation.
In this scheme, add 1 to your value. Adding 1 ensures there will be be a leading
1. In binary form, the value will be a 1 followed by K other bits.
11
……
These K bits ("info bits") are interleaved with K zeroes ("follow bits") each of
which mean "another bit coming", followed by a terminating 1.
0
0……0
1
Conventional exp-Golomb coding has the K zeroes at the beginning, followed by
the 1 i.e 00...01bk-1bk-2 .. b0, but interleaving allows the decoder to run a single
loop, two bits at a time, rather than in two loops.
5.4.2 Context Modeling
Context Modelling estimates the probabilites of symbols as they occur. It calculates
the probabilities for 0 and 1 to be given to Arithmetic coding module. The context
modelling in Dirac is based on the principle that whether a coefficient is small (or zero,
in particular) or not is well-predicted by its neighbours and its parents. Therefore the
codec conditions the probabilities used by the arithmetic coder for coding the first
follow bit on whether the neighbouring coefficients or the parent coefficient are
zero.The reason for this approach is that, whereas the wavelet transform largely
Search WWH ::
Custom Search