Graphics Reference
In-Depth Information
distortion terms
SSE
luma
and
SSE
chroma
are set to zero when both input and PCM
coded samples have the same bit depth. Term
R
mode
includes all the bits required for
signaling PCM mode and PCM coded samples.
Finally, by applying this CU level mode decision at each level of CU recursion
tree a coding tree unit (CTU) level coding mode decision can be obtained.
The goal of Rate distortion optimized quantization (RDOQ), in quantization
process stage, is the adjustment of transform coefficient levels, in R-D sense [
17
].
For an insight to the general concept of minimization process, assume
c
k
to be the
last non-zero coefficient in a transformed block for a given position k; then, for
each transform coefficient level
l
i
, at position i
D
k
1, ::: 0, RDOQ tries to find
the optimal transform coefficient level,
l
i
that minimizes the cost function,
J
k
(
l
i
),
below:
J
k
.l
i
/
D
D
k
.l
i
/
C
R
k
.l
i
/
(9.14)
For computational simplicity, possible values of
l
i
are limited to be either zero, or
truncated
l
i
, or rounded-up
l
i
(i.e.,
l
floor
and
l
ceiling
). Distortion term
D
k
.
l
i
/ is due to
the quantization error and is calculated as normalized SSE in transform domain and
R
k
.
l
i
/ denotes number of bits used for transmitting level
l
i
. The optimal solution is
the vector of re-quantized transform levels at position
k
* with minimum
J
k
over all
possible positions, k.
9.3
Objective Performance Analysis
This section summarizes the comparison of coding efficiency of HEVC and AVC.
The test conditions are summarized in Table
9.4
and all encoders settings described
in Table
9.3
are used for the comparisons.
The results for the test sequences in Table
9.1
are summarized in Tables
9.5
,
9.6
,
9.7
and
9.8
. In case of Random Access Main, coding efficiency of HEVC is 42.7 %
higher than that of AVC. In case of All Intra Main however the improvement is
21.9 % which indicates that the improvement in Intra picture is lower than that in
predictive pictures (P or B picture).
As an example, R-D curves of the sequence, Four People (Class E, RA Main)
are shown in Figs.
9.6
,
9.7
,and
9.8
.
Tabl e 9. 4
Test conditions
Test conditions
Encoder
HM12.1 (HEVC) and JM18.5 (AVC)
Test sequences
All sequences defined in Table
9.1
Bit depth
8-bit (Main configuration)
Prediction structure
AI, RA, LDP and LDB
QPI
22, 27, 32 and 37
