Technical Field
[0001] The present invention relates to a video coding device and a video decoding
device using an adaptive color transform in residual domain and a chroma (color difference)
quantization offset.
Background Art
10 [0002] In a video coding system based on high efficiency video coding (HEVC)/H.265,
each frame of a digitized image is divided into coding tree units (CTUs) and the respective CTUs
are coded in the order of raster scan. The CTU is divided into coding units (CUs) in a quad-tree
structure and then coded. Each CU is divided into prediction units (PUs) and predicted.
Moreover, a prediction error of each CU is divided into transform units (TUs) in a quad-tree
15 structure and then frequency-transformed.
[0003] The CU is a coding unit for an intra prediction or inter-frame prediction.
[0004] The intra prediction (intra-frame prediction) is prediction for generating a
prediction signal from a reconstmcted image of a frame to be encoded. In HEVC/H.265, 33
types of angular intra prediction and the like are defined. In angular intra prediction, a
20 reconstmcted pixel around a block to be encoded is extrapolated in any of 33 directions depicted
in Fig. 14, to generate an intra prediction signal.
[0005] In addition to angular intra prediction, DC prediction and planar prediction are
specified as intra prediction. In DC prediction, a mean value of a reference image is used as the
prediction values of all pixels in a TU to be predicted. In planar prediction, a prediction image
25 is generated by linear interpolation fi·om pixels in a reference image.
[0006] Inter-frame prediction is a prediction based on an image of a reconstructed
fi·ame (reference picture) different in display time from a frame to be encoded. Inter-frame
prediction is also referred to as inter prediction. In inter prediction, an inter prediction signal is
generated based on a reconstructed image block of a reference picture (using pixel interpolation
30 if necessary).
[0007] A digital color image is composed of RGB digital images. A digital color
image is generally transformed to signals in a color space other than the RGB space to increase
the compression efficiency (to decrease the data amount) when the color image is transmitted
through a transmission line. For example, the image signal is transformed to a signal in a color
2
space (YCoCr space) where a combination of a luma signal (Y) and chroma signals (Cb, Cr) are
composed.
[0008] A quantization parameter (QP) for the chroma signal is generated by
transforming the QP for the luma signal by using an offset value "chroma_ qp index_ offset." In
------ --- --- -- - -- --------------
5 HEVC, cb_qp_index_offset (a first chroma quantization offset) is applied to Cb, and
cr_qp_index_offset (a second chroma quantization offset) is applied to Cr.
[0009] In RExt (range extension) ofHEVC, the standardization of an extension
function was carried out (see Non Patent Literature 1).
[0010] As a method for further increasing the compression efficiency of the extension
10 function using RExt, there has been proposed a technique called "adaptive color transform in
residual domain" in Non Patent Literature 2. As illustrated in Fig. 17, the adaptive color
transform in residual domain is a technique of switching a prediction error signal of an image
signal of the RGB space to a signal of the YCoCr space adaptively in block units (per block).
[0011] Specifically, it is possible to select, in block units (per block), whether to
15 compress the prediction error signal of the RGB space directly or to transform the prediction
error signal to a signal of the YCoCr space by using a forward color space transform matrix (see
equation (I)) described below before compression. Moreover, Fig. 17 illustrates an example
that data is compressed in the YCoCr space with respect to the shaded blocks and data is
compressed in the RGB space with respect to other blocks.
20 [0012] Information on the color space used for the data compression of blocks is
signaled by a cu_residual_csc_flag syntax. Equation "cu_residual_csc_flag = 0" represents that
the signal of the RGB space is compressed, while "cu_residual_csc_flag =I" represents that the
signal is compressed after it is transformed to the YCoCr space.
[0013] If cu_residual_csc_flag = I, a receiver (a video decoding device) restores the
25 signal of the YCoCr space to the signal of the RGB space by using a backward color space
transform matrix described below and then performs decoding processing.
[0014]
[Math. I]
3
[£~] [_:
2 Forward - 0 :J [~] /4
- - - ------- ----------------- 3
( 1)
[~] [: 1
Backward: 0 ::J [£:] /4
- 1
[00 15] Moreover, the norm is not constant in the aforementioned color transform
matrix and therefore, when cu_residual_csc_flag = 1, a different chroma quantization offset is
added to the quantization parameter for each YCoCr component in quantization processing and
5 inverse quantization processing of a block prediction error signal.
[00 16] Moreover, Patent Literature 1 describes a video coding device and a video
decoding device which perform signal processing which is different according to whether an
input image signal is a signal of the RGB space or a signal of the YCoCr space. Specifically,
when performing a weighted prediction based on the H.264/AVC, the video coding device
10 applies the same offset to the R, G, and B signals and to theluma signal (Y signal) and applies
different offsets to the clu·oma signals, regarding an offset added to a prediction signal.
However, Patent Literature 1 does not teach any new knowledge regarding the chroma
quantization offset.
15 Citation List
Patent Literature
[0017] Patent Literature 1: Japanese Patent Application Laid-Open No. 2011-151683
Non Patent Literatures
[00 18] Non Patent Literature 1: D. Flynn, et al.,"High Efficiency Video Coding
20 (HEVC) Range Extensions text specification: Draft 7", JCTVC-Q1005, Joint Collaborative Team
on Video Coding (JCT-VC) ofiTU-T SG 16 WP 3 and ISO/IEC JTC 1 I SC 29/ WG 11 17th
Meeting: Valencia, ES, 27 March- 4 April2014
Non Patent Literature 2: L. Zhang et al., "SCCE5 Test 3.2.1: In-loop color-space
transform", JCTVC-R0147, Joint Collaborative Team on Video Coding (JCT-VC) of ITU
25 -T SG 16 WP 3 and ISO/IEC JTC 1/ SC 29/ WG 11 18th Meeting: Sapporo, JP, 30 Ju
ne - 9 July 2014
"
Summary of Invention
Technical Problem
4
[0019] Next, referring to Fig. 18, the following describes the configuration and
operation of a general video coding device which outputs a bitstream with each CU of each
5 -fratneofa-dtgitlzed-inurgeas-an-Jnputimage. ----
[0020] A video coding device illustrated in Fig. 18 includes a switch 101, a color space
transformer 102, a switch 103, a frequency transformer/quantizer 104, an inverse
quantizer/inverse fi·equency transformer 105, a switch 106, an inverse color space transformer
107, a switch 108, a buffer 109, a predictor 110, a prediction parameter determiner 111, an
10 entropy encoder 112, a subtractor 115, and an adder 116.
15
[0021] The predictor 110 generates a prediction signal for a CU input image signal.
Specifically, the predictor 110 generates a prediction signal (an intra-prediction signal) on the
basis of intra prediction and generates a prediction signal (an inter-prediction signal) on the basis
of inter-prediction.
[0022] A prediction image supplied from the predictor 110 is subtracted fi·om an image
input to the video coding device at the subtractor 115 and then input to the switch 101 as a
prediction error image. In the example illustrated in Fig. 18, an input image signal is a signal of
the RGB space. Moreover, the video coding device has a function of an adaptive color
transform in residual domain. For example, the video coding device is able to adaptively
20 switch a prediction error signal of an image signal of the RGB space to a signal of the YCoCr
space in block units (per block).
[0023] When a prediction error signal of the RGB space is used, the switch 101 is set
so that the prediction error image is input to the switch 1 03. When the prediction error signal of
the YCoCr space is used, the switch 101 is set so that the prediction error image is input to the
25 color space transformer 102. Moreovet~ the switch 101 sets an output destination of a
prediction error image according to the control of the prediction parameter determiner 111, for
example.
[0024] The color space transformer 102 transforms the prediction error signal of the
RGB space to a signal of the YCoCr space by using the aforementioned equation (1) (a forward
30 color space transform matrix) and then outputs the signal to the switch 103.
[0025] When a prediction error signal of the RGB space is used, the switch 1 03 outputs
a prediction enor signal, which has been received fi·om the switch 1 01, to the fi·equency
transformer/quantizer 104. When a prediction error signal of the YCoCr space is used, the
switch I 03 outputs a prediction error signal, which has been received from the color space
35 transformer 102, to the frequency transformer/quantizer 104. Moreover, the switch 103 selects
5
the input source of the prediction error image according to the control of the prediction
parameter determiner 111, for example.
[0026] The frequency transformer/quantizer 104 frequency-transforms the prediction
error image and quantizes the frequency-transformed prediction error image (a coefficient
--------------------------------·------
5 image). The entropy encoder 112 performs entropy coding of a prediction parameter and of a
quantization coefficient image and then outputs a bitstream
[0027] The inverse quantizer/inverse frequency transformer 105 inverse-quantizes the
quantization coefficient image. Futiher, the inverse quantizer/inverse frequency transformer
105 performs inverse frequency transform on the inverse-quantized coefficient image. The
10 inverse frequency transformed reconstmcted prediction error image is input to the switch 106.
[0028] When the prediction error signal of the RGB space is used, the switch 106 is set
so that the reconstructed prediction error image is input to the switch 108. When the prediction
error signal of the YCoCr space is used, the switch I 06 is set so that the reconstructed prediction
error image is input to the inverse color space transformer 107. The switch 106 selects an
15 output destination of the reconstructed prediction error image according to the control of the
prediction parameter determiner 111.
[0029] The inverse color space transformer 107 transforms a reconstmcted prediction
error signal of the YCoCr space to a signal of the RGB space by using the aforementioned
equation (1) (a backward color space transform matrix) and then outputs the to the switch 108.
20 [003 0] The switch 108 selects the reconstructed prediction error signal received from
the switch 106 when the prediction error signal of the RGB space is used. When the prediction
error signal of the YCoCr space is used, the switch I 08 selects the reconstructed prediction error
signal received from the inverse color space transformer I 07. Moreover, the switch 108 selects
any of the reconstmcted prediction error images according to the control of the prediction
25 parameter determiner Ill, for example.
[0031] The reconstmcted prediction error image received from the switch 108 is
supplied with a prediction signal by the adder 116 and then supplied as a reconstructed image to
the buffer 109. The buffer 1 09 stores the reconstmcted image.
[0032] The prediction parameter determiner Ill instructs the predictor 110 to decide
30 the prediction parameter which minimizes coding cost by comparing the input image signal with
the prediction signal, for example. The prediction parameter determiner Ill supplies the
decided prediction parameter to the entropy encoder 112. The prediction parameter is
information related to block prediction such as a prediction mode (intra prediction, interprediction),
an intra prediction block size, an intra prediction direction, an inter-prediction block
5
6
size, a motion vector, or the like.
[0033] The prediction parameter determiner 111 fmther instructs to decide whether the
prediction error signal of the RGB space or the prediction error signal of the YCoCr space is
used for each block.
[0034] The bitstream output from the video coding device is transmitted to the video
decoding device. The video decoding device reconstructs images of a video by performing
decoding processing. Fig. 19 is a block diagram illustrating an example of the configuration of
a general video decoding device which decodes the bitstream output fi·om a general video coding
device to acquire a decoded image. Refetl'ing to Fig. 19, the configuration and operation of the
10 general video decoding device will be described below.
[0035] The video decoding device illustrated in Fig. 19 includes an entropy decoder
212, an inverse quantizer/inverse frequency transformer 205, a switch 206, an inverse color
space transformer 207, a switch 208, a buffer 209, a predictor 210, and an adder 216.
[0036] The entropy decoder 212 entropy-decodes an input bitstream. The entropy
15 decoder 212 supplies a quantization coefficient image to the inverse quantizer/inverse frequency
transformer 205 and supplies a prediction parameter to the predictor 210.
[0037] The inverse quantizer/inverse frequency transformer 205 inverse-quantizes the
input quantization coefficient image and outputs it as a coefficient image. Furthermore, the
inverse quantizer/inverse frequency transformer 205 transforms the coefficient image in the
20 frequency domain to an image in the space domain and outputs it as a prediction error image.
The prediction error image is input to the switch 206.
[0038] When the prediction error signal of the RGB space is used, the switch 206 is set
so that the prediction error image is input to the switch 208. When the prediction error signal of
the YCoCr space is used, the switch 206 is set so that the prediction enor image is input to the
25 inverse color space transformer 207. Moreover, the switch 206 is able to recognize whether the
prediction error signal ofthe RGB space or the prediction error signal of the YCoCr space should
be used according to signaling from the video coding device.
[0039] The inverse color space transformer 207 transforms the prediction error signal
of the YCoCr space to a signal of the RGB space by using the aforementioned equation (1) (the
30 backward color space transform matrix) and then outputs the prediction error signal to the switch
208.
[0040] When the prediction error signal of the RGB space is used, the switch 208
selects a prediction error signal received from the switch 206. When the prediction error signal
of the YCoCr space is used, the switch 208 selects the prediction error signal received fi·om the
11 ! 7
inverse color space transformer 207. The switch 208 is able to recognize whether the
prediction error signal of the RGB space or the prediction error signal of the YCoCr space should
be used according to signaling from the video coding device.
[0041] The prediction error image from the switch 208 is added to a prediction signal
5 supplied from the predictor 210 at the adder 216 and then supplied as a reconstructed image to
the buffer 209. The buffer 209 stores the reconstmcted image.
[0042] Moreover, the reconstmcted image stored in the buffer 209 is output as a
decoded image (decoded video).
[0043] The buffer 209 stores images, which have been decoded in the past, as
I 0 reference images. When performing the intra prediction, the predictor 21 0 generates a
prediction image by predicting an image to be decoded on the basis of the adjacent reconstructed
images, which have been decoded in the past within the image currently under decoding. When
performing the inter-prediction, the predictor 210 generates a prediction image on the basis of
reference images supplied from the buffer 209.
15 [0044] In RExt, a chroma quantization offset (a chroma QP offset) technique for a
20
purpose of subjective image quality improvement has been proposed. The chroma quantization
offset technique is used to adjust a quantization parameter for each color component by signaling
cluoma quantization offset values for a second color component and a third color component.
Specifically, this technique is able to change the quantization strength.
[0045] The following syntaxes are provided for signaling chroma quantization offset
values:
- Picture unit: pps cb qp offset/pps cr_ qp offset/slice_ qp delta_ cb/slice _ qp _delta_ cr
- Slice unit: slice_ qp delta_ cb/slice _ qp delta_ cr
-Block unit: cu_chroma_qp_offset_idx
25 [0046] The subjective image quality is able to be increased by adjusting the
30
quantization strength for each color component by using any of the above syntaxes.
[0047] The video coding device illustrated in Fig. 18 and the video decoding device
illustrated in Fig. 19 also apply a chroma quantization offset. As illustrated in Fig. 18, a
previously determined chroma quantization offset is input to the video coding device.
[0048) In the video coding device, the frequency transformer/quantizer I 04 increases
or decreases the quantization parameter of the B component according to the first cluoma
quantization offset and increases or decreases the quantization parameter of the R component
according to the second chroma quantization offset as illustrated in Fig. 20 when quantizing a
coefficient image when the prediction error signal of the RGB space is used. The inverse
" !
5
8
quantizer/inverse frequency transformer 105 increases or decreases the inverse quantization
parameter of the B component according to the first chroma quantization offset and increases or
decreases the inverse quantization parameter of the R component according to the second
cluoma quantization offset.
[0049] When the prediction error signal of the YCoCr space is used, the frequency
transformer/quantizer 104 increases or decreases the quantization parameter of the Co
component according to the first chroma quantization offset and increases or decreases the
quantization parameter of the Cr component according to the second chroma quantization offset
as illustrated in Fig. 20 when quantizing a coefficient image. The inverse quantizer/inverse
I 0 frequency transformer 105 increases or decreases the inverse quantization parameter of the Co
component according to the first chroma quantization offset and increases or decreases the
inverse quantization parameter of the Cr component according to the second chroma
quantization offset.
[0050] In the video decoding device, the inverse quantizer/inverse frequency
15 transformer 205 operates in the same manner as the inverse quantizer/inverse frequency
transformer 105 in the video coding device.
[0051] The cln·oma quantization offset technique is a technique of signaling the
cln·oma quantization offset values for the second color component and the third color component.
Therefore, if the adaptive color transform in residual domain is combined with the cln·oma
20 quantization offset, the block compressed in the RGB space and the block compressed in the
YCoCr space share the quantization strength as illustrated in Fig. 20. Thereby, the quantization
strength cannot be set appropriately according to the color space. Thus, the subjective image
quality improvement effect cannot be acquired by the chroma quantization offset technique.
[0052] It is an object of the present invention to provide a video coding device, a video
25 decoding device, a video coding method, a video decoding method, and a program capable of
preventing a deterioration in the subjective image quality improvement effect in the case where
the adaptive color transform in residual domain and the cln·oma quantization offset are used in
combination.
30 Solution to Problem
[0053] According to the present invention, there is provided a video coding device
capable of selecting, from among a plurality of color spaces, a color space of a prediction error
signal in coded block units, the device including: adaptive cln·oma quantization offset derivation
means for deriving a cln·oma quantization offset for each color space; and inverse quantization
" ! 9
means for inverse-quantizing a quantization coefficient image by using the chroma quantization
offset for each color space.
[0054] Fmthermore, according to the present invention, there is provided a video
decoding device capable of selecting, from among a plurality of color spaces, a color space of a
- ---------- ---- ---- ---- ------ --- ---- - - -
5 prediction error signal in coded block units, the device including: adaptive chroma quantization
offset derivation means for deriving a chroma quantization offset for each color space; and
inverse quantization means for inverse-quantizing a quantization coefficient image by using the
chroma quantization offset for each color space.
[0055] Furthermore, according to the present invention, there is provided a video
10 coding method capable of selecting, from among a plurality of color spaces, a color space of a
prediction error signal in coded block units, the method including: deriving a chroma
quantization offset for each color space; and inverse-quantizing a quantization coefficient image
by using the chroma quantization offset for each color space.
[0056] Furthermore, according to the present invention, there is provided a video
15 decoding method capable of selecting, from among a plurality of color spaces, a color space of a
prediction error signal in coded block units, the method including: deriving a chroma
quantization offset for each color space; and inverse-quantizing a quantization coefficient image
by using the chroma quantization offset for each color space.
[0057] Fmthermore, according to the present invention, there is provided a video
20 coding program for performing a video coding method capable of selecting, fi·om among a
plurality of color spaces, a color space of a prediction error signal in coded block units, the
program causing a computer to perform: a process of deriving a chroma quantization offset for
each color space; and a process of inverse-quantizing a quantization coefficient image by using
the chroma quantization offset for each color space.
25 [0058] Furthermore, according to the present invention, there is provided a video
decoding program for performing a video decoding method capable of selecting, from among a
plurality of color spaces, a color space of a prediction error signal in coded block units, the
program causing a computer to perform: a process of deriving a chroma quantization offset for
each color space; and a process of inverse-quantizing a quantization coefficient image by using
30 the chroma quantization offset for each color space.
Advantageous Effects oflnvention
[0059] According to the present invention, it is possible to prevent a deterioration in a
subjective image quality improvement effect.
Brief Description of Drawings
10
[0060]
[Fig. 1] Fig. 1 is a block diagram illustrating an exemplary embodiment of a video
coding device.
[Fig. 2] Fig. 2 is a flowchart illustrating processing related to signaling of chroma
5 quantization offsets.
[Fig. 3] Fig. 3 is a block diagram illustrating an exemplary embodiment of a video
decoding device.
[Fig. 4] Fig. 4 is a flowchart illustrating processing related to derivation of the chroma
quantization offsets.
10 [Fig. 5] Fig. 5 is an explanatory diagram illustrating an example of a syntax for
transmitting alt_pps _ cb qp offset and alt_pps cr_ qp _offset.
[Fig. 6] Fig. 6 is an explanat01y diagram illustrating an example of a syntax for
transmitting alt_ slice_ qp delta_ cb and alt_ slice_ qp delta_ cr.
[Fig. 7] Fig. 7 is an explanat01y diagram illustrating an example of a syntax for
15 transmitting alt_slice_qp_delta_cb and alt_slice_qp_delta_cr.
20
25
[Fig. 8] Fig. 8 is an explanat01y diagram illustrating an example of a syntax for
transmitting cb_qp_offset_list [i] and cr_qp_offset_list [i].
[Fig. 9] Fig. 9 is an explanat01y diagram illustrating an example of a syntax for
transmitting alt_cb_qp_offset_list [i] and alt_cr_qp_offset_list [i].
[Fig. 1 0] Fig. I 0 is a block diagram illustrating a configuration example of an
information processing system capable of implementing the functions of the video coding device
and the video decoding device.
[Fig. II] Fig. II is a block diagram illustrating the main part of the video coding
device.
[Fig. 12] Fig. 12 is a block diagram illustrating another example of the main patt of the
video coding device.
[Fig. 13] Fig. 13 is a block diagram illustrating still another example of the main part
of the video coding device.
[Fig. 14] Fig. 14 is a block diagram illustrating the main pati of the video decoding
30 device.
[Fig. 15] Fig. 15 is a block diagram illustrating another example of the main part of the
video decoding device.
[Fig. 16] Fig. 16 is a block diagram illustrating still another example of the main part
of the video decoding device.
5
11
[Fig. 17] Fig. 17 is an explanatory diagram illustrating an example of an adaptive color
transform in residual domain.
[Fig. 18] Fig. 18 is a block diagram illustrating the configuration of a general video
coding device.
[Fig. 19] Fig. 19 is a block diagram illustrating the configuration of a general video
decoding device.
[Fig. 20] Fig. 20 is an explanatory diagram illustrating an example of the use of
chroma quantization offsets.
10 Description of Embodiment
[0061]
Exemplary Embodiment 1
Fig. 1 is a block diagram illustrating a first exemplary embodiment of a video coding
device. Referring to Fig. I, description will be made on the configuration of a video coding
15 device which outputs a bitstream with each fi·ame of a digitized video as an input image.
[0062] As illustrated in Fig. 1, the video coding device of the first exemplary
embodiment includes a switch 101, a color space transformer 1 02, a switch 103, a frequency
transformer/quantizer 104, an inverse quantizer/inverse frequency transformer 105, a switch 106,
an inverse color space transformer 107, a switch 108, a buffer 109, a predictor 110, a prediction
20 parameter determiner 111, an entropy encoder 112, a subtractor 115, and an adder 116, similarly
to the general video coding device illustrated in Fig. 18.
[0063] As illustrated in Fig. 1, the video coding device fiuther includes an adaptive
chroma quantization offset derivation unit 121 and a switch 122.
[0064] Since the switch 101, the color space transformer 102, the switch 103, the
25 frequency transformer/quantizer 104, the inverse quantizer/inverse frequency transformer 105,
the switch 106, the inverse color space transformer 107, the switch 108, the buffer 109, the
predictor 110, the subtractor 115, and the adder 116 operate in the same manner as those
illustrated in Fig. 18, the following mainly describes the operations of the adaptive cln·oma
quantization offset derivation unit 121 and the switch 122 and the operations of the prediction
30 parameter determiner Ill and the entropy encoder 112 related to the signaling of the c!n·oma
quantization offsets. Moreover, the adaptive chroma quantization offset derivation unit 121
receives an input of a chroma quantization offset for the RGB space and an input of a c!n·oma
quantization offset for the YCoCr space.
[0065] Fig. 2 is a flowchart illustrating processing related to signaling of c!n·oma
12
quantization offsets.
[0066] The video coding device signals information indicating whether the adaptive
color transform in residual domain is performed or not by adaptive_color_trans_flag. Moreover,
when performing the adaptive color transform in residual domain, the video coding device
---------------- - -------------------
5 signals information indicating the color space ofthe block by cu_residual_csc_flag.
[0067] Unless the adaptive color transform in residual domain is performed, the
entropy encoder 112 signals adaptive_color_trans_flag = 0 and transmits the chroma
quantization offset for the RGB space derived by the adaptive chroma quantization offset (the
chroma quantization offset for the RGB space input by the adaptive chroma quantization offset
10 derivation unit 121) derivation unit 121 by using the following syntax (steps S101 and S102).
If the adaptive color transform in residual domain is performed, the entropy encoder 112 sets
adaptive_color_trans_flag = l. In addition, when compression is performed in the RGB space,
the entropy encoder 112 transmits the clll'oma quantization offset for the RGB space derived by
the adaptive chroma quantization offset derivation unit 121 by using the following syntax (steps
15 S103 and S104).
[0068] - Picture unit:
pps cb qp _ offset/pps _ cr qp offset/slice_ qp _delta_ cb/slice qp delta_ cr
- Slice unit: slice_ qp delta_ cb/slice qp delta_ cr
[0069] When the compression is performed in the YCoCr space, the entropy encoder
20 112 transmits the chroma quantization offset for the YCoCr space derived by the adaptive
chroma quantization offset derivation unit 121 by using the following syntax (steps S103 and
S105).
25
[0070] - Picture unit:
alt_pps cb qp offset/alt_pps cr_ qp offset/alt_slice qp delta_ cb/alt_ slice_ qp delta_ cr
-Slice unit: alt_slice_qp_delta_cb/alt_slice_qp_delta_cr
[0071] Moreover, in the case where the adaptive color transform in residual domain is
performed, the entropy encoder 112 signals cu_residual_csc_flag =I when the compression is
performed in the YCoCr space (when the compression is not performed in the RGB space).
The adaptive chroma quantization offset derivation unit 121 outputs the derived chroma
30 quantization offsets (the first chroma quantization offset and the second chroma quantization
offset) for the YCoCr space to the switch 122.
[0072] When the compression is performed in the RGB space, the entropy encoder 112
signals cu_residual_csc_flag = 0. The adaptive clu·oma quantization offset derivation unit 121
outputs the derived chroma quantization offsets (the first chroma quantization offset and the
"
13
second chroma quantization offset) for the ROB space to the switch 122.
[0073] Moreover, the adaptive chroma quantization offset derivation unit 121
recognizes whether the compression is performed in the ROB space or in the YCoCr space
according to cu _residual_ esc_ flag.
5 [0074] Moreover, the frequency transfonner/quantizer 104 adjusts the quantization
parameter by using the cln·oma quantization offset decided by the prediction parameter
determiner 111.
[0075] The prediction parameter determiner Ill stores the values of the cln·oma
quantization offsets for the ROB space and the values of the cln·oma quantization offsets for the
10 YCoCr space in advance, for example, and appropriately supplies the values of the cln·oma
quantization offsets for the ROB space or the values of the cln·oma quantization offsets for the
YCoCr space to the frequency transformer/quantizer 104. In that case, the values of the chroma
quantization offsets for the ROB space and the values of the cln·oma quantization offsets for the
YCoCr space are included in the prediction parameters supplied to the entropy encoder 112.
15 The entropy encoder 112 signals the values of the chroma quantization offsets for the ROB space
and the values of the chroma quantization offsets for the YCoCr space.
20
[0076] In this case, the video coding device explicitly signals the cln·oma quantization
offsets. Fmthermore, the video coding device signals the values of the cln·oma quantization
offsets.
[0077] Moreover, the operations of the adaptive chroma quantization offset derivation
unit 121 will be described in more detail in the second exemplary embodiment.
[0078] The operations of the video coding device other than the above operations are
the same as the operations of the video coding device illustrated in Fig. 18.
[0079]
25 Exemplary Embodiment 2
30
Fig. 3 is a block diagram illustrating the configuration of a video decoding device
which acquires a decoded image by decoding the bitstream output from the video coding device
which signals cln"Oma quantization offsets. Referring to Fig. 3, the configuration of the video
decoding device of the second exemplary embodiment will be described.
[0080] As illustrated in Fig. 3, the video decoding device of this exemplary
embodiment includes an entropy decoder 212, an inverse quantizer/inverse frequency
transformer 205, a switch 206, an inverse color space transformer 207, a switch 208, a buffer 209,
a predictor 210, and an adder 216 similarly to the general video decoding device illustrated in
Fig. 19.
14
[0081] As illustrated in Fig. 3, the video decoding device further includes an adaptive
chroma quantization offset derivation unit 221 and a switch 222.
[0082] Since the inverse quantizer/inverse frequency transformer 205, the switch 206,
the inverse color space transfmmer 207, the switch 208, the buffer 209, the predictor 210, and
5 the adder 216 operate in the same manner as those illustrated in Fig. 19, the following mainly
describes the operations of the adaptive chroma quantization offset derivation unit 221 and the
switch 222 and the operations of the entropy decoder 212 related to derivation of the clu·oma
quantization offsets.
[0083] Fig. 4 is a flowchart illustrating processing related to derivation of the chroma
10 quantization offsets.
[0084] If the entropy decoder 212 parses a bitstream to get adaptive_color_trans_flag =
1 (indicating that the adaptive color transform in residual domain is to be performed) (step 820 I)
and gets cu_residual_csc_flag = 1 (indicating that data is compressed in the YCoCr space) (step
8202), the adaptive chroma quantization offset derivation unit 221 derives the chroma
15 quantization offsets for the YCoCr space (step 8204). If the entropy decoder 212 gets
cu_residual_csc_flag = 0 (indicating that data is compressed in the RGB space) (step 8202), the
adaptive clu·oma quantization offset derivation unit 221 derives the chroma quantization offsets
for the RGB space (step 8203).
[0085] The adaptive chroma quantization offset derivation unit 221 derives the chroma
20 quantization offsets for the RGB space (the first clu·oma quantization offset qPicb and the second
clu·oma quantization offset qPicr) as follows.
25
[0086] qPicb = Clip3 (- QpBdOffsetc, 57, Qpy + pps_cb_qp_offset +
slice_ cb qp offset + CuQpOffsetcb)
qPicr = Clip3 (- QpBdOffsetc, 57, Qpy + pps_cr_qp_offset + slice_cr_qp_offset +
CuQpOffsetc,) (2)
[0087] In equation (2), Clip3 (x, y, z) is a function of clipping an input z into the [x, y]
range. Qpy is a quantization parameter of a first color component, CuQpOffsetcb is a clu·oma
quantization offset for each block of a second color component, and CuQpOffsetc, is a chroma
quantization offset for each block of a third color component. Although descriptions qPicb and
30 qPicr are used, qPicb corresponds to a chroma quantization offset for the B component and qPic,
corresponds to a chroma quantization offset for the R component in the case of the RGB space in
which the first color component is the G component, the second color component is the B
component, and the third color component is the R component.
[0088] The adaptive clu·oma quantization offset derivation unit 221 derives the clu·oma
15
quantization offsets for the YCoCr space (the first clll'oma quantization offset qPicb and the
second chroma quantization offset qPicr) as described in the following equation (3).
[0089] qPich = Clip3(- QpBdOffsetc, 57, Qpy + alt_pps_cb_qp_offset +
alt_slice_cb_qp_offset + CuQpOffsetcb)
5 qPic, = Clip3(- QpBdOffsetc, 57, Qpy + alt_pps_cr_qp_offset + alt_slice_cr_qp_offset +
CuQpOffsetc,) (3)
[0090] Moreover, quantization parameter (Qp'cb, Qp'c,) is calculated as described in
the following equation ( 4).
[0091] Qp'cb = qPcb + QpBdOffsetc
10 Qp'cr = qPc, + QpBdOffsetc (4)
[0092] The following describes a specific example of the procedure for deriving a
chroma quantization offset. In the description below, wordings enclosed by quotation marks
indicate characteristics in this exemplary embodiment.
[0093] - "If cu residual_ esc_ flag is equal to 0," the variables qPcb and qPcr are derived
15 as follows:
qPich = Clip3(- QpBdOffsetc, 57, Qpy + pps_cb_qp_offset + slice_cb_qp_offset +
CuQpOffsetcb)
qPicr = Clip3(- QpBdOffsetc, 57, Qpy + pps_cr_qp_offset + slice_cr_qp_offset +
CuQpOffsetc,)
20 -"Otherwise (cu_residual_csc_flag is equal to 1), the variables qPcb and
qPcr are derived as follows:"
"qPich = Clip3(- QpBdOffsetc, 57, Qpy + alt_pps_cb_qp_offset + alt_slice_cb_qp_offset +
CuQpOffsetc,)"
"qPicr = Clip3(- QpBdOffsetc, 57, Qpy + alt_pps_cr_qp_offset + alt_slice_cr_qp_offset +
25 CuQpOffsetc,)"
- IfClu·omaArrayType is equal to 1, the variables qPcb and qPc, are set equal to the value ofQpc
as specified in Predetermined Table based on the index qPi equal to qPicb and qPicr, respectively.
- Otherwise, the variables qP cb and qP cb are set equal to Min ( qPi, 51), based on the index qPi
equal to qPicb and qPicr, respectively.
30 -The chroma quantization parameters for the Cb and Cr components, Qp'cb and Qp'c,, are
derived as follows:
Qp'cb = qPcb + QpBdOffsetc
Qp'cr = qPcr + QpBdOffsetc
[0094] The inverse quantizerlinverse frequency transformer 205 increases or decreases
16
the quantization parameter according to the chroma quantization offsets from the adaptive
chroma quantization offset derivation unit 221 when inverse-quantizing the input quantization
coefficient image and outputting it as a coefficient image.
[0095]
5 Exemplary Embodiment 3
Subsequently, a video coding device according to a third exemplary embodiment will
be described. Fig. 5 is an explanatory diagram illustrating an example of a syntax for
transmitting alt_pps cb qp offset and alt_pps cr_ qp offset (an improvement of a syntax
described in "7.3.2.3.2 Picture parameter set range extensions syntax" of Non Patent Literature
10 1). In Fig. 5, wordings written in italics indicate characteristics of this exemplaty embodiment.
[0096] Figs. 6 and 7 are explanatory diagrams illustrating examples of syntaxes for
transmitting alt_slice qp delta_ cb and alt_slice qp delta_ cr (improvements of a syntax
described in "7.3.6.1 General slice segment header syntax"). In Figs. 6 and 7, wordings written
in italics indicate characteristics of this exemplary embodiment.
15 [0097] Moreover, the configuration of the video coding device of this exemplary
embodiment is the same as the configuration illustrated in Fig. I. In the video coding device,
the entropy encoder 112 transmits information by which the chroma quantization offsets for the
RGB space are able to be identified (for example, an index specifying a data table in which
chroma quantization offsets retained in the video decoding device are set or values of the chroma
20 quantization offsets) to the video decoding device.
[0098] When data is compressed in the YCoCr space, the entropy encoder 112 signals
the information by which the cluoma quantization offsets for the YCoCr space are able to be
identified (for example, the values of the chroma quantization offsets themselves) by using the
syntaxes illustrated in Fig. 5 and Figs. 6 and 7.
25 [0099]
Exemplary Embodiment 4
Subsequently, a video decoding device according to a foutih exemplary embodiment
will be described. The video decoding device of this exemplaty embodiment corresponds to the
video coding device ofthe third exemplaty embodiment. Note that the configuration of the
30 video decoding device of this exemplary embodiment is the same as the configuration illustrated
in Fig. 3.
[0100] When the entropy decoder 212 interprets that data is compressed in the YCoCr
space by using the syntaxes illustrated in Fig. 5 and Figs. 6 and 7 in the video decoding device,
the adaptive chroma quantization offset derivation unit 221 derives the chroma quantization
5
17
offsets in the same manner as in the second exemplary embodiment.
[0 10 1] Moreover, in the video coding device, the adaptive chroma quantization offset
derivation unit 121 operates in the same manner as the adaptive chroma quantization offset
derivation unit 221.
[0102]
Exemplary Embodiment 5
Subsequently, a video coding device according to a fifth exemplary embodiment will
be described. Fig. 8 is an explanatory diagram illustrating an example of a syntax for
additionally transmitting cb_qp_offset_list [i] and cr_qp_offset_list [i] for the YCoCr space (an
10 improvement of a syntax described in "7.3.2.3.2 Picture parameter set range extensions syntax"
ofNPL 1). In Fig. 8, wordings written in italics indicate characteristics of this exemplary
embodiment (specifically, the parts indicate that the size of cb qp _ offset_list!cr _ qp offset_list
[the range of chroma_ qp offset_list_len _minus!] is increased according to the value of
adaptive_ color_ trans_ flag). In the video coding device according to this exemplmy
15 embodiment, the quantization offsets are able to be switched between those for the RGB space
and those for the YCoCr space in block units by adjusting the value of the
cu_chroma_qp_offset_idx syntax transmitted in block units according to the value of the
cu residual_ esc_ flag syntax.
[0103] Note that the configuration of the video coding device of this exemplaty
20 embodiment is the same as the configuration illustrated in Fig. 1. In the video coding device,
the entropy encoder 112 transmits information by which the chroma quantization offsets for the
RGB space are able to be identified (for example, a cu chroma_ qp offset_idx syntax which is an
index for specifying a data table in which chroma quantization offsets retained in the video
decoding device are set) to the video decoding device.
25 [0104] According to this exemplaty embodiment, in the video coding device, the
entropy encoder 112 transmits information by which the chroma quantization offsets for the
YCoCr space are able to be identified (for example, a cu_chroma_qp_offset_idx syntax which is
an index for specifying a data table in which chroma quantization offsets retained in the video
decoding device are set) to the video decoding device. In the video decoding device of this
30 exemplaty embodiment, the chroma quantization offsets are able to be switched between those
for the RGB space and those for the YCoCr space in block units on the basis of the value of the
cu_chroma_qp_offset_idx syntax transmitted in block units according to the value of the
cu residual_ esc_ flag syntax.
[0105] Moreover, wordings written in italics (cb_qp_offset_list [i] and
5
10
18
cr_ qp offset_ list [i]) correspond to the chroma quantization offsets for the YCoCr space
described above in Fig. 8.
[0106]
Exemplary Embodiment 6
Next, a video decoding device according to a sixth exemplary embodiment will be
described. The video decoding device of this exemplary embodiment corresponds to the video
coding device of the fifth exemplmy embodiment. Moreover, the configuration of the video
decoding device of this exemplmy embodiment is the same as the configuration illustrated in Fig.
3.
[0107] When the entropy decoder 212 interprets that data is compressed in the YCoCr
space by the syntax illustrated in Fig. 8 in the video decoding device, the chroma quantization
offsets are read from the data table specified by the index, for example, and the adaptive chroma
quantization offset derivation unit 221 calculates the chroma quantization parameter in the same
manner as in the second exemplary embodiment.
15 [0 1 08] Moreover, in the video coding device, the adaptive chroma quantization offset
20
derivation unit 121 operates in the same manner as the adaptive clu·oma quantization offset
derivation unit 221.
[0109]
Exemplary Embodiment 7
Next, a video coding device according to a seventh exemplary embodiment will be
described. Fig. 9 is an explanatmy diagram illustrating an example of a syntax for transmitting
alt_ cb qp offset_list [i] and alt_ cr _qp offset_list [i] for the YCoCr space (an improvement of a
syntax described in "7.3.2.3.2 Picture parameter set range extensions syntax" described in NPL
1). In Fig. 9,wordings written in italics indicate characteristics of this exemplaty embodiment.
25 [0110] In the seventh exemplaty embodiment, the interpretation of the value of the
cu_chroma_qp_offset_idx syntax changes according to the value of the cu_residual_csc_flag
syntax in comparison with the fifth exemplary embodiment as described later, and therefore the
bits of the cu_clu·oma_qp_offset_idx syntax transmitted for each block are able to be saved.
For example, in the seventh ex em platy embodiment, even if cu chroma_ qp _ offset_idx = 0,
30 cb_qp_offset_list [0] and cr_qp_offset_list [0] for RGB are derived when cu_residual_csc_flag =
0, and alt_cb_qp_offset_list [0] and alt_cr_qp_offset_list [0] for YCoCr are derived when
cu_residual_csc_flag = 1. On the other hand, in the fifth exemplaty embodiment, when
cu_chroma_qp_offset_idx = 0, cb_qp_offset_list [0] and cr_qp_offset_list [0] for RGB are
derived. Therefore, in the fifth exemplary embodiment, if the list size is 4 (when
19
chroma_qp_offset_list_1en_minus1 is 3), it is necessary to transmit cu_chroma_qp_offset_idx =
4 in order to derive cb_qp_offset_list [4] and a1t_cr_qp_offset_list [4] forYCoCr.
[0111] Moreover, the configuration of the video coding device of this exemplary
embodiment is the same as the configuration illustrated in Fig. 1. In the video coding device,
-------------------------------------- ----- ------ -- ------ ------- -------
5 the entropy encoder 112 transmits information by which the chroma quantization offsets for the
ROB space are able to be identified (for example, an index for specifying a data table in which
the chroma quantization offsets retained in the video decoding device are set) to the video
decoding device.
[0112] According to this exemplary embodiment, in the video coding device, the
10 entropy encoder 112 transmits information by which the chroma quantization offsets for the
YCoCr space are able to be identified (for example, an index for specifying a data table in which
the chroma quantization offsets retained in the video decoding device are set) to the video
decoding device.
[0113]
15 Exemplary Embodiment 8
Subsequently, a video decoding device according to an eighth exemplary embodiment
will be described. The video decoding device of this exemplmy embodiment corresponds to the
video coding device of the seventh exemplary embodiment. Moreover, the configuration of the
video decoding device of this exemplary embodiment is the same as the configuration illustrated
20 in Fig. 3.
[0114] When the entropy decoder 212 interprets that data is compressed in the YCoCr
space by the syntax illustrated in Fig. 9 in the video decoding device, the chroma quantization
offsets are read from the data table specified by the index, for example, and the adaptive chroma
quantization offset derivation unit 221 calculates the chroma quantization parameter in the same
25 manner as in the second exemplaty embodiment.
[0115] Moreover, in the video coding device, the adaptive chroma quantization offset
derivation unit 121 operates in the same manner as the adaptive chroma quantization offset
derivation unit 221.
[0 116] The following describes a specific example of a procedure for deriving chroma
30 quantization offsets. In the following description, wordings enclosed by quotation marks
indicate characteristics in this exemplaty embodiment.
[0 117]
cu_chroma_qp_offset_idx, when present, specifies the index into the cb_qp_offset_list []and
cr_qp_offset_list []or the alt_cb_qp_offset_list []and alt_cr_qp_offset_list []that is used to
" ! 20
determine the value of CuQpOffsetCb and CuQpOffsetCr. When present, the value of
cu _chroma_ qp _ offset_idx shall be in the range of 0 to clu·oma qp _ offset_list_len minus 1,
inclusive. When not present, the value of cu _ clu·oma_qp offset_idx is inferred to be equal to 0.
When cu clu·oma qp offset_ flag is present, the following applies:
5 - The variable IsCuClu·omaQpOffsetCoded is set equal to 1.
-The variables CuQpOffsetCb and CuQpOffsetCr are derived as follows:
- Ifcu_chroma_qp_offset_flag is equal to 1 and "cu_residual_csc_flag is equal to 0", the
following applies:
CuQpOffsetCb = cb_qp_offset_list [cu_chroma_qp_offset_idx]
10 CuQpOffsetCr = cr_ qp offset_list [ cu _ clu·oma qp offset_idx]
-"Otherwise if cu_chroma_qp_offset_flag is equal to 1 and cu_residual_csc_flag is equal to 1,
the following applies:"
"CuQpOffsetCb = alt_cb_qp_offset_list [cu_clu·oma_qp_offset_idx]"
"CuQpOffsetCr = alt_ cr _ qp offset_ list [ cu cluoma _ qp offset_idx ]"
15 - Othe1wise ( cu _chroma_ qp offset_ flag is equal to 0), CuQpOffsetCb and CuQpOffsetCr are
both set equal to 0.
[0118]
Exemplary Embodiment 9
Although the video coding device explicitly signals the chroma quantization offsets in
20 the above exemplary embodiments, the signaling of the chroma quantization offsets may be
omitted though it is signaled that the color space of the prediction error signal is selected in block
units. In the present description, the above is referred to as chroma quantization offsets are
implicitly signaled.
[0 119] In the case where the video coding device implicitly signals clu·oma
25 quantization offsets, the entropy encoder signals adaptive_ color_ trans_ flag = 1 and then signals
cu_residual_csc_flag in block units, for example, while not signaling information by which the
values of the clu·oma quantization offsets are able to be identified.
[0120] In case where the entropy decoder parses a bitstream to get
adaptive_ color_ trans_ flag = 1 in the video decoding device and gets cu residual_ esc_ flag = 0
30 (indicating that data is compressed in the RGB space), the adaptive chroma quantization offset
derivation unit 221 reads out the values of the chroma quantization offsets for the RGB space
previously stored in the video decoding device. In addition, if the entropy decoder gets
cu_residual_csc_flag = 1 (indicating that data is compressed in the YCoCr space), the adaptive
clu·oma quantization offset derivation unit 221 calculates the values of the cluoma quantization
21
offsets for the YCoCr space fi·om the values of the previously-stored chroma quantization offsets
for the RGB space.
[0121] Since the c!n·oma quantization offsets for the RGB space correlate with the
c!n·oma quantization offsets for the YCoCr space to some extent, in other words, since it is
---------------- --- ----- --- --- ----
5 possible to define a calculation formula for calculating the chroma quantization offsets for the
YCoCr space from the c!n·oma quantization offsets for the RGB space, the adaptive chroma
quantization offset derivation unit 221 is able to derive the c!n·oma quantization offsets for the
YCoCr space by using the calculation formula.
[0122] Specifically, the video decoding device implicitly derives the c!n·oma
I 0 quantization offsets.
[0123] Moreover, in the video coding device, the adaptive c!n·oma quantization offset
derivation unit 121 operates in the same manner as the adaptive chroma quantization offset
derivation unit 221.
[0124] Moreover, if the video coding device implicitly signals c!n·oma quantization
15 offsets, the amount of data to be transmitted is able to be reduced.
[0125] Although the RGB space and the YCoCr space are illustrated as two color
spaces in the above exemplary embodiments, the system of the above exemplmy embodiments is
applicable even if one or both of the two color spaces are other color spaces. Furthermore,
although the first color component G, the second color component B, and the third color
20 component Rare used in the RGB space in the above exemplary embodiments (see Fig. 20), the
way of allocating color signals to the respective color components is not limited thereto, but
arbitrary color signals can be allocated to the respective color components.
[0 126] Although the video coding device and the video decoding device use two color
spaces in the above exemplaty embodiments, these devices are also allowed touse t!n·ee or more
25 color spaces.
[0 127] Furthermore, each of the exemplary embodiments described above may be
realized by hardware, but may also be realized by a computer program.
[0128] The information processing system illustrated in Fig. 10 includes a processor
100 I, a program memory 1002, a storage medium 1003 for storing video data, and a storage
30 medium 1004 for storing a bitstream. The storage medium 1003 and the storage medium 1004
may be separate storage media, or storage areas included in the same storage medium. A
magnetic storage medium such as a hard disk is available as such a storage medium.
[0129] In the information processing system illustrated in Fig. 10, the program
memory I 002 stores a program for realizing the functions of the blocks (except the block of the
1l ! 22
buffer) illustrated in each of Figs. 1 and 3. The processor 1001 realizes the functions of the
video coding device or video decoding device illustrated in Fig. 1 or Fig. 3 by executing the
processes according to the program stored in the program memory 1002.
[0130] Fig. 11 is a block diagram illustrating the main pati of the video coding device.
5 As illustrated in Fig. 11, the video coding device 301 includes an adaptive chroma quantization
offset derivation unit 311 (corresponding to the adaptive chroma quantization offset derivation
unit 121 illustrated in Fig. 1, for example) which derives chroma quantization offsets for each
color space and an inverse quantization unit 312 (corresponding to the inverse quantizer/inverse
frequency transformer 105 illustrated in Fig. I, for example) which inverse-quantizes a
10 quantization coefficient image by using the chroma quantization offsets for each color space.
[0131] Fig. 12 is a block diagram illustrating another example of the main part of the
video coding device. As illustrated in Fig. 12, the video coding device 302 further includes a
color space selection notification unit 313 (corresponding to the entropy encoder 112 illustrated
in Fig. I, for example) which signals that the color space of a prediction error signal is to be
15 selected in block units.
20
[0132] Moreover, unless the video coding device 302 includes a means for signaling
information by which the values of the quantization offsets for each color space are able to be
identified in the configuration illustrated in Fig. 12, the chroma quantization offsets are
implicitly derived in the video coding device 302.
[0133] Fig. 13 is a block diagram illustrating still another example of the main part of
the video coding device. As illustrated in Fig. 13, the video coding device 303 futiher includes
a quantization offset information transmission unit 314 (corresponding to the entropy encoder
112 illustrated in Fig. I, for example) which signals information by which the values of the
chroma quantization offsets for each color space are able to be identified. The information by
25 which the values of chroma quantization offsets are able to be identified is the values themselves
of the chroma quantization offsets or an index for specifYing the data table in which the chroma
quantization offsets retained in the video decoding device are set, for example.
[0134] Fig. 14 is a block diagram illustrating the main part of the video decoding
device. As illustrated in Fig. 14, the video decoding device 40 I includes an adaptive c!U'oma
30 quantization offset derivation unit 411 (corresponding to the adaptive chroma quantization offset
derivation unit 221 illustrated in Fig. 3, for example) which derives chroma quantization offsets
for each color space and an inverse quantization unit 412 (corresponding to the inverse
quantizer/inverse frequency transformer 205 illustrated in Fig .3, for example) which inversequantizes
a quantization coefficient image by using the c!U'oma quantization offsets for each
23
color space.
[0135] Fig. 15 is a block diagram illustrating another example of the main part of the
video decoding device. As illustrated in Fig. 15, the video decoding device 402 further
includes a color space selection parsing unit 413 (corresponding to the entropy decoder 212
5 illustrated in Fig. 3, for example) which parses a bitstream to interpret that the color space of a
prediction error signal is to be selected in block units.
[0 136] Moreover, unless the video decoding device 402 includes a means for parsing a
bitstream to interpret information by which the values of the chroma quantization offsets for each
color space are able to be identified in the configuration illustrated in Fig. 15, the video decoding
10 device 402 implicitly derives the chroma quantization offset.
[0137] Fig. 16 is a block diagram illustrating still another example of the mainpatt of
the video decoding device. As illustrated in Fig. 16, the video decoding device 403 further
includes a clu·oma quantization offset parsing unit 414 (corresponding to the entropy decoder
212 illustrated in Fig. 3, for example) which identifies the values of the chroma quantization
15 offsets for each color space on the basis of the information interpreted from the received
bitstream.
[0138] Although the present invention has been described with reference to the
foregoing exemplary embodiments and examples, the present invention is not limited to the
foregoing exemplary embodiments and examples. Various changes understandable by those
20 skilled in the art can be made to the structures and details of the present invention within the
scope of the present invention.
25
30
[0 139] This application claims priority to Japanese Patent Application No. 2014-
204392 filed on October 3, 2014, and the entire disclosure thereof is hereby incorporated herein
by reference.
Reference Signs List
[0140]
101 Switch
102 Color space transformer
103 Switch
1 04 Frequency transformer/quantizer
105 Inverse quantizer/inverse frequency transformer
106 Switch
107 Inverse color space transformer
" ! 24
108 Switch
109 Buffer
110 Predictor
111 Prediction parameter detenniner
5 T12 Entropy encoder
115 Subtracter
116 Adder
121 Adaptive chroma quantization offset derivation unit
122 Switch
10 205 Inverse quantizer/inverse frequency transformer
206 Switch
207 Inverse color space transformer
208 Switch
209 Buffer
15 210 Predictor
212 Entropy decoder
216 Adder
221 Adaptive clu-oma quantization offset derivation unit
222 Switch
20 301, 302, 303 Video coding device
311 Adaptive chroma quantization offset derivation unit
312 Inverse quantization unit
313 Color space selection notification unit
314 Quantization offset information transmission unit
25 401,402,403 Video decoding device
411 Adaptive clu·oma quantization offset derivation unit
412 Inverse quantization unit
413 Color space selection parsing unit
414 Chrominance quantization offset parsing unit
30 1001 Processor
1002 Program memory
1003 Storage medium
1004 Storage medium

Claims
1. A video coding device capable of selecting, fi·om among a plurality of color spaces, a
color space of a prediction error signal in coded block units, the device comprising:
.... aifaptivechio-maquani1zafion offset defivationmeaiiaor denvinga chi-oma
quantization offset for each color space; and
inverse quantization means for inverse-quantizing a quantization coefficient image by
using the chroma quantization offset for each color space.
2. The video coding device according to claim 1, further comprising color space selection
notification means for signaling that the color space of the prediction error signal is to be
selected in block units.
3. The video coding device according to claim 2, futiher comprising chroma quantization
15 offset information transmission means for signaling information by which the value of the
chroma quantization offset for each color space is able to be identified.
20
25
4. A video decoding device capable of selecting, from among a plurality of color spaces,
a color space of a prediction error signal in coded block units, the device comprising:
adaptive chroma quantization offset derivation means for deriving a chroma
quantization offset for each color space; and
inverse quantization means for inverse-quantizing a quantization coefficient image by
using the chroma quantization offset for each color space.
5. The video decoding device according to claim 4, further comprising color space
selection parsing means for parsing a received bitstream to interpret that the color space of the
prediction error signal is to be selected in block units.
6. The video decoding device according to claim 5, further comprising chroma
30 quantization offset parsing means for identifying the value of the quantization offset for each
color space based on information obtained from the received bitstream.
7. A video coding method capable of selecting, fi·om among a plurality of color spaces, a
color space of a prediction error signal in coded block units, the method comprising:
5
26
deriving a chroma quantization offset for each color space; and
inverse-quantizing a quantization coefficient image by using the chroma quantization
offset for each color space.
8. The video coding method according to claim 7, further comprising signaling that the
color space of the prediction error signal is to be selected in block units.
9. The video coding method according to claim 8, further comprising signaling
information by which the value of the chroma quantization offset for each color space is able to
1 0 be identified.
10. A video decoding method capable of selecting, from among a plurality of color spaces,
a color space of a prediction error signal in coded block units, the method comprising:
deriving a chroma quantization offset for each color space; and
15 inverse-quantizing a quantization coefficient image by using the cln·oma quantization
offset for each color space.
11. The video decoding method according to claim 10, further comprising parsing a
received bitstream to interpret that the color space of the prediction error signal is to be selected
20 in block units.
25
30
12. The video decoding method according to claim 11, further comprising identifying the
value of the quantization offset for each color space based on information obtained from the
received bitstream.
13. A video coding program for performing a video coding method capable of selecting,
fi·om among a plurality of color spaces, a color space of a prediction error signal in coded block
units, the program causing a computer to perform:
a process of deriving a cln·oma quantization offset for each color space; and
a process of inverse-quantizing a quantization coefficient image by using the cln·oma
quantization offset for each color space.
14. The video coding program according to claim 13, causing the computer to perform a
process of signaling that the color space of the prediction error signal is to be selected in block
27
units.
I
15. ,The video codingptogram accordii1g\:o claim 14, clausing the computer to perfonn a
process of signaling informatifn by. which the value of the clu·oma quantization offset for each
--5---colorspace-is-ableto-be-ident\fied-.· ------'- ------
10
t6. A video decoding program for performing a video decoding method qtpable of
' selecting, from among a plurality of color spaces, a color space of a prediction (\rror signal in
coded block units, the program causing a computer to perfmm:
a process of deriving a cln·oma quantization offset for each color spac~; and
a process of inverse-quantizing a quantization coefficient image by using the clu·oma
_quantization offset for each color space.
17. The video decoding program according to claiml6, causiug a computer to perform a
15 process of parsing a receiv~d bitstream to interpret that the color space of the prediction enor
signal is to be selected in block units.
18. The video decoding program according to claim 17, causing the computer to perfonn a
_ process of identifying the value of the clu·oma quantization offset for each color space based on
20 information obtained from the received bitstream.