Description
Title of Invention
IMAGE PROCESSING DEVICE AND IMAGE PROCESSING METHOD
5
Technical Field
[OOOl]
The present disclosure relates to an image processing device and an image
processing method.
10
Background Art
[0002]
H.264/AVC, one of standard specifications for image encoding scheme,
applies a deblocking filter to a block boundary in units of blocks each containing 4x4
15 pixels, for example, in order to prevent image quality degradation due to block
distortion while an image is encoded. The deblocking filter requires a large amount
of processing and may account for 50% of the entire processing amount in image
decoding, for example.
[0003]
20 The standards work for High Efficiency Video Coding (HEVC), a nextgeneration
image encoding scheme, proposes application of the deblocking filter in
units of blocks each containing 8x8 pixels or more according to JCTVC-A119 (see
Non-Patent Literature 1). The technique proposed in JCTVC-A119 increases the
block size to a minimum unit which allows for applying the deblocking filter to
25 perform filtering processes in parallel on block boundaries in the same direction
within one macro block.
Citation List
Non-Patent Literature
[0004]
30 Non-Patent Literature 1 : K.Ugur (Nokia), K.R.Andersson (LM Ericsson),
A.Fuldseth (Tandberg Telecom), "JCTVC-A119:Video coding technology proposal
by Tandberg, Nokia, and Ericsson", Documents of the first meeting of the Joint
Collaborative Team on Video Coding (JCT-VC), Dresden, Germany, 15-23 April,
20 10.
Summary of Invention
Technical Problem
[0005]
Even if the technique proposed in JCTVC-A119 is used, there remains
dependency between a process on the vertical block boundary and a process on the
10 horizontal block boundary. Specifically, a process on the vertical boundary for one
macro block waits until a process on the horizontal boundary for a neighboring
macro block is performed. A process on the horizontal boundary for one macro
block waits until a process on the vertical boundary for the same macro block is
performed. The above-described technique can just provide a parallel process of
15 the deblocking filter to a very limited extent. Accordingly, the above-described
technique may not successfully solve problems of a delay and a decrease in data rates
due to a large processing amount while the deblocking filter is applied.
[0006]
The technology according to the disclosure aims at providing an image
20 processing device and an image processing method capable of providing further
parallel processing when a deblocking filter is applied.
Solution to Problem ,
[0007]
25 According to an embodiment of the present disclosure, there is provided an
image processing device including a decoding section configured to decode an image
from an encoded stream, a horizontal filtering section configured to apply a
deblocking filter to a vertical block boundary within an image to be decoded by the
decoding section, a vertical filtering section configured to apply a deblocking filter to
30 a horizontal'block boundary within an image to be decoded by the decoding section,
and a control section configured to cause the horizontal filtering section to filter in
parallel a plurality of vertical block boundaries included in a processing unit
containing a plurality of coding units and cause the vertical filtering section to filter
in parallel a plurality of horizontal block boundaries included in the processing unit.
[OOOS]
5 The image processing device can be realized typically as an image decoding
device for decoding an image.
[0009]
According to an embodiment of the present disclosure, there is provided an
image processing method including decoding an image from an encoded stream,
10 performing horizontal filtering to apply a deblocking filter to a vertical block
boundary within an image to be decoded, performing vertical filtering to apply a
deblocking filter to a horizontal block boundary within an image to be decoded, and
controlling the horizontal filtering and the vertical filtering so as to filter in parallel a
plurality of vertical block boundaries included in a processing unit containing a
15 plurality of coding units and filter in parallel a plurality of horizontal block
boundaries included in the processing unit.
[OO 1 01
According to an embodiment of the present disclosure, there is provided an
image processing device including a horizontal filtering section configured to apply a
20 deblocking filter to a vertical block boundary within an image to be locally decoded
when encoding an image to be encoded, a vertical filtering section configured to
apply a deblocking filter to a horizontal block boundary within the image, a control
section configured to cause the horizontal filtering section to filter in parallel a
plurality of vertical block boundaries included in a processing unit containing a
25 plurality of coding units and cause the vertical filtering section to filter in parallel a
plurality of horizontal block boundaries included in the processing unit, and an
encoding section configured to encode the image to be encoded using an image
filtered by the horizontal filtering section and the vertical filtering section.
[OOll]
30 The image processing device can be realized typically as an image encoding
device for encoding an image.
[OO 1 21
According to an embodiment of the present disclosure, there is provided an
image processing method including performing horizontal filtering to apply a
deblocking filter to a vertical block boundary within an image to be locally decoded
5 when encoding an image to be encoded, performing vertical filtering to apply a
deblocking filter to a horizontal block boundary within the image, controlling the
horizontal filtering and the vertical filtering so as to filter in parallel a plurality of
vertical block boundaries included in a processing unit containing a plurality of
coding units and filter in parallel a plurality of horizontal block boundaries included
10 in the processing unit, and encoding the image to be encoded using an image filtered
by the horizontal filtering and the vertical filtering.
Advantageous Effects of Invention
[00 131
15 As described above, the image processing device and the image processing
method according to the present disclosure further improves parallel processing
when a deblocking filter is applied.
Brief Description of Drawings
20 [0014]
[Fig. 11 Fig. 1 is a block diagram showing an example of a configuration of an image
encoding device according to an embodiment.
[Fig. 21 Fig. 2 is a block diagram showing an example of a configuration of an image
decoding device according to an embodiment.
25 [Fig. 31 Fig. 3 is an explanatory diagram showing an example of neighboring pixels
around a boundary.
[Fig. 41 Fig. 4 is an explanatory diagram illustrating reference pixels during filtering
need determination processes according to an existing technique.
[Fig. 51 Fig. 5 is an explanatory diagram illustrating pixels updated by filtering
30 processes.
[Fig. 61 Fig. 6 is an explanatory diagram illustrating identification of edges for
description of the embodiment.
[Fig. 71 Fig. 7 is an explanatory diagram illustrating a parallel process according to
an existing technique.
[Fig. 81 Fig. 8 is a first explanatory diagram illustrating dependency between
5 processes according to an existing technique.
[Fig. 91 Fig. 9 is a second explanatory diagram illustrating dependency between
processes according to an existing technique.
[Fig. 101 Fig. 10 is an explanatory diagram illustrating a sequence of processes
according to an existing technique.
10 [Fig. 111 Fig. 11 is an explanatory diagram illustrating a sequence of processes
according to a first working example.
[Fig. 121 Fig. 12 is a block diagram illustrating a detailed configuration of a
deblocking filter according to the first embodiment.
[Fig. 131 Fig. 13 is a block diagram illustrating a detailed configuration of a
15 determination section.
[Fig. 141 Fig. 14 is an explanatory diagram illustrating neighboring blocks around a
slice boundary.
[Fig. 151 Fig. 15 is an explanatory diagram illustrating a first example of a sequence
of processes for each slice.
20 [Fig. 161 Fig. 16 is an explanatory diagram illustrating a second example of a
sequence of processes for each slice.
[Fig. 171 Fig. 17 is an explanatory diagram illustrating first and second examples of a
determination technique provided by a modification.
[Fig. 181 Fig. 18 is an explanatory diagram illustrating third and fourth examples of a
25 determination technique provided by a modification.
[Fig. 191 Fig. 19 is an explanatory diagram illustrating fifth and sixth examples of a
determination technique provided by a modification.
[Fig. 201 Fig. 20 is a flowchart illustrating a process flow for the deblocking filter
according to the first working example.
30 [Fig. 211 Fig. 21 is a flowchart illustrating a flow of a filtering need determination
process.
[Fig. 221 Fig. 22 is an explanatory diagram illustrating a sequence of processes
according to a second working example.
[Fig. 231 Fig. 23 is a block diagram illustrating a detailed configuration of the
deblocking filter according to the second working example.
5 [Fig. 241 Fig. 24 is a flowchart illustrating a process flow for the deblocking filter
according to the second working example.
[Fig. 251 Fig. 25 is an explanatory diagram illustrating a process sequence for each
LCU.
[Fig. 261 Fig. 26 is a flowchart illustrating a process flow for each LCU.
10 [Fig. 271 Fig. 27 is an explanatory diagram illustrating an overview of a third
working example.
[Fig. 281 Fig. 28 is a block diagram illustrating a detailed configuration of a
deblocking filter according to the third working example.
[Fig. 291 Fig. 29 is an explanatory diagram illustrating determination of a weight for
15 weighted average.
[Fig. 301 Fig. 30 is an explanatory diagram illustrating an example of a weight for
weighted average.
[Fig. 3 11 Fig. 3 1 is an explanatory diagram illustrating an output pixel value from a
calculation section according to the third working example.
20 [Fig. 321 Fig. 32 is an explanatory diagram illustrating a first example of process
sequence for comparison.
[Fig. 331 Fig. 33 is an explanatory diagram illustrating a first example of process
sequence provided by the third working example.
[Fig. 341 Fig. 34 is an explanatory diagram illustrating a second example of process
25 sequence for comparison.
[Fig. 351 Fig. 35 is an explanatory diagram illustrating a second example of process
sequence provided by the third working example.
[Fig. 361 Fig. 36 is a flowchart illustrating a first example of a process flow for the
deblocking filter according to the third working example.
30 [Fig. 371 Fig. 37 is a flowchart illustrating a flow of a pixel value calculation process
shown in Fig. 36.
[Fig. 381 Fig. 38 is an explanatory diagram illustrating multiview codec.
[Fig. 391 Fig. 39 is an explanatory diagram illustrating an image encoding process
according to an embodiment applied to multiview codec.
[Fig. 401 Fig. 40 is an explanatory diagram illustrating an image decoding process
5 according to an embodiment applied to multiview codec.
[Fig. 411 Fig. 41 is an explanatory diagram illustrating scalable codec.
[Fig. 421 Fig. 42 is an explanatory diagram illustrating an image encoding process
according to an embodiment applied to scalable codec.
[Fig. 431 Fig. 43 is an explanatory diagram illustrating an image decoding process
10 according to an embodiment applied to scalable codec.
[Fig. 441 Fig. 44 is a block diagram illustrating a schematic configuration of a
television apparatus.
[Fig. 451 Fig. 45 is a block diagram illustrating a schematic configuration of a mobile
phone.
15 [Fig. 461 Fig. 46 is a block diagram illustrating a schematic configuration of a
recording/reproduction device.
[Fig. 471 Fig. 47 is a block diagram illustrating a schematic configuration of an
image capturing device.
20 Description of Embodiment
[00 1 51
Hereinafter, preferred embodiments of the present invention will be
described in detail with reference to the appended drawings. Note that, in this
specification and the drawings, elements that have substantially the same function
25 and structure are denoted with the same reference signs, and repeated explanation is
omitted.
[00 1 61
Description of Embodiment will be described in the following sequence.
1. Apparatus Overview
1 - 1. Image Encoding Device
1-2. Image Decoding Device
2. Existing Technique
2-1. Basic Configuration of Deblocking Filter
2-2. Dependency Between Processes According to an Existing
Technique
5 3. First Working Example
3- 1. Deblocking Filter Configuration Example
3-2. Determination Condition Modifications
3-3. Process Flow
4. Second Working Example
4-1. Deblocking Filter Configuration Example
4-2. Process Flow
4-3. Process Example for Each LCU
5. Third Working Example
5-1. Overview
5-2. Deblocking Filter Configuration Example
5-3. Process Sequence Example
5-4. Process Flow
6. Application to Various Codecs
6- 1. Multiview Codec
20 6-2. Scalable Codec
7. Example Applications
8. Summing-up
[00 1 71
1. Apparatus Overview
2 5 With reference to Figs. 1 and 2, the following describes an overview of an
apparatus to which the technology disclosed in this specification is applicable. The
technology disclosed in this specification is applicable to an image encoding device
and an image decoding device, for example.
[00 1 81
30 [I - 1. Image Encoding Device]
Fig. 1 is a block diagram showing an example of a configuration of an
image encoding device 10 according to an embodiment. Referring to Fig. 1, the
image encoding device 10 includes an A/D (Analogue to Digital) conversion section
11, a reordering buffer 12, a subtraction section 13, an orthogonal transform section
14, a quantization section 15, a lossless encoding section 16, an accumulation buffer
5 17, a rate control section 18, an inverse quantization section 21, an inverse
orthogonal transform section 22, an addition section 23, a deblocking filter 24a, a
frame memory 25, a selector 26, an intra prediction section 30, a motion estimation
section 40, and a mode selection section 50.
[00 191
10 The A/D conversion section 11 converts an image signal input in an
analogue format into image data in a digital format, and outputs a series of digital
image data to the reordering buffer 12.
[0020]
The reordering buffer 12 reorders the images included in the series of image
15 data input from the A/D conversion section 11. After reordering the images
according to the a GOP (Group of Pictures) structure according to the encoding
process, the reordering buffer 12 outputs the image data which has been reordered to
the subtraction section 13, the intra prediction section 30, and the motion estimation
section 40.
20 [0021]
The image data input from the reordering buffer 12 and predicted image
data selected by the mode selection section 50 described later are supplied to the
subtraction section 13. The subtraction section 13 calculates predicted error data
which is a difference between the image data input from the reordering buffer 12 and
25 the predicted image data input from the mode selection section 50, and outputs the
calculated predicted error data to the orthogonal transform section 14.
[0022]
The orthogonal transform section 14 performs orthogonal transform on the
predicted error data input from the subtraction section 13. The orthogonal
30 transform to be performed by the orthogonal transform section 14 may be discrete
cosine transform (DCT) or Karhunen-Loeve transform, for example. The
orthogonal transform section 14 outputs transform coefficient .data acquired by the
orthogonal transform process to the quantization section 15.
[0023]
The transform coefficient data input from the orthogonal transform section
5 14 and a rate control signal from the rate control section 18 described later are
supplied to the quantization section 15. The quantization section 15 quantizes the
transform coefficient data, and outputs the transform coefficient data which has been
quantized (hereinafter, referred to as quantized data) to the lossless encoding section
16 and the inverse quantization section 21. Also, the quantization section 15
10 switches a quantization parameter (a quantization scale) based on the rate control
signal from the rate control section 18 to thereby change the bit rate of the quantized
data to be input to the lossless encoding section 16.
[0024]
The quantized data input from the quantization section 15 and information
15 described later about intra prediction or inter prediction generated by the intra
prediction section 30 or the motion estimation section 40 and selected by the mode
selection section 50 are supplied to the lossless encoding section 16. The
information about intra prediction may include prediction mode information
indicating an optimal intra prediction mode for each block, for example. Also, the
20 information about inter prediction may include prediction mode information for
prediction of a motion vector for each block, difference motion vector information,
reference image information, and the like, for example.
[0025]
The lossless encoding section 16 generates an encoded stream by
25 performing a lossless encoding process on the quantized data. The lossless
encoding by the lossless encoding section 16 may be variable-length coding or
arithmetic coding, for example. Furthermore, the lossless encoding section 16
multiplexes the information about intra prediction or the information about inter
prediction mentioned above to the header of the encoded stream (for example, a
30 block header, a slice header or the like). Then, the lossless encoding section 16
outputs the generated encoded stream to the accumulation buffer 17.
[0026]
The accumulation buffer 17 temporarily stores the encoded stream input
from the lossless encoding section 16 using a storage medium, such as a
semiconductor memory. Then, the accumulation buffer 17 outputs the accumulated
5 encoded stream at a rate according to the band of a transmission line (or an output
line from the image encoding device 10).
[0027]
The rate control section 18 monitors the free space of the accumulation
buffer 17. Then, the rate control section 18 generates a rate control signal
10 according to the free space on the accumulation buffer 17, and outputs the generated
rate control signal to the quantization section 15. For example, when there is not
much free space on the accumulation buffer 17, the rate control section 18 generates
a rate control signal for lowering the bit rate of the quantized data. Also, for
example, when the free space on the accumulation buffer 17 is sufficiently large, the
15 rate control section 18 generates a rate control signal for increasing the bit rate of the
quantized data.
[0028]
The inverse quantization section 21 performs an inverse quantization
process on the quantized data input from the quantization section 15. Then, the
20 inverse quantization section 21 outputs transform coefficient data acquired by the
inverse quantization process to the inverse orthogonal transform section 22.
[0029]
The inverse orthogonal transform section 22 performs an inverse orthogonal
transform process on the transform coefficient data input from the inverse
25 quantization section 21 to thereby restore the predicted error data. Then, the inverse
orthogonal transform section 22 outputs the restored predicted error data to the
addition section 23.
[0030]
The addition section 23 adds the restored predicted error data input from the
30 inverse orthogonal transform section 22 and the predicted image data input from the
mode selection section 50 to thereby generate decoded image data. Then, the
addition section 23 outputs the generated decoded image data to the deblocking filter
24a and the frame memory 25.
[003 11
A deblocking filter 24a performs filtering processes to decrease block
5 distortion that occurs during image encoding. For example, the deblocking filter
24a determines necessity of filtering for each block boundary of decoded image data
supplied from an addition section 23 and applies the deblocking filter to a boundary
that is determined to require the filter. The deblocking filter 24a is also supplied
with information used for the determination of filtering necessity (e.g., mode
10 information, transform coefficient information, and motion vector information) as
well as decoded image data from the addition section 23. After the filtering, the
block distortion is eliminated from the decoded image data and the deblocking filter
24a outputs the decoded image data to frame memory 25. The process for the
deblocking filter 24a will be described in detail later.
15 [0032]
The frame memory 25 stores, using a storage medium, the decoded image
data input from the addition section 23 and the decoded image data after filtering
input from the deblocking filter 24a.
[0033]
20 The selector 26 reads, from the frame memory 25, the decoded image data
before filtering that is to be used for the intra prediction, and supplies the decoded
image data which has been read to the intra prediction section 30 as reference image
data. Also, the selector 26 reads, from the frame memory 25, the decoded image
data after filtering to be used for the inter prediction, and supplies the decoded image
25 data which has been read to the motion estimation section 40 as reference image data.
[0034]
The intra prediction section 30 performs an intra prediction process in each
intra prediction mode, based on the image data to be encoded that is input from the
reordering buffer 12 and the decoded image data supplied via the selector 26. For
30 example, the intra prediction section 30 evaluates the prediction result of each intra
prediction mode using a predetermined cost function. Then, the intra prediction
section 30 selects an intra prediction mode by which the cost function value is the
smallest, that is, an intra prediction mode by which the compression ratio is the
highest, as the optimal intra prediction mode. Furthermore, the intra prediction
section 30 outputs, to the mode selection section 50, prediction mode information
5 indicating the optimal intra prediction mode, the predicted image data, and the
information about intra prediction such as the cost function value.
[003 51
A motion estimation section 40 performs an inter prediction process
(prediction process between frames) based on image data for encoding supplied from
10 a reordering buffer 12 and decoded image data supplied via a selector 26. For
example, the motion estimation section 40 evaluates the prediction result of each
prediction mode using a predetermined cost function. Then, the motion estimation
section 40 selects an optimal prediction mode, namely, a prediction mode that
minimizes the cost function value or maximizes the compression ratio. The motion
15 estimation section 40 generates predicted image data according to the optimal
prediction mode. The motion estimation section 40 outputs information about the
inter prediction such as prediction mode information indicating the optimal intra
prediction mode, the predicted image data, and the cost function value to a mode
selection section 50.
20 [0036]
The mode selection section 50 compares the cost function value related to
the intra prediction input from the intra prediction section 30 and the cost function
value related to the inter prediction input from the motion estimation section 40.
Then, the mode selection section 50 selects a prediction method with a smaller cost
25 function value, from the intra prediction and the inter prediction. In the case of
selecting the intra prediction, the mode selection section 50 outputs the information
about intra prediction to the lossless encoding section 16, and also, outputs the
predicted image data to the subtraction section 13 and the addition section 23. Also,
in the case of selecting the inter prediction, the mode selection section 50 outputs the
30 information about inter prediction described above to the lossless encoding section
16, and also, outputs the predicted image data to the subtraction section 13 and the
addition section 23.
[0037]
[I-2. Image Decoding Device]
Fig. 2 is a block diagram showing an example of a configuration of an
5 image decoding device 60 according to an embodiment. With reference to Fig. 2,
the image decoding device 60 includes an accumulation buffer 61, a lossless
decoding section 62, an inverse quantization section 63, an inverse orthogonal
transform section 64, an addition section 65, a deblocking filter 24b, a reordering
buffer 67, a DIA (Digital to Analogue) conversion section 68, a frame memory 69,
10 selectors 70 and 71, an intra prediction section 80, and a motion compensation
section 90.
[003 81
The accumulation buffer 61 temporarily stores an encoded stream input via
a transmission line using a storage medium.
15 [0039]
The lossless decoding section 62 decodes an encoded stream input from the
accumulation buffer 61 according to the encoding method used at the time of
encoding. Also, the lossless decoding section 62 decodes information multiplexed
to the header region of the encoded stream. Information that is multiplexed to the
20 header region of the encoded stream may include information about intra prediction
and information about inter prediction in the block header, for example. The
lossless decoding section 62 outputs the information about intra prediction to the
intra prediction section 80. Also, the lossless decoding section 62 outputs the
information about inter prediction to the motion compensation section 90.
25 [0040]
The inverse quantization section 63 inversely quantizes quantized data
which has been decoded by the lossless decoding section 62. The inverse
orthogonal transform section 64 generates predicted error data by performing inverse
orthogonal transformation on transform coefficient data input from the inverse
30 quantization section 63 according to the orthogonal transformation method used at
the time of encoding. Then, the inverse orthogonal transform section 64 outputs the
generated predicted error data to the addition section 65.
[004 11
The addition section 65 adds the predicted error data input from the inverse
orthogonal transform section 64 and predicted image data input from the selector 71
5 to thereby generate decoded image data. Then, the addition section 65 outputs the
generated decoded image data to the deblocking filter 24b and the frame memory 69.
[0042]
The deblocking filter 24b performs filtering processes to decrease block
distortion appearing on a decoded image. The deblocking filter 24b determines the
10 necessity of filtering at each block boundary for decoded image data input from the
addition section 65, for example, and applies the deblocking filter to a boundary that
is determined to require the filter. The deblocking filter 24b is also supplied with
information used for the determination of filtering necessity as well as decoded
image data from the addition section 65. After the filtering, the block distortion is
15 eliminated from the decoded image data and the deblocking filter 24b outputs the
decoded image data to the reordering buffer 67 and the frame memory 69. The
process for the deblocking filter 24b will be described in detail later.
[0043]
The reordering buffer 67 generates a series of image data in a time sequence
20 by reordering images input from the deblocking filter 24b. Then, the reordering
buffer 67 outputs the generated image data to the D/A conversion section 68.
[0044]
The D/A conversion section 68 converts the image data in a digital format
input from the reordering buffer 67 into an image signal in an analogue format.
25 Then, the D/A conversion section 68 causes an image to be displayed by outputting
the analogue image signal to a display (not shown) connected to the image decoding
device 60, for example.
[0045]
The frame memory 69 uses a storage medium to store the decoded image
30 data input from the addition section 65 before filtering and the decoded image data
input from the deblocking filter 24b after filtering.
[0046]
The selector 70 switches the output destination of the image data from the
frame memory 69 between the intra prediction section 80 and the motion
compensation section 90 for each block in the image according to mode information
5 acquired by the lossless decoding section 62. For example, in the case the intra
prediction mode is specified, the selector 70 outputs the decoded image data before
filtering that is supplied from the frame memory 69 to the intra prediction section 80
as reference image data. Also, in the case the inter prediction mode is specified, the
selector 70 outputs the decoded image data after filtering that is supplied from the
10 frame memory 69 to the motion compensation section 90 as the reference image data.
[0047]
The selector 71 switches the output source of predicted image data to be
supplied to the addition section 65 between the intra prediction section 80 and the
motion compensation section 90 for each block in the image according to the mode
15 information acquired by the lossless decoding section 62. For example, in the case
the intra prediction mode is specified, the selector 71 supplies to the addition section
65 the predicted image data output from the intra prediction section 80. In the case
the inter prediction mode is specified, the selector 71 supplies to the addition section
65 the predicted image data output from the motion compensation section 90.
20 [0048]
The intra prediction section 80 performs in-screen prediction of a pixel
value based on the information about intra prediction input from the lossless
decoding section 62 and the reference image data from the frame memory 69, and
generates predicted image data. Then, the intra prediction section 80 outputs the
25 generated predicted image data to the selector 71.
[0049]
The motion compensation section 90 performs a motion compensation
process based on the information about inter prediction input from the lossless
decoding section 62 and the reference image data from the frame memory 69, and
30 generates predicted image data. Then, the motion compensation section 90 outputs
the generated predicted image data to the selector 71.
[0050]
<2. Existing Technique>
[2-1. Basic Configuration of Deblocking Filter]
Generally, processes using the deblocking filter in an existing image
5 encoding scheme such as H.2641AVC or HEVC include two types of processes,
namely, filtering need determination processes and filtering processes. The
following describes these two processes in HEVC, for example.
[005 11
(1) Filtering Need Determination Processes
10 The filtering need determination processes determine whether the
deblocking filter needs to be applied to each boundary of blocks within an input
image. Block boundaries include a vertical boundary between blocks horizontally
adjacent to each other and a horizontal boundary between blocks vertically adjacent
to each other. JCTVC-A119 uses a block size of 8x8 pixels as a minimum
15 processing unit. For example, a macro block of 16x 16 pixels includes four blocks
of 8x8 pixels. The process is applied to one (left) vertical boundary and one (top)
horizontal boundary for each block, namely, four boundaries plus four boundaries
equal to eight boundaries in total. The specification assumes that the macro block
as a technical term includes an coding unit (CU) in the context of HEVC.
20 [0052]
Fig. 3 is an explanatory diagram showing an example of pixels in two
blocks (neighboring blocks) Ba and Bb adjacent to each other around a boundary.
The following describes the vertical boundary as an example and the description is
obviously applicable to the horizontal boundary. The example in Fig. 3 uses
25 symbol p, to represent a pixel in block Ba. In this symbol, i denotes a column
index and j denotes a row index. The column index i is numbered as 0, 1,2, and 3
in order (from right to left) from the column nearest to the vertical boundary. The
row index j is numbered as 0, 1, 2, . . . , 7 from the top to the bottom. The left half
of block Ba is omitted from the drawing. Symbol qk, is used to represent a pixel in
30 block Bb. In this symbol, k denotes a column index and j denotes a row index.
The column index k is numbered as 0, 1, 2, and 3 in order (from left to right) from
the column nearest to the vertical boundary. The right half of block Bb is omitted
from the drawing.
[0053]
The following conditions can be used to determine the necessity of applying
5 the deblocking filter to the vertical boundary between blocks Ba and Bb shown in
Fig. 3.
[0054]
Determination condition of luma component (Luma) ... The deblocking
filter is applied if conditions A and B are both true.
10 Condition A:
(Al) Block Ba or Bb enters the intra prediction mode;
(A2) Block Ba or Bb has a nonzero orthogonal transform coefficient; or
(A3) IMVAx-MVBxl24 or IMVAy-MVBylT4
Condition B:
15 ~p22-2p12+po2~+~q22-2q12+qo2~p2~-2p1~+po~~+~q2s-~q1~+qo~~<~
[0055]
Condition A3 assumes a motion vector for block Ba to be (MVAx,MVAy)
and a motion vector for block Bb to be (MVBx,MVBy) according to the Qpel (114
pixel) accuracy. Condition B uses P as an edge determination threshold value. An
20 initial value of P is given according to a quantization parameter. The value for P is
user-specifiable using a parameter within the slice header.
[0056]
Determination condition of chroma component (Chroma) ... The deblocking
filter is applied if condition A1 is true.
2 5 Condition A1 : Block Ba or Bb enters the intra prediction mode.
[0057]
As indicated by broken-line frames L3 and L6 in Fig. 4, the filtering need
determination processes on general vertical boundaries (particularly under
determination condition B of luma component) reference pixels on the third and sixth
30 rows (assuming the top row to be the first) in each block. Similarly, the filtering
need determination processes on horizontal boundaries reference pixels (not shown
in Fig. 4) on the third and sixth columns in each block. The above-described
determination conditions are used to determine that the deblocking filter needs to be
applied to a boundary on which the filtering processes described below are
performed.
5 [0058]
(2) Filtering Processes
If it is determined that the deblocking filter needs to be applied to a
boundary, the filtering processes are performed on pixels to the right and the left of
the vertical boundary and on pixels above and below the horizontal boundary. On
10 luma components, the filter strength is switched between a strong filter and a weak
filter according to pixel values.
[0059]
Filtering luma components
Selecting the strength ... The filter strength is selected for each row or
15 column. The strong filter is selected if all of the following conditions C1 through
C3 are satisfied. The weak filter is selected if even any one of the conditions is not
satisfied.
(C l)d<(P>>2)
20 (C3)lpoj-qoj1<((5tc+1)>>1)
where j denotes a row index for the vertical boundary or a column index for
the horizontal boundary. d=lp22-2~12+~021+1q22-2q12+qo21+IP2~-2~1~+~o~l+lq2~-
2qi5+qo51
[0060]
25 Weak filtering
A=Clip(-tc,tc,(l 3(qoj-poj)+4(qlj-plj)- 5(q.~j-p2~)+16)>>5>)
~0j=Cli~0-255@0j+A)
qoj=Clip~-255(q0j-A)
30 q lj'Clip~-255(qlj-A/2)
[006 11
Strong filtering
~0~'C1ip0-2~~(@2~+2p~~+2p0j+2q0~+qlj+4)>>3)
qoj=c1ip0-255(@1 j+2~0j+2qoj+2lqj+ q2j+4)>>3)
p lj'clip~-255(@2j+~~j+p~j+q~j+2)>>2)
5 lj=c1ip~-255(@~j+q~j+q~j+q2j+2)>>2)
p2j'Cl ip0-2~5((2~3~+3~I j2+jp+0pj+ qoj+4)>>3)
q2j'C lipo-255(@oj+qoj+ql j+3qzj+2q3j+4)>>3)
[0062]
where Clip(a,b,c) denotes a process to clip value c within the range of alclb
10 and C l i ~ ~ - ~de~n~ote(sc a) p rocess to clip value c within the range of 05~1255.
[0063]
Filtering chroma components
A=clip(-tc,t~,((((qoj-~oj)<<2)+p lj-qlj+4)>>3))
~0j=Clip0-255@0j+A)
15 qoj=clip0-255(qoj-A)
[0064]
As indicated by broken-line frames C6 through C8 and C1 through C3 in
Fig. 5, the filtering processes (particularly strong filtering on luma components) on
general vertical boundaries update pixel values on the first through third and sixth
20 through eighth columns in each block. Similarly, the filtering processes on
horizontal boundaries update pixel values on the first through third and sixth through
eighth rows in each block.
[0065]
[2-2. Dependency Between Processes According to an Existing Technique]
2 5 For the purpose of description, as shown in Fig. 6, macro block MBx (MBO,
MB 1 ...) each having the size of 16x 16 pixels includes the top left vertical boundary
represented as Vx,O, the top center vertical boundary represented as Vx,l, the bottom
left vertical boundary represented as Vx,2, the bottom center vertical boundary
represented as Vx,3, the top left horizontal boundary represented as Hx,O, the top
30 right horizontal boundary represented as Hx,l, the left center horizontal boundary
represented as Hx,2, and the right center horizontal boundary represented as Hx,3.
Concerning boundary Z, for example, the filtering need determination process is
represented as JZ and the filtering process is represented as Fz.
[0066]
The above-described existing technique causes no dependency between
5 processes on boundaries in the same direction within one macro block. Therefore,
the technique can perform parallel filtering on vertical boundaries and horizontal
boundaries within one macro block, for example. As an example, Fig. 7 makes it
clear that there is no dependency among four filtering processes Fvo,o, Fvo,~F,V 0,2,
and FV0,3 (no pixel updated redundantly) within macro block MBO and the filtering
10 processes can be performed in parallel.
[0067]
However, the above-described existing technique leaves the dependency
between the filtering processes on vertical boundaries and the filtering need
determination processes on horizontal boundaries. The existing technique also
15 leaves the dependency between the filtering processes on horizontal boundaries and
the filtering need determination processes on vertical boundaries. If a vertical
boundary is processed prior to a horizontal boundary, for example, the filtering need
determination processes need to be performed on horizontal boundaries within a
given macro block after termination of the filtering processes on vertical boundaries.
20 As an example, Fig. 8 shows that, within macro block MBO, filtering need
determination process JHO,0d epends on results of filtering processes Fvo,o and Fvo,~
and filtering need determination process JHO,d~e pends on a result of filtering
processes Fvo,~. Similarly, the filtering need determination processes need to be
performed on vertical boundaries within a given macro block after termination of the
25 filtering process on the horizontal boundary for the adjacent macro block. As an
example, Fig. 9 shows that filtering need determination process Jvl,~fo r macro block
MBl depends on results of filtering processes FHO,i and FH0,3 for macro block MBO
and filtering need determination process JV1,2 for macro block MB1 depends on a
result of filtering process FHO,f~o r macro block MBO.
30 [0068]
The existing technique involves the dependency between processes and
therefore provides parallel processing of the deblocking filter to a very limited extent
even if the technique proposed in JCTVC-A119 is used.
[0069]
Fig. 10 is an explanatory diagram illustrating a sequence of deblocking filter
5 processes according to an existing technique. The example assumes that the
deblocking filter is supplied with an image having the size of 32x32 pixels. The
input image includes four macro blocks MBO through MB3 each having the size of
16x 16 pixels.
[0070]
10 In Fig. 10, each broken-line frame represents a process to be performed in
parallel. For example, the first step performs, in parallel, filtering need
determination processes JVO,~Jv, o,,, JV0,2 and Jvo,3 on four vertical boundaries in
macro block MBO. The second step performs, in parallel, filtering processes Fvo,~,
Fvo,,, FVo,2 and Fvoz on four vertical boundaries in macro block MBO. After
15 termination of the second step, the third step performs, in parallel, filtering need
determination processes JHO,OJH, O,lJ, H0,2 and JHOo,~n four horizontal boundaries in
macro block MBO. The third step uses a pixel value after the filtering process on
the vertical boundary at the second step for the filtering need determination process
on the horizontal boundary. The fourth step performs, in parallel, filtering
20 processes FHO,0F, HO,,F, H0,2a nd FH0,3o n four horizontal boundaries in macro block
MBO. After termination of the fourth step, processes (fifth to eighth steps) for
macro block MBl are performed successively. The fifth step uses a pixel value
after the filtering process on the horizontal boundary of the macro block MBO at the
fourth step for the filtering need determination process on the vertical boundary of
25 the macro block MBl . After termination of the process on the macro block MBl,
processes (ninth to twelfth steps) for macro block MB2 are performed successively.
After termination of the process on the macro block MB2, processes (thirteenth to
sixteenth steps) for macro block MB3 are performed successively.
[007 11
30 Such parallel processing within the limited extent cannot satisfactorily solve
the problem of delay or data rate degradation due to a large processing amount when
the deblocking filter is applied. Three working examples described below further
improve parallel processing when the definition is applied.
[0072]
<3. First Working Example>
5 [3-1. Deblocking Filter Configuration Example]
The following describes example configurations of the deblocking filter 24a
for the image encoding device 10 shown in Fig. 1 and the deblocking filter 24b for
the image decoding device 60 shown in Fig. 2 according to the first working example.
The configurations of the deblocking filter 24a and the deblocking filter 24b may be
10 common to each other. In the following description, the deblocking filter 24a and
the deblocking filter 24b are generically referred to as a deblocking filter 24 when
there is no need for distinction between them.
[0073]
(1) Dependency between new processes
15 According to the working example, processes using the deblocking filter 24
also include two types of processes, namely, a filtering need determination process
and a filtering process. The deblocking filter 24 uses pixel values for an image
input to the deblocking filter for the determination across a plurality of macro blocks
in the filtering need determination process on one of the vertical boundary and the
20 horizontal boundary. If the vertical boundary is processed prior to the horizontal
boundary, for example, the deblocking filter 24 can perform the filtering need
determination process on the vertical boundary for a given block without waiting for
the filtering process on the horizontal boundary for the neighboring blocks. If the
horizontal boundary is processed prior to the vertical boundary, for example, the
25 deblocking filter 24 can perform the filtering need determination process on the
horizontal boundary for a given block without waiting for the filtering process on the
horizontal boundary for the neighboring blocks. The result is to relieve the
dependency of processes between macro blocks.
[0074]
30 Relieving the dependency of processes between macro blocks can
parallelize processes between the plurality of macro blocks within an image. For
example, this enables to perform filtering need determination processes in parallel on
vertical boundaries for all blocks within an input image. This also enables to
perform filtering need determination processes in parallel on horizontal boundaries
for all blocks within an input image.
5 [0075]
Fig. 11 is an explanatory diagram illustrating a sequence of processes
available for the working example. The example also assumes that the deblocking
filter is supplied with an image having the size of 32x32 pixels. The input image
includes four macro blocks MBO through MB3 each having the size of 16x 16 pixels.
10 [0076]
In Fig. 11, each broken-line frame represents a process to be performed in
parallel. While the example in Fig. 10 requires 16 process steps for a sequence of
processes, the example in Fig. 11 aggregates the same number of processes into four
process steps. The first step performs, in parallel, filtering need determination
15 processes Jvo,o through JV3,3 a nd J H ~th,ro~u gh JH3,3o n all vertical boundaries and all
horizontal boundaries of all macro blocks MBO through MB3. The second step
performs, in parallel, filtering processes Fvo,t~h rough FV3,3o n 16 vertical boundaries
of all macro blocks MBO through MB3. The third step performs, in parallel,
filtering need determination processes FHO,0 through FH3,~o n all horizontal
20 boundaries of all macro blocks MBO through MB3. The fourth step performs, in
parallel, filtering processes FHO,0 through FH3,3 on 16 horizontal boundaries of all
macro blocks MBO through MB3. The third and fourth steps may precede the first
and second steps if the horizontal boundary is processed prior to the vertical
boundary.
25 [0077]
Fig. 11 provides the example of maximizing the parallelism (processes
performed in parallel) by parallelizing processes over all macro blocks in an image.
While not limited to the example, processes may be parallelized over some macro
blocks instead of all macro blocks in an image.
30 [0078]
(2) Detailed configuration of deblocking filter
Fig. 12 is a block diagram illustrating a detailed configuration of the
deblocking filter 24 according to the first working example for performing the
above-described parallel processes. With reference to Fig. 12, the deblocking filter
24 includes a vertical determination block 110, a horizontal determination block 114,
5 a horizontal filtering block 130, a vertical filtering block 140, and a parallelization
control section 150.
[0079]
(2- 1) Vertical determination block
The vertical determination block 110 includes a plurality of vertical
10 boundary determination sections 1 12- 1 through 1 12-n. Each vertical boundary
determination section 112 is supplied with images input to the deblocking filter 24
and determination information used to determine whether filtering is needed.
[OOSO]
The vertical boundary determination sections 1 12- 1 through 11 2-n
15 determine whether to apply the deblocking filter to vertical boundaries using pixel
values for an image input to the deblocking filter 24 across the plurality of macro
blocks within the image. Each vertical boundary determination section 112 supplies
the horizontal filtering block 130 with information indicating a determination result
about each vertical boundary such as binary information indicating a determination
20 result that value " 1 " forces application of the deblocking filter.
[OOS 11
(2-2) Horizontal filtering block
The horizontal filtering block 130 includes a plurality of horizontal filtering
sections 132-1 through 132-n. Each horizontal filtering section 132 is supplied with
25 an input image and the determination result about each vertical boundary from the
vertical determination block 110.
[0082]
A determination result from the corresponding vertical boundary
determination section 112 may indicate that the filter needs to be applied. In such a
30 case, each horizontal filtering section 132 applies the deblocking filter for vertical
boundary to right and left elements with reference to the vertical boundary. Each
horizontal filtering section 132 supplies the horizontal determination block 11 4 and
the vertical filtering block 140 with pixel values after filtering for filter-applied
pixels and pixel values of the input image for the other pixels.
[0083]
5 (2-3) Horizontal determination block
The horizontal determination block 114 includes a plurality of horizontal
boundary determination sections 11 6-1 through 1 16-n. Each horizontal boundary
determination section 116 is supplied with pixel values after the filtering performed
by the horizontal filtering block 130 and the determination information used to
10 determine whether filtering is needed.
[0084]
The horizontal boundary determination section 1 16- 1 through 1 16-n
determine whether to apply the deblocking filter to horizontal boundaries using pixel
values after the filtering performed by the horizontal filtering block 130 across the
15 plurality of macro blocks within the image. Each horizontal boundary
determination section 116 supplies the vertical filtering block 140 with information
indicating a determination result about each horizontal boundary.
[0085]
(2-4) Vertical filtering block
20 The vertical filtering block 140 includes a plurality of vertical filtering
sections 142-1 through 142-n. Each vertical filtering section 142 is supplied with
pixel values after the filtering performed by the horizontal filtering block 130 and a
determination result about each horizontal boundary from the horizontal
determination block 1 14.
25 [0086]
A determination result from the corresponding horizontal boundary
determination section 11 6 may indicate that the filter needs to be applied. In such a
case, each vertical filtering section 142 applies the deblocking filter for horizontal
boundary to top and bottom elements with reference to the horizontal boundary.
30 Each vertical filtering section 142 supplies filter-applied pixels with pixel values
after the filtering and the other pixels with pixel values supplied from the horizontal
filtering block 130. An output from each vertical filtering section 142 may
configure an output image from the deblocking filter 24.
[0087]
(3) More detailed configuration of the determination section
5 Fig. 13 is a block diagram illustrating a detailed configuration of each of the
vertical boundary determination sections 112 and the horizontal boundary
determination sections 116. With reference to Fig. 13, each determination section
includes a tap constitution section 121, a calculation section 122, a threshold
comparison section 123, a distortion evaluation section 124, and a filtering
10 determination section 125.
[0088]
The tap constitution section 121 acquires a reference pixel value from pixel
values of two blocks neighboring across a focused boundary in the input image and
constitutes a tap (a set of reference pixel values) for determining determination
15 condition B for the above-described luma component. For example, a vertical
boundary may be focused in the blocks each of which has the size of 8x8 pixels. In
this case, the tap constitution section 121 constitutes a tap from pixel values
belonging to the third and sixth rows of two blocks at the right and left. If a
horizontal boundary is focused, the tap constitution section 121 constitutes a tap from
20 pixel values belonging to the third and sixth columns of two blocks at the top and
bottom. The calculation section 122 assigns the tap constituted by the tap
constitution section 121 to the left-hand side of the determination expression in
determination condition B and calculates an edge value to be compared with edge
determination threshold value P. The threshold comparison section 123 compares
25 the value calculated by the calculation section 122 with edge determination threshold
value p and outputs a comparison result to the filtering determination section 125.
[0089]
The distortion evaluation section 124 evaluates determination condition A of
the above-described luma component using mode information (MB mode), transform
30 coefficient information, and motion vector information supplied as the determination
information. The distortion evaluation section 124 outputs an evaluation result to
the filtering determination section 125. The distortion evaluation section 124
evaluates only determination condition A1 of a chroma component based on the
mode information.
[0090]
5 The filtering determination section 125 determines whether to apply the
deblocking filter to a focused boundary based on a comparison result of
determination condition B supplied from the threshold comparison section 123 and
an evaluation result of determination condition A supplied from the distortion
evaluation section 124. The filtering determination section 125 outputs information
10 indicating the determination result.
[0091]
(4) Parallelization control section
The parallelization control section 150 shown in Fig. 12 controls the
parallelism of filtering need determination processes in the vertical determination
15 block 110 and the horizontal determination block 114, and the parallelism of filtering
processes in the horizontal filtering block 130 and the vertical filtering block 140.
[0092]
For example, the parallelization control section 150 may control the
parallelism of processes for each block based on an input image size. More
20 specifically, the parallelization control section 150 increases the parallelism of
processes for each block if the input image size is relatively large. This can
adaptively prevent delay or data rate degradation due to a processing amount that
increases according to image sizes. For example, the parallelization control section
150 may control the parallelism of processes for each block based on a sequence
25 parameter set, a picture parameter set, or parameters contained in the slice header.
This enables to flexibly configure the parallelism according to requirements of users
who develop apparatuses. For example the parallelism may be configured
according to restrictions on the installation environment such as the number of
processor cores or the number of software threads.
30 [0093]
The working example can parallelize processes between macro blocks.
This signifies that any sequence of processes on blocks within an image has no effect
on a finally output result. Accordingly, the parallelization control section 150 can
control a sequence of filtering need determination processes in the vertical
determination block 110 and the horizontal determination block 114, and a sequence
5 of filtering processes in the horizontal filtering block 130 and the vertical filtering
block 140 on a block basis.
[0094]
More specifically, the parallelization control section 150 may control a
sequence of filtering processes according to the dependency of the filtering processes
10 between macro blocks. According to an existing technique, for example, the
dependency of processes between neighboring macro blocks around a slice boundary
may delay parallel processes on each slice within an image. However, the
parallelization control section 150 according to the working example can perform
filtering processes on neighboring macro blocks around the slice boundary prior to
15 the other macro blocks.
[0095]
For example, Fig. 14 illustrates eight macro blocks MBlO through MB13
and MB20 through MB23 around a slice boundary. Macro blocks MBlO through
MB 13 belong to slice SL 1. Macro blocks MB20 through MB23 belong to slice SL2.
20 Concerning these macro blocks, the filtering processes for horizontal boundaries on
macro block MB20 in slice SL2 depend on the filtering processes for vertical
boundaries on macro block MB12 in slice SLl. Similarly, the filtering processes
for horizontal boundaries on macro block MB21 in slice SL2 depend on the filtering
processes for vertical boundaries on macro block MB 13 in slice SL1.
25 [0096]
According to an example in Fig. 15 under these conditions, the
parallelization control section 150 performs filtering processes on the vertical
boundaries of macro blocks MB 12 and MB 13 out of filtering processes for slice SL1
in preference to processes on the other boundaries. The result is to prevent a large
30 delay from occurring in filtering processes on the horizontal boundaries of macro
blocks MB20 and MB21 out of filtering processes for slice SL2. An example in Fig.
16 initially performs filtering processes in parallel on vertical boundaries for all
macro blocks included in slice SLl. Also in this case, no delay occurs in the
filtering process on the horizontal boundaries of macro blocks MB20 and MB21 in
slice SL2.
5 [0097]
[3-2.D etermination Condition Modifications]
As described above, each vertical boundary determination section 112
references pixels corresponding to the third and sixth rows in a block and determines
for vertical boundaries of each block whether filtering is needed similarly to the
10 existing technique as illustrated in Fig. 4. Likewise, each horizontal boundary
determination section 116 references pixels corresponding to the third and sixth
columns in a block and determines for horizontal boundaries of each block whether
filtering is needed. In such a case, the configuration according to the working
example can be easily embodied without changing determination conditions for the
15 filtering need determination process provided for the existing apparatus.
[0098]
Each vertical boundary determination section 112 and each horizontal
boundary determination section 116 may perform the determination using
determination conditions different from the existing technique. For example, each
20 vertical boundary determination section 112 may reference pixels corresponding to
three or more columns in a block. Each horizontal boundary determination section
116 may reference pixels corresponding to three or more columns in a block. In
addition, each vertical boundary determination section 112 and each horizontal
boundary determination section 116 may use determination condition expressions
25 different from the existing technique. With reference to Figs. 17 through 19, the
following describes six examples of the determination technique according to the
working example.
[0099]
(1) First example
30 Fig. 17 is an explanatory diagram illustrating first and second examples of
the determination technique. In the first and second examples, the filtering need
determination processes (particularly the determination using determination
condition B for luma components) for vertical boundaries references pixels of all
rows L1 through L8 from the first to the eighth in each block. The filtering need
determination processes for horizontal boundaries also references pixels of all
5 columns from the first to the eighth in each block.
[O 1001
The first example may define determination conditions for luma
components as follows.
Determination condition of luma component (Luma) ... The deblocking
10 filter is applied if conditions A and B are both true.
Condition A:
(Al) Block Ba or Bb enters the intra prediction mode;
(A2) Block Ba or Bb has a nonzero orthogonal transform coefficient; or
(A3) IMVAx-MVBxl24 or IMVAy-MVByl24
15 Condition B:
iD0=lp20-2p10+p00l+lq20-2q~0+q001+IP27-2p1+p0l+lq-q17+qo71
iD1=(~21-2p11+~(+0(1q 21-2ql1+ q01(+(~26-2~16+~06(+lq26-2q16+q06(
iD2=1p22-2p12+p021+lq22-2q12+q021+IP25-2~1+pol+lq-q15+qo~1
iD~=lp2~-2p13+p03l+lq23-2q~3+q03l+l~24-2~14+~~l+lq24-2q~4+qo4l
20 iDave=(iDo+iD +iD2+iD3)>>2
Under this condition, iDave

1>
Under this condition, iDave

1>
Under this condition, iDaVe

[4- 1. Deblocking Filter Configuration Example]
The following describes example configurations of the deblocking filter 24
according to the second working example.
[0 1 241
5 (1) Dependency between new processes
According to the working example, the deblocking filter 24 performs the
filtering need determination process on vertical boundaries of each block without
waiting for application of the deblocking filter to the other blocks in the macro block
to which the block belongs. The deblocking filter 24 performs the filtering need
10 determination process on horizontal boundaries of each block without waiting for
application of the deblocking filter to the other blocks in the macro block to which
the block belongs. This can relieve the dependency of processes within a macro
block.
[0 1251
15 As described above, relieving the dependency of processes can
consequently parallelize filtering need determination processes on vertical
boundaries and horizontal boundaries in a macro block.
[0 1261
Fig. 22 is an explanatory diagram illustrating a first example of process
20 sequence available on the deblocking filter 24. The example also assumes that the
deblocking filter is supplied with an image having the size of 32x32 pixels. The
input image includes four macro blocks MBO through MB3 each having the size of
16x 16 pixels.
[0 1271
2 5 In Fig. 22, each broken-line frame represents a process to be performed in
parallel. While the example in Fig. 10 requires 16 process steps for a sequence of
processes, the example in Fig. 22 aggregates the same number of processes into 12
process steps. The first step performs, in parallel, filtering need determination
processes Jvo,t~h rough JV0,3a nd JHO,0th rough JH0,3o n four vertical boundaries and
30 four horizontal boundaries of macro block MBO. The second step performs, in
parallel, filtering processes Fvo,o through Fvoz on four vertical boundaries in macro
block MBO. The third step performs, in parallel, filtering need determination
processes JVI,O through JV1,3 and JHIp through JH1,3 on four vertical boundaries and
four horizontal boundaries of macro block MBl. The fourth step performs, in
parallel, filtering processes FvlPo through FV1,3 on four vertical boundaries in macro
5 block MBl. The fifih step performs, in parallel, filtering processes FHO,0 through
FH0,3 on four horizontal boundaries in macro block MBO. The sixth step performs,
in parallel, filtering need determination processes JV2,0 through JV2,3 and Jm,0 through
Jm,3 on four vertical boundaries and four horizontal boundaries of macro block MB2.
The seventh step performs, in parallel, filtering processes FV2,o through FV2,3 on four
10 vertical boundaries in macro block MB2. The eighth step performs, in parallel,
filtering processes FHI,0 through FH1,3 on four horizontal boundaries in macro block
MB 1. The ninth step performs, in parallel, filtering need determination processes
JV3,0 through Jv3,3 and JH3,O through JH3,3 on four vertical boundaries and four
horizontal boundaries of macro block MB3. The tenth step performs, in parallel,
15 filtering processes FV3,0 through FV3,3 on four vertical boundaries in macro block
MB3. The eleventh step performs, in parallel, filtering processes F~2,0th rough Fm,3
on four horizontal boundaries in macro block MB2. The twelfih step performs, in
parallel, filtering processes FH3,0 through FH3,3 on four horizontal boundaries in
macro block MB3. In this case, the deblocking filter 24 can perform a process on
20 the entire input image using process steps fewer than those of the existing technique.
[0 1281
(2) Detailed configuration of deblocking filter
Fig. 23 is a block diagram illustrating a detailed configuration of the
deblocking filter 24 according to the second working example for performing the
25 above-described parallel processes. With reference to Fig. 23, the deblocking filter
24 includes a vertical determination block 2 10, a horizontal determination block 2 14,
the horizontal filtering block 130, the vertical filtering block 140, and the
parallelization control section 150.
[0 1291
3 0 (2- 1) Vertical determination block
The vertical determination block 210 includes a plurality of vertical
boundary determination sections 2 12- 1 through 2 12-n. Each vertical boundary
determination section 212 determines whether to apply the deblocking filter to
vertical boundaries of each block without waiting for application of the deblocking
filter to the other blocks in the macro block to which the block belongs. Each
5 vertical boundary determination section 212 supplies the horizontal filtering block
130 with information indicating a determination result about each vertical boundary
such as binary information indicating a determination result that value "1" forces
application of the deblocking filter.
[0130]
10 (2-2) Horizontal determination block
The horizontal determination block 214 includes a plurality of horizontal
boundary determination sections 2 16- 1 through 2 16-n. Each horizontal boundary
determination section 216 determines whether to apply the deblocking filter to
horizontal boundaries of each block without waiting for application of the deblocking
15 filter to the other blocks in the macro block to which the block belongs. Each
horizontal boundary determination section 2 16 supplies the vertical filtering block
140 with information indicating a determination result about each horizontal
boundary.
[0131]
20 Also according to the working example, each vertical boundary
determination section 2 12 and each horizontal boundary determination section 2 16
may determine for each boundary whether filtering is needed by referencing pixels at
positions similarly to the existing technique. Instead, each vertical boundary
determination section 2 12 and each horizontal boundary determination section 2 16
25 may determine for each boundary whether filtering is needed according to the
technique described in "3-2. Determination Condition Modifications."
[0132]
[3-2. Process Flow]
Fig. 24 is a flowchart illustrating a process flow for the deblocking filter 24
30 according to the second working example. With reference to Fig. 24, the vertical
boundary determination sections 2 12- 1 through 2 12-n determine in parallel for all
vertical boundaries included in a focused macro blocks within an input image
whether filtering is needed (step S202). The horizontal boundary determination
sections 214-1 through 214-n determine in parallel whether filtering is needed for all
horizontal boundaries included in the focused macro block(step S204). Steps S202
5 and S204 are also performed in parallel.
[0133]
The horizontal filtering sections 132-1 through 132-n apply the deblocking
filter in parallel to vertical boundaries in the focused macro block determined at step
S202 to require the deblocking filter to be applied (step S210).
10 [0134]
The process at step S220 aims at a focused macro block in the most recent
loop. The process at step S220 may be skipped for the first focused macro block.
The vertical filtering sections 142-1 through 142-n apply the deblocking filter in
parallel to horizontal boundaries determined, at step S204 in the most recent loop, to
15 require the deblocking filter to be applied (step S220).
[0135]
The process at steps S202 through S220 is repeated for a newly focused
macro block if focused macro blocks remain unprocessed in the input image (step
S230).
20 [0136]
If there remains no focused macro block unprocessed, the vertical filtering
sections 142-1 through 142-n apply the deblocking filter in parallel to horizontal
boundaries determined to require the deblocking filter to be applied in the focused
macro block for the last loop (step S240).
25 [0137]
The flow of processes described above is also a mere example. The
parallelism and sequence of processes may be changed. Further, the parallelization
control section 150 may control the parallelism and sequence of processes.
[0138]
30 [4-3. Process Example for Each LCU]
As already mentioned, the technology according to various working
examples described in this specification may be provided as a process based on an
HEVC coding unit (CU). According to HEVC, a coding unit having the largest size
is referred to as a largest coding unit (LCU) that can be selected as 64x64 pixels, for
example. The minimum selectable CU size is 8x8 pixels. Normally, an image is
5 encoded and decoded corresponding to each LCU in accordance with a raster scan
sequence from the LCU at the top left of a picture (or a slice). The following
describes process examples corresponding to LCUs in the deblocking filter 24.
[0139]
Fig. 25 is an explanatory diagram illustrating a process sequence for each
10 LCU according to the second working example described above. The example
assumes the LCU size to be 16x 16 pixels and the CU size to be 8x8 pixels.
[0 1 401
With reference to Fig. 25, the first stage is shown at the top left of the
drawing and indicates that the filtering on LCUs has completed up to the (n-1)th
15 LCU. Shaded pixels are targeted for filtering on vertical boundaries. Filled pixels
are targeted for filtering on horizontal boundaries.
[0141]
In Fig. 25, the second process at the top right and the third process at the
bottom left are targeted for the nth LCU. Prior to the second process, filtering need
20 determination processes are performed in parallel on all vertical boundaries and
horizontal boundaries belonging to the nth LCU. Namely, the filtering need
determination process on boundaries belonging to CUs in the nth LCU is performed
without waiting for application of the deblocking filter to the other CUs in the nth
LCU. The second process performs filtering processes in parallel on vertical
25 boundaries that belong to the nth LCU and are determined to require the deblocking
filter to be applied. The second process performs filtering processes in parallel on
horizontal boundaries that belong to the nth LCU and are determined to require the
deblocking filter to be applied.
[0 1421
30 A process for the fourth stage at the bottom right of Fig. 25 is targeted for
the (n+l)th LCU. At the fourth stage, the filtering process is performed in parallel
on a vertical boundary determined to require the deblocking filter to be applied after
the filtering need determination processes are performed in parallel on boundaries
belonging to all CUs in the (n+l)th LCU.
[0 1431
5 While the example assumes the LCU size to be 16x 16 pixels, it may be set
to 32x32 or 64x64 pixels. The effect of shortening the processing time according to
the parallelization is further improved because increasing the size of an LCU to be
selected also increases the number of vertical boundaries and horizontal boundaries
belonging to one LCU.
10 [0144]
Fig. 26 is a flowchart illustrating a process flow of the deblocking filter 24
for each LCU.
[0 1451
With reference to Fig. 26, the vertical boundary determination sections 212-
15 1 through 212-n determine in parallel whether filtering is needed for all vertical
boundaries included in a focused LCU within an input image (step S252). The
horizontal boundary determination sections 2 16- 1 through 2 16-n determine in
parallel whether filtering is needed for all horizontal boundaries included in the
focused LCU (step S254). Steps S252 and S254 are also performed in parallel.
20 [0146]
The horizontal filtering sections 132- 1 through 132-n apply the deblocking
filter in parallel to vertical boundaries in the focused LCU determined at step S252 to
require the deblocking filter to be applied (step S260).
[0147]
2 5 The vertical filtering sections 142-1 through 142-n apply the deblocking
filter in parallel to horizontal boundaries in the focused LCU determined at step S254
to require the deblocking filter to be applied (step S270).
[0 1481
The process at steps S252 through S270 is repeated for a newly focused
30 LCU if an LCU remains unprocessed in the input image (step S280). The process
terminates if there remains no LCU unprocessed.
[0 1 491
<5. Third Working Example>
[5-1. Overview]
The first and second working examples change the existing sequence of
5 processes for the deblocking filter to improve the parallelism of processes.
Particularly, the first working example relieves the process dependency by extending
the scope of pixel values for an input image supplied to the deblocking filter when
the necessity of filtering is determined. The third working example to be described
enhances this concept. The third working example further parallelizes processes by
10 filtering input pixel values supplied to the deblocking filter during a filtering process
on vertical boundaries and horizontal boundaries.
[0150]
Fig. 27 is an explanatory diagram illustrating an overview of the working
example. At the bottom left of Fig. 27, there is shown a shape representing input
15 pixels (also referred to as reconstruct pixels) before being processed by the
deblocking filter. The working example allows pixel values input to the deblocking
filter to be referenced from a filtering need determination process for vertical
boundaries and horizontal boundaries as well as a filtering process for vertical
boundaries and a filtering process for horizontal boundaries. Accordingly, this
20 enables to solve the dependency between the two filtering need determination
processes and the dependency between the two filtering processes.
[0151]
The filtering processes for vertical boundaries and the filtering processes for
horizontal boundaries may update values of duplicate pixels. Filled pixels in Fig.
25 27 illustrate positions of the pixels likely to be duplicated. The deblocking filter
according to the working example calculates one output pixel value from two filter
outputs in terms of pixels that are duplicately updated by two filters operating in
parallel.
[0 1521
30 [5-2. Deblocking Filter Configuration Example]
Fig. 28 is a block diagram -illustrating a detailed configuration of the
deblocking filter 24 according to the third working example. With reference to Fig.
28, the deblocking filter 24 includes the line memory 308, the determination block
3 10, a horizontal filtering block 330, a vertical filtering block 340, the parallelization
control section 150, and a calculation section 360.
5 [0153]
The line memory 308 stores pixel values for an input image supplied to the
deblocking filter 24. Filtering processes in the horizontal filtering block 330 and
the vertical filtering block 340 do not update pixel values stored in the line memory
308. Filtering need determination processes performed by sections described below
10 in the determination block 310 reference pixel values stored in the line memory 308.
The apparatus includes another memory for purposes different from processes of the
deblocking filter 24. This memory may be reused (shared) as the line memory 308.
[0 1 541
The determination block 3 10 includes vertical boundary determination
15 sections 3 12- 1 through 3 12-n and horizontal boundary determination sections 3 14- 1
through 314-n. The vertical boundary determination sections 312 and the
horizontal boundary determination sections 3 14 are supplied with pixel values stored
in the line memory 308 for an image input to the deblocking filter 24 and
determination information used to determine the need for filtering.
20 [0155]
The vertical boundary determination sections 3 12 use pixel values input to
the deblocking filter 24 to determine whether to apply the deblocking filter to each
vertical boundary. The vertical boundary determination sections 3 12 output, to the
horizontal filtering block 330, information indicating a determination result about
25 each vertical boundary.
[0156]
The horizontal boundary determination sections 314 also use pixel values
input to the deblocking filter 24 to determine whether to apply the deblocking filter
to each horizontal boundary. The horizontal boundary determination sections 314
30 perform determination processes in parallel to determination processes performed by
the vertical boundary determination sections 312. The horizontal boundary
determination sections 314 output, to the vertical filtering block 340, information
indicating a determination result about each horizontal boundary.
[0157]
Also according to the working example, each vertical boundary
5 determination section 312 and each horizontal boundary determination section 314
may determine for each boundary whether filtering is needed by referencing pixels at
positions similarly to the existing technique. Instead, each vertical boundary
determination section 3 12 and each horizontal boundary determination section 3 14
may determine for each boundary whether filtering is needed according to the
10 technique described in "3-2. Determination Condition Modifications."
[0158]
The horizontal filtering block 330 includes horizontal filtering sections 332-
1 through 332-n. The horizontal filtering sections 332 are supplied with an input
image value from the line memory 208 and a determination result concerning each
15 vertical boundary from the determination block 3 10.
[0159]
The horizontal filtering sections 332 apply the deblocking filter for vertical
boundaries to right and left pixels around the corresponding vertical boundary if the
determination result from the vertical boundary determination section 3 12 indicates
20 that the filter needs to be applied. The horizontal filtering sections 332 output, to
the calculation section 360, a pixel value after the filtering in terms of the filtered
pixel or an input pixel value in terms of the other pixels.
[0 1601
The vertical filtering block 340 includes vertical filtering sections 342-1
25 through 342-n. The vertical filtering sections 342 are supplied with an input pixel
value from the line memory 308 and a determination result concerning each
horizontal boundary from the determination block 3 10.
[0161]
The vertical filtering sections 342 apply the deblocking filter for horizontal
30 boundaries to top and bottom pixels around the corresponding horizontal boundary if
the determination result from the horizontal boundary determination section 314
indicates that the filter needs to be applied. Filtering processes of the vertical
filtering sections 342-1 through 342-n are performed in parallel to filtering processes
of the horizontal filtering sections 332-1 through 332-n. The vertical filtering
sections 342 output, to the calculation section 360, a pixel value after the filtering in
5 terms of the filtered pixel or an input pixel value in terms of the other pixels.
[0 1 621
The calculation section 360 is supplied with an output pixel value from the
horizontal filtering block 330 and an output pixel value from the vertical filtering
block 340 in parallel. Further, the calculation section 360 is supplied with
10 determination results from the vertical boundary determination section 3 12 and the
horizontal boundary determination section 3 14. According to a determination result,
the calculation section 360 calculates output pixel values for pixels filtered from the
horizontal filtering block 330 and the vertical filtering block 340 based on filter
outputs from the horizontal filtering block 330 and the vertical filtering block 340.
15 [0163]
According to the working example, for example, the calculation section 360
calculates an average of two filter outputs for duplicately filtered pixels. The
calculation section 360 may calculate a simple average of two filter outputs.
Instead, the calculation section 360 may calculate a weighted average of two filter
20 outputs. For example, the calculation section 360 may determine a weight for
weighted averages of pixels according to the distance from each pixel to the vertical
boundary and to the horizontal boundary.
[0 1 641
Fig. 29 is an explanatory diagram illustrating determination of a weight for
25 weighted average calculated by the calculation section 360. Fig. 29 shows focused
pixel PZ in black corresponding to one of the duplicated positions illustrated in Fig.
27. There are three pixels corresponding to distance Dv between focused pixel PZ
and nearest vertical boundary VZ. There are two pixels corresponding to distance
DH between focused pixel Pz and nearest horizontal boundary Hz. Distance DH is
30 smaller than distance Dv. In this case, the calculation section 360 may set a weight
for output from the deblocking filter applied to horizontal boundary Hz to be larger
than a weight for output from the deblocking filter applied to vertical boundary VZ.
The example in Fig. 29 assumes that a ratio of filter output Vout for vertical boundary
Vz to filter output Hout for horizontal boundary HZ is 2:3.
[0 1651
5 As seen from Fig. 29, calculating a weighted average of two filter outputs
can consequently provide each focused pixel with an output pixel value similar to the
case of applying one two-dimensional filter having a filter tap along the horizontal
direction and a filter tap along the vertical direction. Parallelizing filtering
processes on the vertical boundary and the horizontal boundary can also
10 appropriately reduce block distortion appearing on the vertical boundary and the
horizontal boundary. As another working example, the deblocking filter 24 may
include one two-dimensional filter that simultaneously calculates horizontal filtering,
vertical filtering, and a weighted average. In this case, however, the installation is
very complicated because filter coefficients need to be variously changed
15 correspondingly to pixels. On the other hand, the third working example performs
two one-dimensional filters in parallel and then calculates a weighted average. This
can easily provide processes substantially equal to a two-dimensional filter while
ensuring the functionality of existing deblocking filters.
[0 1661
20 Fig. 30 is an explanatory diagram illustrating an example of the weight for
weighted averages calculated based on the example in Fig. 29. Fig. 30 shows 36
pixels (6x6) around an intersection between the vertical boundary and the horizontal
boundary. The pixels correspond to the above-described duplicated positions. The
ratio of the weight for filter output Vout to the weight for filter output Hout is 1 : 1 (2:2
25 or 3:3) for pixels positioned at an equal distance from the vertical boundary and the
horizontal boundary. The weight for filter output Vout is larger than the weight for
filter output HOut for pixels nearer to the vertical boundary. For example, the ratio
of weights for pixel PI is Vout:Hou=t 3:l. The weight for filter output Vouti s smaller
than the weight for filter output Hout for pixels nearer to the horizontal boundary.
30 For example, the ratio of weights for pixel P2 is Vout:Hou=t 1: 3.
[0 1 671
The block distortion can be more effectively suppressed and the image
quality can be improved by varying the weight for weighted averages depending on
the distance between each pixel and the boundary.
[0168]
5 The above-described weights are mere examples. For example, the
calculation section 360 may determine the weight of weighted averages for pixels
according to the edge strengths of the vertical boundary and the horizontal boundary
corresponding to each pixel instead of or in addition to the distance between each
pixel and the boundary. The edge strength may be represented with a parameter
10 such as an edge value calculated from the calculation section 122 as shown in Fig. 13,
for example. In this case, the weight for filer output on a boundary having a
stronger edge may be set to be larger than the weight for filer output on a boundary
having a weaker edge. Varying the weight of weighted averages according to the
edge strength can adaptively improve the effect of the deblocking filter at a boundary
15 that remarkably causes block distortion.
[0 1691
The calculation section 360 selects outputs from actually filtered blocks in
terms of pixels filtered by one of the horizontal filtering block 330 and the vertical
filtering block 340. The calculation section 360 directly outputs an input pixel
20 value to be output to the deblocking filter 24 in terms of pixels not filtered by the
horizontal filtering block 330 or the vertical filtering block 340. A table in Fig. 3 1
lists output pixel values from the calculation section 360 according to results of the
determination whether to require the filtering.
[0 1 701
2 5 [5-3. Process Sequence Example]
The following describes two examples of process sequences available for
the deblocking filter 24 according to the working example. The example also
assumes that the deblocking filter is supplied with an image having the size of 32x32
pixels. The input image includes four macro blocks MBO through MB3 each
30 having the size of 16x 16 pixels.
[0171]
(1) First example
For comparison, Fig. 32 illustrates a process sequence when the dependency
remains between a filtering process on the vertical boundary and a filtering process
on the horizontal boundary. In Fig. 32, the first step performs, in parallel, filtering
5 need determination processes Jvo,t~h rough JV3,3a nd JHO,0th rough JH3,3o n all vertical
boundaries and all horizontal boundaries of all the four macro blocks MBO through
MB3. The second step performs fiItering processes FvoPo through Fv3,3 on 16
vertical boundaries of the four macro blocks MBO through MB3. The third step
performs filtering processes FHOt,h~r ough F~3,3o n 16 horizontal boundaries of the
10 four macro blocks MBO through MB3. The fourth step stores pixel values after the
filtering process on the horizontal boundary in the memory used for outputs from the
deblocking filter 24.
[0 1 721
Fig. 33 illustrates a first example of process sequence provided by the
15 working example. In Fig. 33, the first step performs, in parallel, filtering need
determination processes Jvo,~th rough JV3,3 and JHO,th~r ough JH3,3 on all vertical
boundaries and all horizontal boundaries of the four macro blocks MBO through
MB3. The second step performs, in parallel, filtering processes Fvo,~th rough Fv3,3
and FHO,0th rough FH3,3o n all vertical boundaries and all horizontal boundaries of the
20 four macro blocks MBO through MB3. Actually, the second step filters only a
boundary determined to require the filtering. The third step stores pixel values in
the memory used for outputs from the deblocking filter 24. A weighted average of
two filter outputs may be calculated as an output pixel value in terms of pixels
filtered by the horizontal filtering block 330 and the vertical filtering block 340.
25 [0173]
(2) Second example
While the first example maximizes the parallelism, the deblocking filter 24
according to the second example can also perform a process for each macro block.
[0 1 741
30 For comparison, Fig. 34 illustrates a process sequence for each macro block
when the dependency remains between a filtering process on the vertical boundary
and a filtering process on the horizontal boundary. The process sequence in Fig. 34
substantially equals the process sequence in Fig. 22 according to the first working
example. Fig. 36 explicitly shows the four process steps (sixth, tenth, 14th, and
16th) that store pixel values in the memory for output and are omitted from Fig. 22
5 for simplicity. Sixteen process steps including the four process steps configure the
process in Fig. 34.
[0 1 751
Fig. 35 illustrates a second example of process sequence provided by the
working example. In Fig. 35, the first step performs, in parallel, filtering need
10 determination processes Jvo,o through JvoJ and JHO,0t hrough JHO,~o n four vertical
boundaries and four horizontal boundaries of macro block MBO. The second step
performs, in parallel, filtering processes Fvo,o through Fvo,a~n d FHOt,h~ro ugh FHO,o~n
four vertical boundaries and four horizontal boundaries of macro block MBO. The
third step stores pixel values of macro block MBO in the memory used for outputs
15 from the deblocking filter 24. A weighted average of two filter outputs may be
calculated as an output pixel value in terms of pixels duplicately filtered by two
filters. The fourth to sixth steps similarly process macro block MB1. The seventh
to ninth steps similarly process macro block MB2. The tenth to twelfth steps
similarly process macro block MB3. The process in Fig. 35 includes twelve process
20 steps fewer than those of the process in Fig. 34.
[0 1761
The third working example eliminates the dependency between filtering
processes for vertical boundaries and a filtering process for horizontal boundaries.
The process of the deblocking filter 24 can be performed using fewer process steps
25 than those used for the first and second working examples. One of advantages of
allowing a filtering process to reference only pixels input to the deblocking filter is
that any configuration of filter taps causes no dependency between filtering processes
for vertical boundaries and filtering processes for horizontal boundaries. The third
working example can improve the image quality by configuring a filter tap using
30 more pixels than used for existing techniques. For example, the existing technique
uses a filter tap of three pixels for each side of each boundary as described with
reference to Fig. 7. The working example causes no dependency between processes
even if a filter tap of five pixels or more is used at each boundary. No dependency
occurs between processes even by further decreasing the block size as a process unit
for the deblocking filter.
5 [0177]
Also in the third working example as well as the first and second working
examples, the parallelization control section 150 may control the parallelism and
sequence of processes in the deblocking filter 24.
[0178]
10 [5-4. Process Flow]
Fig. 36 is a flowchart illustrating an example of a process flow for the
deblocking filter according to the third working example. Fig. 37 is a flowchart
illustrating a flow of the pixel value calculation process shown in Fig. 36.
[0 1 791
15 With reference to Fig. 36, the vertical boundary determination sections 3 12-
1 through 3124 determine in parallel whether filtering is needed for all vertical
boundaries within an input image or a macro block (step S302). The horizontal
boundary determination sections 314-1 through 3 14-n determine in parallel whether
filtering is needed for all horizontal boundaries within the input image or the macro
20 block (step S304). Steps S302 and S304 are also performed in parallel.
[0 1 801
The horizontal filtering sections 332-1 through 332-n apply the deblocking
filter in parallel to all vertical boundaries determined at step S302 to require the
deblocking filter to be applied (step S306). The vertical filtering sections 342-1
25 through 342-n apply the deblocking filter in parallel to all horizontal boundaries
determined at step S304 to require the deblocking filter to be applied (step S308).
Steps S306 and S308 are also performed in parallel.
[0181]
The calculation section 360 then performs the pixel value calculation
30 process as shown in Fig. 37 (step S310). With reference to Fig. 37, the process
from step S3 14 to step S326 loops for each pixel to be processed (step S3 12).
[0 1 821
At step S314, the calculation section 360 determines whether two filters for
vertical boundaries and horizontal boundaries have filtered a focused pixel (step
S3 14). The process proceeds to step S322 if the two filters have filtered the focused
5 pixel. The process proceeds to step S316 if the two filters have not filtered the
focused pixel.
[0 1 831
At step S3 16, the calculation section 360 determines whether one of the two
filters for vertical boundaries and horizontal boundaries has filtered the focused pixel
10 (step S3 16). The process proceeds to step S320 if one of the two filters has filtered
the focused pixel. The process proceeds to step S318 if none of the filters has
filtered the focused pixel.
[0 1 841
At step S318, the calculation section 360 acquires an input pixel value to the
15 deblocking filter 24 (step S318). At step S320, the calculation section 360 acquires
a filter output from the filter that actually filters the focused pixel (step S320).
[0 1 851
At step S322, the calculation section 360 determines weight values for
calculating a weighted average of filter outputs from the two filters concerning the
20 focused pixel according to distances from the focused pixel to the vertical boundary
and the horizontal boundary or the edge strengths of the vertical boundary and the
horizontal boundary corresponding to the focused pixel (step S322). The
calculation section 360 calculates a weighted average of filter outputs from the two
filters using the determined weight (step S324).
25 [0186]
The calculation section 360 stores the pixel value of the focused pixel in the
memory while the pixel value is acquired at step S3 18 or S320 or is calculated at step
S324 (step S326). The sequences of processes as shown in Figs. 36 and 37
terminate when the process is performed on all pixels to be processed.
30 [0187]
<6. Application to Various Codecs>
The technology according to the disclosure is applicable to various codecs
related to image encoding and decoding. The following describes examples of
applying the technology according to the disclosure to multiview codec and scalable
codec.
5 [0188]
[6- 1. Multiview Codec]
The multiview codec is an image encoding scheme that encodes and
decodes multiple-perspective video. Fig. 38 is an explanatory diagram illustrating
the multiview codec. Fig. 38 illustrates sequences of frames for three views
10 captured at three observing points. Each view is provided with a view ID (view-id).
One of the views is specified as a base view. Views other than the base view are
referred to as non-base views. The example in Fig. 38 represents a base view with
view ID "0" and two non-base views with view ID "I" or "2." Encoding multiview
image data may compress the data size of the encoded stream as a whole by encoding
15 frames of the non-base view based on encoding information about frames of the base
view.
[0 1 891
The deblocking filter may be applied to each view during the encoding
process and the decoding process according to the multiview codec described above.
20 Application of the deblocking filter to each view may parallelize horizontal filtering
and vertical filtering in units of processes including multiple CUs for each view
according to the technology of the disclosure. The above-described process unit
may represent several CUs, LCUs, or pictures. The parameter (such as the one
described in the preceding paragraph 0092) to control a parallel process may be set
25 for each view. The parameter set for the base view may be reused for the non-base
view.
[0 1 901
The horizontal filtering and the vertical filtering may be parallelized over a
plurality of views. The plurality of views may share the parameter (such as the one
30 described in the preceding paragraph 0092) to control the parallel process. It may
be advantageous to additionally specify a flag indicating whether the plurality of
views share the parameter.
[0191]
Fig. 39 is an explanatory diagram illustrating the image encoding process
applied to the multiview codec described above. Fig. 39 shows a configuration of a
5 multiview encoding device 71 0 as an example. The multiview encoding device 71 0
includes a first encoding section 720, a second encoding section 730, and a
multiplexing section 740.
[0 1 921
The first encoding section 720 encodes a base view image and generates an
10 encoded stream for the base view. The second encoding section 730 encodes a nonbase
view image and generates an encoded stream for the non-base view. The
multiplexing section 740 multiplexes an encoded stream for the base view generated
from the first encoding section 720 and one or more encoded streams for the nonbase
view generated from the second encoding section 730 to generate a multiplexed
15 stream for multiview.
[0193]
The first encoding section 720 and the second encoding section 730
illustrated in Fig. 39 are configured similarly to the image encoding device 10
according to the above-described embodiment. Applying the deblocking filter to
20 views enables to parallelize the horizontal filtering and the vertical filtering in units
of processes containing multiple CUs. A parameter to control these processes may
be inserted into a header area of the encoded stream for each view or into a common
header area in the multiplexed stream.
[0 1 941
2 5 Fig. 40 is an explanatory diagram illustrating an image decoding process
applied to the multiview codec described above. Fig. 40 shows a configuration of a
multiview decoding device 760 as an example. The multiview decoding device 760
includes a demultiplexing section 770, a first decoding section 780, and a second
decoding section 790.
30 [0195]
The demultiplexing section 770 demultiplexes a multiplexed stream for -
multiview into an encoded stream for the base view and an encoded stream for one or
more non-base views. The first decoding section 780 decodes a base view image
from an encoded stream for the base view. The second decoding section 730
decodes a non-base view image from an encoded stream for the non-base view.
5 [0196]
The first decoding section 780 and the second decoding section 790
illustrated in Fig. 40 are configured similarly to the image decoding device 60
according to the above-described embodiment. Applying the deblocking filter to
views enables to parallelize horizontal filtering in units of processes including
10 multiple CUs and parallelize vertical filtering. A parameter to control these
processes may be acquired from a header area of the encoded stream for each view or
from a common header area in the multiplexed stream.
[0 1971
[6-2. Scalable Codec]
15 The scalable codec is an image encoding scheme to provide hierarchical
encoding. Fig. 41 is an explanatory diagram illustrating the scalable codec. Fig.
41 illustrates frame sequences for three layers of different space resolutions, time
resolutions, or image qualities. Each layer is provided with a layer ID (layer-id).
These layers include a base layer having the lowest resolution (or image quality).
20 Layers other than the base layer are referred to as enhancement layers. The
example in Fig. 41 represents a base layer with layer ID "0" and two enhancement
layers with layer ID "1" or "2." Encoding multi-layer image data may compress the
data size of the encoded stream as a whole by encoding frames of the enhancement
layer based on encoding information about frames of the base layer.
25 [0198]
The deblocking filter may be applied to each layer during the encoding
process and the decoding process according to the scalable codec described above.
Application of the deblocking filter to each layer may parallelize horizontal filtering
and vertical filtering in units of processes including multiple CUs for each view
30 according to the technology of the disclosure. The above-described process unit
may represent several CUs, LCUs, or pictures. The parameter (such as the one
described in the preceding paragraph 0092) to control a parallel process may be set
for each layer. The parameter set for the base layer view may be reused for the
enhancement layer.
[0 1 991
5 The horizontal filtering and the vertical filtering may be parallelized over a
plurality of layers. The plurality of layers may share the parameter (such as the one
described in the preceding paragraph 0092) to control the parallel process. It may
be advantageous to additionally specify a flag indicating whether the plurality of
layers share the parameter.
10 [0200]
Fig. 42 is an explanatory diagram illustrating the image encoding process
applied to the scalable codec described above. Fig. 42 shows a configuration of a
scalable encoding device 810 as an example. The scalable encoding device 810
includes a first encoding section 820, a second encoding section 830, and a
15 multiplexing section 840.
[020 11
The first encoding section 820 encodes a base layer image and generates an
encoded stream for the base layer. The second encoding section 830 encodes an
enhancement layer image and generates an encoded stream for the enhancement layer.
20 The multiplexing section 840 multiplexes an encoded stream for the base layer
generated from the first encoding section 820 and one or more encoded streams for
the enhancement layer generated from the second encoding section 830 to generate a
multiplexed stream for multi-layer.
[0202]
2 5 The first encoding section 820 and the second encoding section 830
illustrated in Fig. 42 are configured similarly to the image encoding device 10
according to the above-described embodiment. Applying the deblocking filter to
layers enables to parallelize horizontal filtering in units of processes including
multiple CUs and parallelize vertical filtering. A parameter to control these
30 processes may be inserted into a header area of the encoded stream for each layer or
into a common header area in the multiplexed stream.
[0203]
Fig. 43 is an explanatory diagram illustrating an image decoding process
applied to the scalable codec described above. Fig. 43 shows a configuration of a
scalable decoding device 860 as an example. The scalable decoding device 860
5 includes a demultiplexing section 870, a first decoding section 880, and a second
decoding section 890.
[0204]
The demultiplexing section 870 demultiplexes a multiplexed stream for
multi-layer into an encoded stream for the base layer and an encoded stream for one
10 or more enhancement layers. The first decoding section 880 decodes a base layer
image from an encoded stream for the base layer. The second decoding section 830
decodes an enhancement layer image from an encoded stream for the enhancement
layer.
[0205]
15 The first decoding section 880 and the second decoding section 890
illustrated in Fig. 43 are configured similarly to the image decoding device 60
according to the above-described embodiment. Applying the deblocking filter to
layers enables to parallelize horizontal filtering in units of processes including
multiple CUs and parallelize vertical filtering. A parameter to control these
20 processes may be acquired from a header area of the encoded stream for each layer
or from a common header area in the multiplexed stream.
102061
<7. Example Application>
The image encoding device 10 and the image decoding device 60 according
25 to the embodiment described above may be applied to various electronic appliances
such as a transmitter and a receiver for satellite broadcasting, cable broadcasting such
as cable TV, distribution on the Internet, distribution to terminals via cellular
communication, and the like, a recording device that records images in a medium
such as an optical disc, a magnetic disk or a flash memory, a reproduction device that
30 reproduces images from such storage medium, and the like. Four example
applications will be described below.
[0207]
[7-1. First Example Application]
Fig. 44 is a block diagram showing an example of a schematic configuration
of a television adopting the embodiment described above. A television 900 includes
5 an antenna 901, a tuner 902, a demultiplexer 903, a decoder 904, an video signal
processing section 905, a display section 906, an audio signal processing section 907,
a speaker 908, an external interface 909, a control section 9 10, a user interface 9 11,
and a bus 912.
[0208]
10 The tuner 902 extracts a signal of a desired channel from broadcast signals
received via the antenna 901, and demodulates the extracted signal. Then, the tuner
902 outputs an encoded bit stream obtained by demodulation to the demultiplexer
903. That is, the tuner 902 serves as transmission means of the televisions 900 for
receiving an encoded stream in which an image is encoded.
15 [0209]
The demultiplexer 903 separates a video stream and an audio stream of a
program to be viewed from the encoded bit stream, and outputs each stream which
has been separated to the decoder 904. Also, the demultiplexer 903 extracts
auxiliary data such as an EPG (Electronic Program Guide) from the encoded bit
20 stream, and supplies the extracted data to the control section 910. Additionally, the
demultiplexer 903 may perform descrambling in the case the encoded bit stream is
scrambled.
[02 1 01
The decoder 904 decodes the video stream and the audio stream input from
25 the demultiplexer 903. Then, the decoder 904 outputs video data generated by the
decoding process to the video signal processing section 905. Also, the decoder 904
outputs the audio data generated by the decoding process to the audio signal
processing section 907.
[0211]
30 The video signal processing section 905 reproduces the video data input
from the decoder 904, and causes the display section 906 to display the video. The
video signal processing section 905 may also cause the display section 906 to display
an application screen supplied via a network. Further, the video signal processing
section 905 may perform an additional process such as noise removal, for example,
on the video data according to the setting. Furthermore, the video signal processing
5 section 905 may generate an image of a GUI (Graphical User Interface) such as a
menu, a button, a cursor or the like, for example, and superimpose the generated
image on an output image.
[02 121
The display section 906 is driven by a drive signal supplied by the video
10 signal processing section 905, and displays a video or an image on an video screen of
a display device (for example, a liquid crystal display, a plasma display, an OLED, or
the like).
102 1 31
The audio signal processing section 907 performs reproduction processes
15 such as D/A conversion and amplification on the audio data input from the decoder
904, and outputs audio from the speaker 908. Also, the audio signal processing
section 907 may perform an additional process such as noise removal on the audio
data.
[02 141
20 The external interface 909 is an interface for connecting the television 900
and an external appliance or a network. For example, a video stream or an audio
stream received via the external interface 909 may be decoded by the decoder 904.
That is, the external interface 909 also serves as transmission means of the
televisions 900 for receiving an encoded stream in which an image is encoded.
25 [0215]
The control section 910 includes a processor such as a CPU (Central
Processing Unit), and a memory such as an RAM (Random Access Memory), an
ROM (Read Only Memory), or the like. The memory stores a program to be
executed by the CPU, program data, EPG data, data acquired via a network, and the
30 like. The program stored in the memory is read and executed by the CPU at the
time of activation of the television 900, for example. The CPU controls the
operation of the television 900 according to an operation signal input from the user
interface 91 1, for example, by executing the program.
[02 1 61
The user interface 911 is connected to the control section 910. The user
5 interface 91 1 includes a button and a switch used by a user to operate the television
900, and a receiving section for a remote control signal, for example. The user
interface 91 1 detects an operation of a user via these structural elements, generates an
operation signal, and outputs the generated operation signal to the control section 910.
[02 1 71
10 The bus 912 interconnects the tuner 902, the demultiplexer 903, the decoder
904, the video signal processing section 905, the audio signal processing section 907,
the external interface 909, and the control section 91 0.
[02 1 81
In the television 900 configured in this manner, the decoder 904 has a
15 function of the image decoding device 60 according to the embodiment described
above. Accordingly, also in the case of the image decoding in the television 900, it
is possible to enhance the parallelism of deblocking filter processes and ensure highspeed
processing.
[02 1 91
20 [7-2. Second Example Application]
Fig. 45 is a block diagram showing an example of a schematic configuration
of a mobile phone adopting the embodiment described above. A mobile phone 920
includes an antenna 921, a communication section 922, an audio codec 923, a
speaker 924, a microphone 925, a camera section 926, an image processing section
25 927, a demultiplexing section 928, a recording/reproduction section 929, a display
section 930, a control section 93 1, an operation section 932, and a bus 933.
[0220]
The antenna 921 is connected to the communication section 922. The
speaker 924 and the microphone 925 are connected to the audio codec 923. The
30 operation section 932 is connected to the control section 931. The bus 933
- interconnects the communication section 922, the audio codec 923, the camera
section 926, the image processing section 927, the demultiplexing section 928, the
recording/reproduction section 929, the display section 930, and the control section
93 1.
[022 11
5 The mobile phone 920 performs operation such as transmission/reception of
audio signal, transmission/reception of emails or image data, image capturing,
recording of data, and the like, in various operation modes including an audio
communication mode, a data communication mode, an image capturing mode, and a
videophone mode.
10 [0222]
In the audio communication mode, an analogue audio signal generated by
the microphone 925 is supplied to the audio codec 923. The audio codec 923
converts the analogue audio signal into audio data, and A/D converts and compresses
the converted audio data. Then, the audio codec 923 outputs the compressed audio
15 data to the communication section 922. The communication section 922 encodes
and modulates the audio data, and generates a transmission signal. Then, the
communication section 922 transmits the generated transmission signal to a base
station (not shown) via the antenna 921. Also, the communication section 922
amplifies a wireless signal received via the antenna 921 and converts the frequency
20 of the wireless signal, and acquires a received signal. Then, the communication
section 922 demodulates and decodes the received signal and generates audio data,
and outputs the generated audio data to the audio codec 923. The audio codec 923
extends and D/A converts the audio data, and generates an analogue audio signal.
Then, the audio codec 923 supplies the generated audio signal to the speaker 924 and
25 causes the audio to be output.
[0223]
Also, in the data communication mode, the control section 931 generates
text data that makes up an email, according to an operation of a user via the operation
section 932, for example. Moreover, the control section 931 causes the text to be
30 displayed on the display section 930. Furthermore, the control section 931
generates email data according to a transmission instruction of the user via the
operation section 932, and outputs the generated email data to the communication
section 922. Then, the communication section 922 encodes and modulates the
email data, and generates a transmission signal. Then, the communication section
922 transmits the generated transmission signal to a base station (not shown) via the
5 antenna 921. Also, the communication section 922 amplifies a wireless signal
received via the antenna 921 and converts the frequency of the wireless signal, and
acquires a received signal. Then, the communication section 922 demodulates and
decodes the received signal, restores the email data, and outputs the restored email
data to the control section 93 1. The control section 931 causes the display section
10 930 to display the contents of the email, and also, causes the email data to be stored
in the storage medium of the recording/reproduction section 929.
[0224]
The recording/reproduction section 929 includes an arbitrary readable and
writable storage medium. For example, the storage medium may be a built-in
15 storage medium such as an RAM, a flash memory or the like, or an externally
mounted storage medium such as a hard disk, a magnetic disk, a magneto-optical
disk, an optical disc, an USB memory, a memory card, or the like.
[0225]
Furthermore, in the image capturing mode, the camera section 926 captures
20 an image of a subject, generates image data, and outputs the generated image data to
the image processing section 927, for example. The image processing section 927
encodes the image data input from the camera section 926, and causes the encoded
stream to be stored in the storage medium of the recording/reproduction section 929.
[0226]
2 5 Furthermore, in the videophone mode, the demultiplexing section 928
multiplexes a video stream encoded by the image processing section 927 and an
audio stream input from the audio codec 923, and outputs the multiplexed stream to
the communication section 922, for example. The communication section 922
encodes and modulates the stream, and generates a transmission signal. Then, the
30 communication section 922 transmits the generated transmission signal to a base
station (not shown) via the antenna 921. Also, the communication section 922
amplifies a wireless signal received via the antenna 921 and converts the frequency
of the wireless signal, and acquires a received signal. These transmission signal and
received signal may include an encoded bit stream. Then, the communication
section 922 demodulates and decodes the received signal, restores the stream, and
5 outputs the restored stream to the demultiplexing section 928. The demultiplexing
section 928 separates a video stream and an audio stream from the input stream, and
outputs the video stream to the image processing section 927 and the audio stream to
the audio codec 923. The image processing section 927 decodes the video stream,
and generates video data. The video data is supplied to the display section 930, and
10 a series of images is displayed by the display section 930. The audio codec 923
extends and D/A converts the audio stream, and generates an analogue audio signal.
Then, the audio codec 923 supplies the generated audio signal to the speaker 924 and
causes the audio to be output.
[0227]
15 In the mobile phone 920 configured in this manner, the image processing
section 927 has a function of the image encoding device 10 and the image decoding
device 60 according to the embodiment described above. Accordingly, also in the
case of encoding and decoding an image in the mobile phone 920, it is possible to
enhance the parallelism of deblocking filter processes and ensure high-speed
20 processing.
[022 81
[7-3. Third Example Application]
Fig. 46 is a block diagram showing an example of a schematic configuration
of a recording/reproduction device adopting the embodiment described above. A
25 recording/reproduction device 940 encodes, and records in a recording medium,
audio data and video data of a received broadcast program, for example. The
recording/reproduction device 940 may also encode, and record in the recording
medium, audio data and video data acquired from another device, for example.
Furthermore, the recording/reproduction device 940 reproduces, using a monitor or a
30 speaker, data recorded in the recording medium, according to an instruction of a user,
for example. At this time, the recording/reproduction device 940 decodes the audio
data and the video data.
lo2291
The recording/reproduction device 940 includes a tuner 941, an external
interface 942, an encoder 943, an HDD (Hard Disk Drive) 944, a disc drive 945, a
5 selector 946, a decoder 947, an OSD (On-Screen Display) 948, a control section 949,
and a user interface 950.
[0230]
The tuner 941 extracts a signal of a desired channel from broadcast signals
received via an antenna (not shown), and demodulates the extracted signal. Then,
10 the tuner 941 outputs an encoded bit stream obtained by demodulation to the selector
946. That is, the tuner 941 serves as transmission means of the
recording/reproduction device 940.
[023 11
The external interface 942 is an interface for connecting the
15 recording/reproduction device 940 and an external appliance or a network. For
example, the external interface 942 may be an IEEE 1394 interface, a network
interface, an USB interface, a flash memory interface, or the like. For example,
video data and audio data received by the external interface 942 are input to the
encoder 943. That is, the external interface 942 serves as transmission means of the
20 recording/reproduction device 940.
[0232]
In the case the video data and the audio data input from the external
interface 942 are not encoded, the encoder 943 encodes the video data and the audio
data. Then, the encoder 943 outputs the encoded bit stream to the selector 946.
25 [0233]
The HDD 944 records in an internal hard disk an encoded bit stream, which
is compressed content data of a video or audio, various programs, and other pieces of
data. Also, the HDD 944 reads these pieces of data from the hard disk at the time
of reproducing a video or audio.
30 [0234]
The disc drive 945 records or reads data in a recording medium that is
mounted. A recording medium that is mounted on the disc drive 945 may be a
DVD disc (a DVD-Video, a DVD-RAM, a DVD-R, a DVD-RW, a DVD+, a
DVD+RW, or the like), a Blu-ray (registered trademark) disc, or the like, for example.
[0235]
5 The selector 946 selects, at the time of recording a video or audio, an
encoded bit stream input from the tuner 941 or the encoder 943, and outputs the
selected encoded bit stream to the HDD 944 or the disc drive 945. Also, the
selector 946 outputs, at the time of reproducing a video or audio, an encoded bit
stream input from the HDD 944 or the disc drive 945 to the decoder 947.
10 [0236]
The decoder 947 decodes the encoded bit stream, and generates video data
and audio data. Then, the decoder 947 outputs the generated video data to the OSD
948. Also, the decoder 904 outputs the generated audio data to an external speaker.
[0237]
15 The OSD 948 reproduces the video data input from the decoder 947, and
displays a video. Also, the OSD 948 may superimpose an image of a GUI, such as
a menu, a button, a cursor or the like, for example, on a displayed video.
[023 81
The control section 949 includes a processor such as a CPU, and a memory
20 such as an RAM or an ROM. The memory stores a program to be executed by the
CPU, program data, and the like. A program stored in the memory is read and
executed by the CPU at the time of activation of the recording/reproduction device
940, for example. The CPU controls the operation of the recording/reproduction
device 940 according to an operation signal input from the user interface 950, for
25 example, by executing the program.
[0239]
The user interface 950 is connected to the control section 949. The user
interface 950 includes a button and a switch used by a user to operate the
recording/reproduction device 940, and a receiving section for a remote control
30 signal, for example. The user interface 950 detects an operation of a user via these
structural elements, generates an operation signal, and outputs the generated
operation signal to the control section 949.
[0240]
In the recording/reproduction device 940 configured in this manner, the
encoder 943 has a function of the image encoding device 10 according to the
5 embodiment described above. Also, the decoder 947 has a function of the image
decoding device 60 according to the embodiment described above. Accordingly,
also in the case of encoding and decoding an image in the recording/reproduction
device 940, it is possible to enhance the parallelism of deblocking filter processes
and ensure high-speed processing.
10 [0241]
[7-4. Fourth Example Application]
Fig. 47 is a block diagram showing an example of a schematic configuration
of an image capturing device adopting the embodiment described above. An image
capturing device 960 captures an image of a subject, generates an image, encodes the
15 image data, and records the image data in a recording medium.
102421
The image capturing device 960 includes an optical block 961, an image
capturing section 962, a signal processing section 963, an image processing section
964, a display section 965, an external interface 966, a memory 967, a media drive
20 968, an OSD 969, a control section 970, a user interface 971, and a bus 972.
[0243]
The optical block 961 is connected to the image capturing section 962.
The image capturing section 962 is connected to the signal processing section 963.
The display section 965 is connected to the image processing section 964. The user
25 interface 971 is connected to the control section 970. The bus 972 interconnects the
image processing section 964, the external interface 966, the memory 967, the media
drive 968, the OSD 969, and the control section 970.
[0244]
The optical block 961 includes a focus lens, an aperture stop mechanism,
30 and the like. The optical block 961 forms an optical image of a subject on an image
capturing surface of the image capturing section 962. Tbe image capturing section
962 includes an image sensor such as a CCD, a CMOS or the like, and converts by
photoelectric conversion the optical image formed on the image capturing surface
into an image signal which is an electrical signal. Then, the image capturing
section 962 outputs the image signal to the signal processing section 963.
5 [0245]
The signal processing section 963 performs various camera signal processes,
such as knee correction, gamma correction, color correction and the like, on the
image signal input from the image capturing section 962. The signal processing
section 963 outputs the image data after the camera signal process to the image
10 processing section 964.
[0246]
The image processing section 964 encodes the image data input from the
signal processing section 963, and generates encoded data. Then, the image
processing section 964 outputs the generated encoded data to the external interface
15 966 or the media drive 968. Also, the image processing section 964 decodes
encoded data input from the external interface 966 or the media drive 968, and
generates image data. Then, the image processing section 964 outputs the
generated image data to the display section 965. Also, the image processing section
964 may output the image data input from the signal processing section 963 to the
20 display section 965, and cause the image to be displayed. Furthermore, the image
processing section 964 may superimpose data for display acquired from the OSD 969
on an image to be output to the display section 965.
[0247]
The OSD 969 generates an image of a GUI, such as a menu, a button, a
25 cursor or the like, for example, and outputs the generated image to the image
processing section 964.
[0248]
The external interface 966 is configured as an USB inputloutput terminal,
for example. The external interface 966 connects the image capturing device 960
30 and a printer at the time of printing an image, for example. Also, a drive is
connected to the external interface 966 as necessary. A removable medium, such as
a magnetic disk, an optical disc or the like, for example, is mounted on the drive, and
a program read from the removable medium may be installed in the image capturing
device 960. Furthermore, the external interface 966 may be configured as a
network interface to be connected to a network such as a LAN, the Internet or the
5 like. That is, the external interface 966 serves as transmission means of the image
capturing device 960.
[0249]
A recording medium to be mounted on the media drive 968 may be an
arbitrary readable and writable removable medium, such as a magnetic disk, a
10 magneto-optical disk, an optical disc, a semiconductor memory or the like, for
example. Also, a recording medium may be fixedly mounted on the media drive
968, configuring a non-transportable storage section such as a built-in hard disk drive
or an SSD (Solid State Drive), for example.
[02 5 01
15 The control section 970 includes a processor such as a CPU, and a memory
such as an RAM or an ROM. The memory stores a program to be executed by the
CPU, program data, and the like. A program stored in the memory is read and
executed by the CPU at the time of activation of the image capturing device 960, for
example. The CPU controls the operation of the image capturing device 960
20 according to an operation signal input from the user interface 971, for example, by
executing the program.
[025 11
The user interface 971 is connected to the control section 970. The user
interface 971 includes a button, a switch and the like used by a user to operate the
25 image capturing device 960, for example. The user interface 971 detects an
operation of a user via these structural elements, generates an operation signal, and
outputs the generated operation signal to the control section 970.
[02 5 21
In the image capturing device 960 configured in this manner, the image
30 processing section 964 has a function of the image encoding device 10 and the image
decoding device 60 according to the embodiment described above. Accordingly, in
the case of encoding and decoding an image in the image capturing device 960, it is
possible to enhance the parallelism of deblocking filter processes and ensure highspeed
processing.
[0253]
5 <8. Summing-up>
With reference to Figs. 1 through 47, there have been described three
working examples of the deblocking filter for the image encoding device 10 and the
image decoding device 60 according to an embodiment. The three working
examples relieve the dependency of deblocking filter processes inherent to the
10 existing technique. This can improve the parallelism of processes when the
deblocking filter is applied. As a result, it is possible to avoid delay or data rate
degradation due to a large processing amount of the deblocking filter and ensure
high-speed processing. The parallelism and sequences of deblocking filter
processes can be flexibly set according to various conditions such as image sizes or
15 installation environment.
[0254]
According to the first working example, pixel values of an input image
supplied to the deblocking filter are referred to across a plurality of macro blocks
within the image when determining for one of a vertical boundary and a horizontal
20 boundary whether filtering is needed. The result is to relieve the dependency of
processes between macro blocks or coding units. Therefore, it is possible to
parallelize filtering need determination processes across the plurality of macro blocks
or all macro blocks within an image if the parallelism is maximized.
[0255]
2 5 According to the second working example, determinations for vertical
boundaries and horizontal boundaries of each block whether filtering is needed are
made without waiting for applying deblocking filters to the other blocks in the macro
block to which the block belongs. This relieves the dependency of processes
between a vertical boundary and a horizontal boundary in a macro block or a coding
30 unit. Accordingly, it is possible to parallelize filtering need determination processes
on vertical boundaries and horizontal boundaries in a macro block.
[02 5 61
According to the third working example, filtering processes for vertical
boundaries and horizontal boundaries filter pixels input to the deblocking filter.
This configuration can parallelize filtering processes for vertical boundaries and
5 horizontal boundaries each other. This can further accelerate processes of the
deblocking filter. An output pixel value is calculated based on two filter outputs in
terms of a pixel updated by two filtering processes performed in parallel.
Parallelizing two filtering processes can also appropriately reduce block distortion
appearing on the vertical boundary and the horizontal boundary. An output pixel
10 value can be calculated as a weighted average of two filter outputs. This can allow
the deblocking filter to more effectively eliminate the block distortion and further
improve the image quality.
[02 5 71
The specification has mainly described examples where filtering processes
15 for vertical boundaries precedes filtering processes for horizontal boundaries. In
addition, the above-described effects of the technology according to the disclosure
are equally available to a case where filtering processes for horizontal boundaries
precede filtering processes for vertical boundaries. The deblocking filter processing
unit or the macro block may be sized otherwise than described in the specification.
20 An available technique may omit the filtering need determination processes and
parallelize application of the deblocking filter to vertical boundaries and horizontal
boundaries.
[02 5 81
A technique of transmitting information used for deblocking filter process
25 parallelization from the encoding side to the decoding side is not limited to the
technique of multiplexing the information into the encoded stream header. For
example, the information may not be multiplexed into an encoded bit stream but may
be transmitted or recorded as separate data associated with the encoded bit stream.
The term "association" signifies ensuring possibility of linking an image (or part of
30 an image such as a slice or a block) contained in the bit stream with information
corresponding to the image. Namely, the information may be transmitted over a
transmission path different from that used for images (or bit streams). The
information may be recorded on a recording medium (or a different recording area on
the same recording medium) different from that used for images (or bit streams).
The information and the image (or bit stream) may be associated with each other
5 based on any units such as multiple frames, one frame, or part of a frame.
[0259]
The preferred embodiments of the present invention have been described
above with reference to the accompanying drawings, whilst the present invention is
not limited to the above examples, of course. A person skilled in the art may find
10 various alternations and modifications within the scope of the appended claims, and
it should be understood that they will naturally come under the technical scope of the
present invention.
[0260]
The specification represents filtering processes for vertical boundaries as
15 "horizontal filtering" and filtering processes for horizontal boundaries as "vertical
filtering." Generally, filtering processes for vertical boundaries uses horizontally
positioned filter taps. Filtering processes for horizontal boundaries uses vertically
positioned filter taps. For this reason, the above-described nomenclature is used for
the filtering processes.
20
Reference Signs List
102611
10,60 image processing device
1 12,2 12 first determination section (vertical boundary determination section)
25 1 16,2 16 second determination section (horizontal boundary determination section)
132 first filtering section (horizontal filtering section)
142 second filtering section (vertical filtering section)
150 parallelization control section

CLAIMS
Claim 1 (Amended)
An image processing device comprising:
a decoding section configured to decode an encoded stream to generate an
image;
a horizontal filtering section configured to apply a deblocking filter to
neighboring blocks neighboring a vertical block boundary within the image
generated by the decoding section; and
a control section configured to cause the horizontal filtering section to filter
in parallel neighboring blocks neighboring each of a plurality of vertical block
boundaries per a parallel processing unit of the plurality of vertical block boundaries
among a plurality of blocks.
Claim 2 (Amended)
The image processing device according to claim 1,
wherein the control section causes the horizontal filtering section to set the
plurality of vertical block boundaries as the parallel processing unit.
Claim 3 (Amended)
The image processing device according to claim 2,
wherein the control section causes the horizontal filtering section to filter
each vertical block boundary without depending on filtering results for other vertical
block boundaries.
Claim 4 (Amended)
The image processing device according to claim 3,
wherein the control section causes the horizontal filtering section to use a
block size which does not pose a dependency between filterings for the vertical block
boundaries different from each other as a block size of each block.
Claim 5 (Amended)
The image processing device according to claim 4, fkrther comprising
a vertical filtering section configured to apply a deblocking filter to
neighboring blocks neighboring a horizontal block boundary within the image
5 generated by the decoding section,
wherein the control section causes the vertical filtering section to filter in
parallel neighboring blocks neighboring each of a plurality of horizontal block
boundaries per a parallel processing unit of the plurality of horizontal block
boundaries among a plurality of blocks.
10
Claim 6 (Amended)
The image processing device according to claim 5,
wherein the control section causes the vertical filtering section to set the
plurality of horizontal block boundaries as the parallel processing unit.
15
Claim 7 (Amended)
The image processing device according to claim 6,
wherein the control section causes the vertical filtering section to filter each
horizontal block boundary without depending on filtering results for other horizontal
20 block boundaries.
Claim 8 (Amended)
The image processing device according to claim 7,
wherein the control section causes the vertical filtering section to use a
25 block size which does not pose a dependency between filterings for the horizontal
block boundaries different from each other as a block size of each block.
Claim 9 (Amended)
The image processing device according to claim 8,
30 wherein the control section causes the horizontal filtering section and the
vertical filtering section to use 8x8 pixels as the block size.
Claim 10 (Amended)
The image processing device according to claim 9,
wherein the control section causes the horizontal filtering section to filter in
5 parallel the plurality of vertical block boundaries and then causes the vertical
filtering section to filter in parallel the plurality of horizontal block boundaries.
Claim 11 (Amended)
The image processing device according to claim 10,
10 wherein the encoded stream is encoded per a unit of hierarchized block, and
wherein the decoding section decodes the encoded stream per the unit of
hierarchized block.
Claim 12 (Amended)
15 The image processing device according to claim 11, further comprising
a determination section configured to determine in parallel whether to apply
a deblocking filter to neighboring blocks neighboring each of a plurality of vertical
block boundaries per a parallel processing unit of the plurality of vertical block
boundaries among a plurality of blocks,
20 wherein the control section causes the horizontal filtering section to perform
filtering according to a determination result from the determination section.
Claim 13 (Amended)
The image processing device according to claim 12,
2 5 wherein the determination section determines in parallel whether to apply a
deblocking filter to neighboring blocks neighboring each of a plurality of horizontal
block boundaries per a parallel processing unit of the plurality of horizontal block
boundaries among a plurality of blocks, and
wherein the control section causes the vertical filtering section to perform
30 filtering according to a determination result from the determination section.
Claim 14 (Amended)
An image processing method comprising:
decoding an encoded stream to generate an image;
performing horizontal filtering to apply a deblocking filter to neighboring
5 blocks neighboring a vertical block boundary within the generated image; and
controlling the horizontal filtering to filter in parallel neighboring blocks
neighboring each of a plurality of vertical block boundaries per a parallel processing
unit of the plurality of vertical block boundaries among a plurality of blocks.
10 Claim 15 (Amended)
An image processing device comprising:
a horizontal filtering section configured to apply a deblocking filter to
neighboring blocks neighboring a vertical block boundary within a locally decoded
image when encoding an image to be encoded;
15 a control section configured to cause the horizontal filtering section to filter
in parallel neighboring blocks neighboring each of a plurality of vertical block
boundaries per a parallel processing unit of the plurality of vertical block boundaries
among a plurality of blocks; and
an encoding section configured to encode the image to be encoded using an
20 image filtered by the horizontal filtering section.
Claim 16 (Amended)
The image processing device according to claim 15,
wherein the control section causes the horizontal filtering section to set the
25 plurality of vertical block boundaries as the parallel processing unit.
Claim 17 (Amended)
The image processing device according to claim 16,
wherein the control section causes the horizontal filtering section to filter
30 each vertical block boundary without depending on filtering results for other vertical
block boundaries.
Claim 18 (Amended)
The image processing device according to claim 17,
wherein the control section causes the horizontal filtering section to use a
5 block size which does not pose a dependency between filterings for the vertical block
boundaries different from each other as a block size of each block.
Claim 19 (Amended)
The image processing device according to claim 18, further comprising
10 a vertical filtering section configured to apply a deblocking filter to
neighboring blocks neighboring a horizontal block boundary within the image,
wherein the control section causes the vertical filtering section to filter in
parallel neighboring blocks neighboring each of a plurality of horizontal block
boundaries per a parallel processing unit of the plurality of horizontal block
15 boundaries among a plurality of blocks.
Claim 20 (Amended)
The image processing device according to claim 19,
wherein the control section causes the vertical filtering section to set the
20 plurality of horizontal block boundaries as the parallel processing unit.
Claim 21 (Added)
The image processing device according to claim 20,
wherein the control section causes the vertical filtering section to filter each
25 horizontal block boundary without depending on filtering results for other horizontal
block boundaries.
Claim 22 (Added)
The image processing device according to claim 2 1,
30 wherein the control section causes the vertical filtering section to use a
block size which does not pose a dependency between filterings for the horizontal
block boundaries different from each other as a block size of each block.
Claim 23 (Added)
The image processing device according to claim 22,
5 wherein the control section causes the horizontal filtering section and the
vertical filtering section to use 8x8 pixels as the block size.
Claim 24 (Added)
The image processing device according to claim 23,
10 wherein the control section causes the horizontal filtering section to filter in
parallel the plurality of vertical block boundaries and then causes the vertical
filtering section to filter in parallel the plurality of horizontal block boundaries.
Claim 25 (Added)
15 The image processing device according to claim 24,
wherein the encoding section encodes the image to be encoded per a unit of
hierarchized block, and
wherein the image to be encoded is locally decoded per the unit of
hierarchized block.
20
Claim 26 (Added)
The image processing device according to claim 25, further comprising
a determination section configured to determine in parallel whether to apply
a deblocking filter to neighboring blocks neighboring each of a plurality of vertical
25 block boundaries per a parallel processing unit of the plurality of vertical block
boundaries among a plurality of blocks,
wherein the control section causes the horizontal filtering section to perform
filtering according to a determination result from the determination section.
30 Claim 27 (Added)
The image processing device according to claim 26,
wherein the determination section determines in parallel whether to apply a
deblocking filter to neighboring blocks neighboring each of a plurality of horizontal
block boundaries per a parallel processing unit of the plurality of horizontal block
boundaries among a plurality of blocks, and
5 wherein the control section causes the vertical filtering section to perform
filtering according to a determination result fiom the determination section.
Claim 28 (Added)
An image processing method comprising:
10 performing horizontal filtering to apply a deblocking filter to neighboring
blocks neighboring a vertical block boundary within a locally decoded image when
-- encoding an image to be encoded;
controlling the horizontal filtering to filter in parallel neighboring blocks
neighboring each of a plurality of vertical block boundaries per a parallel processing
15 unit of the plurality of vertical block boundaries among a plurality of blocks; and
encoding the image to be encoded using an image filtered by the horizontal
filtering.