Method of decoding video data.
Abstract
A method is provided that derives the index of the reference image and the motion vector from a current prediction unit, generates a prediction block of the current prediction unit using the index of the reference image and the motion vector, generates a residual block by means of inverse scanning, inverse quantization and inverse transformation and generates the reconstructed pixels using the prediction block and the residual block. The prediction pixels of the prediction block are generated using an interpolation filter based on the motion vector. Accordingly, the efficiency of encoding motion information is improved by including several combination candidates. Also, the computational complexity of the encoder and decoder is reduced by selecting different filters according to the location of the prediction pixels determined by the motion vector.

Term
6.1 yearsleft in the term
Expires 2 November 2032.
- Priority
- Filed
- Granted
- Today
- Expires
7 claims: 2 independent, 5 dependent
- 1REIVINDICACIONES IMPI INSTITUTO MEXICANO de la propiedad industrial 1. Un método para decodificación de datos de video en predicción unidireccional mediante un aparato de decodificación, el método comprende:derivar, mediante el aparato de decodificación, un índice de la imagen de referencia y un vector de movimiento de una unidad de predicción actual;generar, mediante el aparato de decodificación, un bloque de predicción de la unidad de predicción actual usando el índice de la imagen de referencia y el vector de movimiento;generar, mediante el aparato de decodificación, un bloque cuantificado mediante la exploración inversa de los componentes del coeficiente cuantificados;generar, mediante el aparato de decodificación, un bloque transformado mediante la cuantificación inversa del bloque cuantificado usando un parámetro de cuantificación;generar, mediante el aparato de decodificación, un bloque residual mediante transformación inversa del bloque transformado;y generar, mediante el aparato de decodificación, pixeles reconstruidos usando el bloque de predicción y el bloque residual, en donde los pixeles de predicción del bloque de predicción se generan usando un filtro de interpolación seleccionado con base en el vector de movimiento, el filtro de interpolación es un filtro asimétrico de 7 derivaciones si el vector de movimiento ' 46 ΙΜΡΙ»5 INSTITUTO MEXICANO DE LA PROPIEDAD indica una posición de un cuarto de pixel, y e 1 tt^-3e interpolación es un filtro simétrico de 8 derivaciones si él vector de movimiento indica una posición de medio pixel, en donde el parámetro de cuantificación se deriva agregando un parámetro de cuantificación diferencial y un predictor del parámetro de cuantificación, y en donde, cuando sólo uno de un parámetro de cuantificación izquierdo y un parámetro de cuantificación anterior está disponible, el predictor del parámetro de cuantificación es un promedio de un parámetro de cuantificación anterior y el disponible del parámetro de cuantificación izquierdo y el parámetro de cuantificación anterior, y en donde, cuando el tamaño del bloque cuantificado es mayor que un tamaño predeterminado, se generan una pluralidad de subconjuntos mediante exploración inversa de los componentes de coeficiente cuantificado y el bloque cuantificado se genera mediante exploración inversa de la pluralidad de subconjuntos usando un patrón de exploración en diagonal.
- 2El método de conformidad con la reivindicación 1, en donde el parámetro de cuantificación se deriva por una unidad de cuantificación, y un tamaño de la unidad de cuantificación es uno de los tamaños permisibles de una unidad de codificación.
- 3El método de conformidad con la reivindicación 1, en donde el número de derivaciones del filtro de IMPIf INSTITUTO MEXICANO | de la monedad. de Indstrbllx por la posición vector de ιτιονί*ΤΤΤ*ττ^τ^·ΒΒ· । ..... conformidad con la reivindicación interpolación se determina predicción indicados por el 4. El método de 1, en donde el índice de la imagen de referencia y el vector de movimiento de la unidad de predicción actual son un índice de la imagen de referencia y un vector de movimiento de un candidato de combinación espacial o temporal especificado por un índice de combinación, y si la unidad de predicción actual es una segunda unidad de predicción dividida mediante división asimétrica, el candidato de combinación espacial correspondiente a una primera unidad de predicción dividida mediante la división asimétrica se establece como no disponible. 5. El método de conformidad con la reivindicación
- 44, en donde, si un tamaño de la unidad de predicción actual es (3/2)Nx2N, el candidato de combinación espacial a la izquierda se establece como no disponible.
- 56. El método de conformidad con la reivindicación 4, en donde un vector de movimiento del candidato de combinación temporal es un vector de movimiento de un bloque candidato de combinación temporal dentro de una imagen candidata de combinación temporal, y una posición del bloque candidato de combinación temporal se determina dependiendo de una posición del bloque actual dentro de una unidad de codificación más larga (LCU). IMPI instituto mexicano OF LA PROPIEDAD INDUSTRIAL
- 67. El método de conformidad con la reivindicación 1, en donde si ambos del parámetro de cuantificación izquierdo y el parámetro de cuantificación anterior no están disponibles, el predictor de parámetro de cuantificación se establece como el parámetro de cuantificación previo.
- 78. El método de conformidad con la reivindicación 1, en donde el tamaño predeterminado es un bloque 4x4. IMPI
Independent claims7
254 paragraphs in 35 sections, as filed
(54) Title: METHOD FOR DECODING VIDEO DATA.
(54) Title: METHOD OF DECODING VIDEO DATA.
(57) Summary
A method is provided that derives the index of the reference image and the motion vector from a current prediction unit, generates a prediction block of the current prediction unit using the index of the reference image and the motion vector, generates a residual block by means of inverse scanning, inverse quantization and inverse transformation and generates the reconstructed pixels using the prediction block and the residual block. The prediction pixels of the prediction block are generated using an interpolation filter based on the motion vector. Accordingly, the efficiency of encoding motion information is improved by including several combination candidates. Also, the computational complexity of the encoder and decoder is reduced by selecting different filters according to the location of the prediction pixels determined by the motion vector.
(57) Abstract
Provided is a method derives a reference picture Index and a motion vector of a current prediction unit, generates a prediction block of the current prediction unit using the reference picture Index and the motion vector, generafing a residual block by inverse-scan, inverse-quantization and inverse transform, and generates reconstructed pixels using the prediction block and the residual block. Prediction pixels of the prediction block is generated using an interpolation filter selected based on the motion vector. Accordingly, the coding efficiency of the motion Information is improved by including various merge candidates. Also, the computational complexity of an encoder and a decoder is reduced by selecting different filter according to location of the prediction pixels determined by the motion vector.
Mexican Institute of Industrial Property
<img file="MX347162B_D0001.tif" />
PATENT TITLE NO. 347162
Owner (s): INFOBRIDGE PTE. LTD.
Address: 10 Anson road # 23-140 International Plaza Singapore, 079903, SINGAPORE
Name: METHOD FOR DECODING VIDEO DATA.
Classification: lnt.CI.8: H04N19 / 105: H04N19 / 117; H04N19 / 122; H04N19 / 126; H04N19 / 159: H04N19 / 176; H04N19 / 182
Inventor (s): SOO MI OH; MOONOCK YANG
REQUEST
Number; International filing date:
MX / a / 2015/004214 November 2, 2012
Divisional Patent Number: 333762
<td>PRIORITY</td><td></td>
<td>Country: Date: KR; November 7, 2011</td><td>Number: 10-2011-0115348</td>
<td>Validity: Twenty years Expiration Date: November 2, 2032</td><td></td>
The reference patent is granted based on articles 1, 2<sup>or</sup> fraction V, 6<sup>or</sup> Section III, and 58 of the Industrial Property Law.
In accordance with article 23 of the Industrial Property Law, this patent is valid for twenty years, renewable, counted from the date of filing of the international application and will be subject to the payment of your fee to keep the rights. >
Whoever signs this title does so based on the provisions of articles 8, sections III and 7<sup>or</sup> bis 2 «and the Industrial Property Law (Official Gazette of the Federation (DOF} 06/27/1991, amended on 08/02/1994, 10/25/1996, 12/26/1907, 05/17/1999 , 01/26/2004, 06/16/2005, 01/25/2006, 05/06/2009, 06/01/2010, 08/18/2010, 08/28/2010, 01/27/2012 and 09 / 04/2012); Articles 1 ·. 3rd section V subsection a), 4th and 12th sections I and III of the Regulations of the Mexican Institute of Industrial Property (DOF 12/14/1998, amended on 07/01/2002,15/07/2004, 07/28/2004 and 08/07/2007); Articles 1 ", 3", 4 ", 5th section V maso a), 16 sections f and III and 30 of the Organic Statute of the Mexican Institute of Industrial Property (DOF 12/27/1999, amended on 10/10/2002, 07/29/2004, 08/04/2004 and 09/13/2007); 1, 3 and 5 subsection a) of the Agreement that delegates powers to the Deputy General Directors, Coordinator, Divisional Directors, Heads of Regional Offices, Divisional Deputy Directors, Departmental Coordinators and other subordinates of the Mexican Institute of Industrial Property. (DOF 12/15/1999, amended on 02/04/2000, 07/29/2004, 08/04/2004 and 09/13/2007).
<img file="MX347162B_D0002.tif" />
METHOD FOR DECODING VIDEO DATA
TECHNICAL FIELD
The present invention relates to a method for decoding video data, and more particularly, to a method for deriving motion information in the merge mode by constructing a merge candidate list using the spatial merge candidates. and temporal and generate prediction blocks using the movement information.
BACKGROUND OF THE INVENTION
Methods for compressing video data include MPEG2, MPEG-4, and H.264 / MPEG-4 AVC. According to these methods, an image is divided into macroblocks to encode an image, the respective macroblocks are encoded generating prediction blogs using the Inter prediction or the intra prediction. The difference between the original blocks and the prediction blocks is transformed to generate transformed blogues, and the transformed blocks are quantized using a quantization parameter and one of a plurality of predetermined quantization matrices. The quantized coefficient of the quantized blocks is scanned by means of a predetermined scan type and then subjected to entropic coding. Quantization parameters are set by macroblock and encoded using a
<img file="MX347162B_D0003.tif" />
previous quantization parameter.
IMPI
INSTITUTO MEXICANO OE LA PROPERTY INDUSTRIAL
In H.264 / MPEG-4 AVC, motion estimation is used. , to eliminate temporal redundancy between consecutive images. To detect temporal redundancy, one or more reference images are used to estimate the motion of the current blocks, and motion compensation is carried out to generate prediction blocks using the motion information. The motion information includes one or more reference image indices and one or more motion vectors.
According to H.264 / MPEG-4 AVC, only the motion vectors are predicted and encoded using the adjacent motion vectors, and the indices of the reference images are encoded without the indices of the adjacent reference images. Also, the computational complexity to generate the prediction blocks is high, since the prediction blocks are interpolated using a long branch filter.
However, if multiple sizes are used for inter prediction, the correlation between the current block movement information and the movement information of one or more adjacent blocks increases. The correlation between the movement vector of a current block and the movement vector of the adjacent block within a reference image becomes greater as the size of the image becomes greater if the Mexican institute movement of the image is almost constant "b ^ Naíi & ^ Éo.v? ^ · Consequently, the compression method with On / On. ~ L- fip.scri.txk— above reduces the efficiency of compression of motion information if the image size is larger than the high definition image and multiple sizes for motion estimation and motion compensation.
BRIEF DESCRIPTION OF THE INVENTION
Technical problem
The present invention is directed to a method for decoding video data by deriving the motion information by constructing a combination candidate list using the spatial combination candidates and the temporal combination candidates and generating the prediction blocks using a filter determined by the motion vector.
Technical Solution
One aspect of the present invention provides a method for decoding video data, comprising: deriving the index of the reference picture and the motion vector of a current prediction unit; generating a prediction block of the current prediction unit using the index of the reference image and the motion vector; generate a quantized block by inverse scanning the coefficient components
<img file="MX347162B_D0004.tif" />
MEXICAN INSTITUTE
Of the property
INDUSTRIAL quantified; generating a transform block by Ta inverse quantization of the quantized block using Tos quantization parameters; generating a residual block by inverse transformation of the transformation block; and generating the reconstructed pixels using the prediction block and the residual block. The prediction pixels of the prediction block are generated using an interpolation filter selected based on the motion vector.
Advantageous Effects
The method according to the present invention derives the index of the reference image and the motion vector of a current prediction unit, generates a prediction block of the current prediction unit using the index of the reference image and the vector of motion, generate a residual block using inverse scan, inverse quantization, and inverse transform, and generate the reconstructed pixels using the prediction blog and residual block. The prediction pixels in the block are generated using an interpolation filter selected based on the motion vector. Accordingly, the efficiency of encoding the motion information is improved by including several combination candidates. Also, the computational complexity of the encoder and decoder is reduced by selecting different input filters.
IMPI ^
INSTITUID MEXICANO
OF THE PROPERTY
INDUSTRIAL, according to the location of the prediction pixels determined by the motion vector.
DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of an image coding apparatus in accordance with the present invention.
FIG. 2 is a flow chart illustrating the method for decoding video data in the inter prediction mode in accordance with the present invention.
FIG. 3 is a conceptual diagram illustrating the positions of the pixels indicated by the motion vector in accordance with the present invention.
FIG. 4 is a flowchart illustrating a method for decoding motion information in combining mode in accordance with the present invention.
FIG. 5 is a conceptual diagram illustrating the positions of the candidate blocks of the spatial combination according to the present invention.
FIG. 6 is a conceptual diagram illustrating the positions of the candidate blocks of the spatial combination in an asymmetric division mode in accordance with the present invention.
FIG. 7 is another conceptual diagram illustrating the positions of the candidate blocks of the spatial combination in another mode of asymmetric division in accordance with the present invention.
<img file="MX347162B_D0005.tif" />
IMPI
Mexican Institute of Industrial Property
FIG. 8 is another conceptual diagram illustrating the positions of candidate blocks of spatial combining in another mode of asymmetric division in accordance with the present invention.
FIG. 9 is another conceptual diagram illustrating the positions of candidate blocks of spatial combination in another mode of asymmetric division in accordance with the present invention. '
FIG. 10 is a conceptual diagram illustrating the position of the candidate blocks of the temporal combination in accordance with the present invention.
FIG. 11 is a conceptual diagram illustrating a method for storing motion information in accordance with the present invention.
FIG. 12 is a block diagram of an image decoding apparatus 200 in accordance with the present invention.
FIG. 13 is a flow chart illustrating a method for decoding images in the inter prediction mode in accordance with the present invention.
FIG. 14 is a flow chart illustrating the method for deriving the motion information in the combination mode.
FIG. 15 is a flow chart illustrating the procedure for generating the residual block in the
IMPI Mexican institute, ·. Z. , _ _, · Ot THE PROPERTY prediction inter in accordance with the present inveiwarom
<img file="MX347162B_D0006.tif" />
DETAILED DESCRIPTION OF THE INVENTION
Hereafter, various embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited to the exemplary embodiments described below, but can be implemented in various types. Therefore, many other modifications and variations of the present invention are possible, and it will be understood that within the scope of the concept described, the present invention may be practiced other than as specifically described.
The image encoding apparatus and the image decoding apparatus in accordance with the present invention may be a user terminal, such as a personal computer, a mobile terminal, a mobile multimedia player, a smartphone, or wireless terminals. communications. The image encoding device and the image decoding device may be included in a communication unit for communicating with various devices, a memory for storing various programs, and data used to encode and decode images.
FIG. 1 is a block diagram of an image coding apparatus 100 in accordance with the present invention.
INSTITUTO MEXICANO DE LA Μ · ΡΙΕΟΑΙ;
Referring to FIG. 1, the apa'PSW *<sup>1</sup> Luü— ^ art decoding of images according to id ^ previously- · invention includes an image division unit 110, an intra prediction unit 120, an inter prediction unit 130, a transformation unit 140, a quantization unit 150 , a scanning unit 160, an entropic coding unit 170, a quantization / transform unit 180, a post-processing unit 190, and an image storage unit 195.
Image dividing unit 110 splits an image or section into several larger coding units (LCUs) and divides each LCU into one or more coding units. The size of the LCUs can be 32x32, 64x64 or 128x128. The image division unit 110 determines the prediction mode and the division mode of each coding unit.
LCUs include one or more encoding units. LCUs have a recursive quaternary tree structure to specify the division structure of LCUs. The parameters to specify the maximum size and minimum size of the encoding units are included in the sequence parameter set. The division structure is specified by one or more division encoding unit flags (division_cu_flags). The size of the encoding units is 2Nx2N. If the size of the<sup>9 </sup>INSTITUTO MEXICANO JH
OF INDUSTRIAL PROPERTY
LCU is 64x64 and the size of the smallest encoding units (SCO) is 8x8, the size of the encoding units can be 64x64, 32x32, 16x16 or 8x8.
The coding units include one or more prediction units. In intra prediction, the size of the prediction units is 2Nx2N. In Inter prediction, the size of the prediction units is specified by the division mode. The division mode is one of 2Nx2N, 2NxN, Nx2N and NxN if the encoding units are divided symmetrically.
The division mode is one of 2NxnU, 2NxnD, nLx2N and nRx2N if the division units are asymmetrically. Split modes are allowed based on the size of the encoding units to reduce equipment complexity. If the encoding units have a minimum size, symmetric division is not allowed. Also, if the encoding units have the minimum size, the NxN splitting mode cannot be allowed.
Coding units include one or more transformation units. The transformation units have a recursive quaternary tree structure to specify the division structure of the coding units. The slice structure is specified by one or more slice transform unit flags (slice_your_flags). The parameters to specify the maximum size and minimum size of the units of
INSTITUTO MEXICANO DE LA FROEUDAD industrial transformation of luma are included in a set of sequence parameters.
The intra prediction unit 120 determines the intra prediction mode of a current prediction unit and generates a prediction block using the intra prediction mode.
The prediction unit 130 inter-determines the motion information of a current prediction unit using one or more reference images stored in the image storage unit 195, and generates a prediction block of the prediction units. The motion information includes one or more indices of the reference images and one or more motion vectors.
The transformation unit 140 transforms the residual blog to generate a transformed block. The residual block has the same size as the transformation units. If the transform units are greater than the transform units, the residual signals between the current block and the prediction block are divided into multiple residual blocks.
Quantization unit 150 determines the quantization parameter for quantizing the transform block. The quantization parameter is the quantization step size. The quantization parameter is determined according to the quantization units. The size of the quantization units can vary and is one of the “« τντο MEXICAN kSLw-SÍP permissible sizes of the coding units ^ I ^ W size of the coding units is equal to 1 U, the minimum size of the coding units quantization, the encoding units become the quantization units. A plurality of coding units may be included in the minimum size quantization units. The minimum size of the quantization units is determined according to the images and the parameter to specify the size of the quantization units is included in the image parameter set.
The quantization unit 150 generates a predictor of the quantization parameter and generates a differential quantization parameter by subtracting the predictor of the quantization parameter from the quantization parameter. The differential quantization parameter is entropic encoded.
The quantization parameter predictor is generated using the quantization parameters of the adjacent coding units and the quantization parameter of the previous coding unit as follows.
The quantization parameter on the left, and the upper quantization parameter and the pre-quantization parameter are retrieved sequentially in this order. The average of the first two available quantization parameters retrieved in that order, is
ΡΙ sets as the predictor of the quantization parameter when two or more quantization parameters are available., And when only one quantization parameter is available, the available quantization parameter is set as the predictor of the quantization parameter. That is, if the left and top quantization parameters are available, the average of the left and top quantization parameters is set as the predictor of the quantization parameter. If only one of the quantization parameters to the left and top is available, the average of the available quantization parameter and the previous quantization parameters are set as the predictor of the quantization parameter. If both the left and top quantization parameters are not available, the previous quantization parameter is set as the predictor of the quantization parameter. The average is rounded.
The differential quantization parameter is converted into binary sequences for the absolute value of the differential quantization parameter and a binary sequence to indicate the sign of the differential quantization parameter through a binarization process, and the binary sequences are arithmetically encoded. If the absolute value of the differential quantization parameter is 0, the binary sequence to indicate the sign can be omitted.
The
IMPI pin it for binarization of the value absoTSVd.
The quantization unit 150 quantTfTc'áT transformed using a quantization matrix and the quantization parameter to generate a quantized block. The quantized block is provided to the quantization / inverse transform unit 180 and the scan unit 160.
The scan unit 160 determines whether a scan pattern is applied to the quantized block.
In inter prediction, the diagonal scan is used as the scan pattern if CABAC is used for entropic coding. The quantized coefficients of the quantized block are divided into the components of the coefficient. The components of the coefficient are the significant flags, the signs of the coefficient, and the levels of the coefficient. The diagonal scan is applied to each of the components of the coefficient. The significant coefficients indicate whether the corresponding quantized coefficient is zero or not. the sign of the coefficient indicates the sign of the non-zero quantization coefficient, and the level of the coefficient indicates the absolute value of the non-zero quantized coefficient.
When the size of the transform unit is larger than the default size, the quantized block is divided into multiple subsets and the diagonal scan is
<img file="MX347162B_D0007.tif" />
<img file="MX347162B_D0008.tif" />
IMPI Mexican fwrnyTo de la fropiedab industrial applies to each subset. The significant flags, the signs of the coefficients, and the levels of the coefficients of each subset are respectively scanned according to the diagonal scan. The default size is 4x4. The subset is a 4x4 block containing 16 transformation coefficients.
The scan pattern for exploring subsets is the same as the scan pattern for exploring the coefficient components. The significant flags, the signs of the coefficients, and the levels of the coefficients of each subset are scanned in the reverse direction. The subsets are also scanned in the reverse direction.
The parameter indicating the position of the last non-zero coefficient is encoded and transmitted next to the decoding. The parameter indicating the position of the last non-zero coefficient specifies the position of the last non-zero coefficient within the quantized block. The non-zero subset flag is defined for each subset other than the first subset and the last subset is transmitted alongside decoding. The first subset covers a coefficient of DC. The last subset covers the last nonzero coefficient. The nonzero subset flag indicates whether the subset contains nonzero coefficients or not.
The entropic coding unit 170 applies the pinch of the entropic coding property to the component scanned by the scan unit 160, the intra prediction information received from the intra prediction unit 120, the motion information received from the intra prediction unit 130. , and so on.
The quantization / inverse transform unit 180, inverse quantizes the quantized coefficients of the quantized block, and inverse transforms the block with inverse quantization to generate the residual signals.
Post-processing unit 190 performs an unlock filtering process to remove lock objects generated in a reconstructed image.
The image storage unit 195 receives the post-processed images from the post-processing unit 190, and stores the images in image units. An image can be a box or a field.
FIG. 2 is a flow chart illustrating a method for encoding video data in the inter-prediction mode in accordance with the present invention.
The movement information of a current block is determined (S110). The current block is a prediction unit. The size of the current block is determined by the size and division mode of the encoding unit.
Movement information varies according to the <sup>16</sup> IMPI fNSIiiytv MEXICAN MU FkOHEDAD INDUSTRiU prediction type. If the prediction type is one-way prediction, the motion information includes the reference index that specifies the image of a reference list 0, and a motion vector. If the prediction type is bidirectional prediction, the motion information includes two reference indices that specify an image from reference list 0 and an image from reference list 1, and a motion vector from list 0 and a motion vector from list 1.
A prediction block of the current block is generated using the movement information (S120).
If the motion vector indicates an entire pixel location, the prediction block is generated by copying a block from the reference image, specified by the motion vector. If the motion vector indicates the location of a subpixel, the prediction block is generated by interpolating the pixels of the reference image. The motion vector is given in quarter pixel units.
FIG. 3 is a conceptual diagram illustrating the positions of the pixels indicated by a motion vector in accordance with the present invention.
In FIG. 3, the pixels labeled LO, RO, Rl, Ll, A0 and B0 are integer position pixels of the reference image and the pixels labeled a<sub>THE</sub>to r<sub>L</sub>oat the subpixel locations are fractional pixels to be
<img file="MX347162B_D0009.tif" />
<img file="MX347162B_D0010.tif" />
- IMPI
MEXICAN INSSTTTVTO <sup>Dr</sup> THE PROPIFDap IMD'JSTM / l interpolated using an interpolation filter «..-... eX -... £ L ^^ selects based on the motion vector.
If a pixel to be interpolated is located at the location of the subpixel to<sub>L</sub>o, b<sub>L0</sub> oc<sub>L0</sub>, the pixel labeled with a<sub>L0</sub>, b<sub>L</sub>ooc<sub>L</sub>or it is generated by applying an interpolation filter to the pixels in the horizontally closest integer position. If the pixel to be interpolated is located at the location of the subpixel d<sub>L</sub>oz h<sub>L0</sub> on<sub>L0</sub>, the pixel labeled d<sub>L0</sub>, h<sub>L0</sub> on<sub>L</sub>or is generated by applying an interpolation filter to the pixels of the vertically closest integer position. If a pixel to be interpolated is located at the subpixel location and<sub>L</sub>or, Ílo op<sub>L0</sub>, the pixel labeled with e<sub>L</sub>o, ile op<sub>L</sub>or it is generated by applying an interpolation filter to the vertically closest interpolated pixels, each of which includes an 'a' character within its label. If a pixel to be interpolated is located at the subpixel location g<sub>L</sub>cu k<sub>L</sub>oor<sub>L0</sub>, the pixel labeled g<sub>L0</sub>, k<sub>L0</sub> or<sub>L0</sub> it is generated by applying an interpolation filter to the vertically closest interpolated pixels, each of which includes a 'c' character within its label. If a pixel to be interpolated is located at the location of the subpixel f<sub>L</sub>Of Ilo oq<sub>L</sub>o, the pixel labeled f<sub>L0</sub>, j<sub>L0</sub> oq<sub>L0</sub> It is generated by applying an interpolation filter to vertically adjacent interpolated pixels, each of which includes a 'c' character within its label.
IMPIAS INSTITUTO MEXICANO 7
Say LA raOP. -DAD industrial ^ g ~ ur ^^
The interpolation filter is determined based on the location of the sub-pixel of the pixel to be interpolated7 ~ o'cbnToase in the prediction mode and the location of the sub-pixel of the pixel to be interpolated.
Table 1 shows the exemplary filters. The location of the subpixel H indicates the location of a half pixel in the direction of interpolation. For example, locations b<sub>L</sub>oz h<sub>L0</sub>, Ílo, j lo, yk<sub>L</sub>o correspond to the location of the H subpixels. The FL and -FR subpixel locations indicate the location of a quarter pixel in the direction of interpolation. For example, the locations to<sub>L</sub>oz d<sub>L</sub>oz e<sub>L</sub>oz f<sub>THE</sub>z and 9lo correspond to the location of the FL subpixels, and the locations c<sub>L</sub>o, n<sub>L0</sub>, Ploz Qlo yr<sub>L0</sub> correspond to the location of the FR subpixels.
Table 1
<td>Prediction mode</td><td>Subpixel location</td><td>Filter coefficients</td>
<td rowspan="3">One-way prediction</td><td>H</td><td> {2, -8, 36, 36, -8,2}</td>
<td>FL</td><td> {-3,51,20, -7,2}</td>
<td>FR</td><td> {2, -7,20,51,-3}</td>
<td rowspan="3">Bidirectional prediction</td><td>H</td><td> {-1,4,-11,40, 40,-11,4.-1)</td>
<td>FL</td><td> {-1,4,-10, 57, 19, -7,3,-1}</td>
<td>FR</td><td> {-1,3,-7, 19, 57,-10, 4,-1}</td>
As shown in Table 1, in one-way prediction, the 6-lead symmetric filter can be used to interpolate the pixels at the half-pixel location H, and the 5-lead asymmetric filter can be used to interpolate the pixels at the location. of a
<img file="MX347162B_D0011.tif" />
IMPI
INSTITUT · MEXICANO DE LA PROPIEDAD INDUSTRIAL quarter pixel FL or FR. In bidirectional prediction, the 8-lead symmetric filter can be used for the H half-skin location and the 8-lead asymmetric filter can be used for the FL 1/4 pixel location and
FR.
Alternatively, the filter can be determined only by the location of the subpixels of the pixel to be interpolated. In unidirectional prediction, the 8-lead symmetric filter can be used to interpolate the pixels of the half-pixel locations and the 7-lead asymmetric filter or the 6-lead asymmetric filter can be used to interpolate the pixels of the locations of a quarter pixel. In bidirectional prediction, the same filter or another filter that has a smaller number of taps can be used to interpolate the pixels from the sub-pixel locations.
A residual block is generated using the current block and the prediction block (130). The residual block has the same size as the transformation unit. If the prediction unit is greater than the transform unit, the residual signals between the current block and the prediction block are in multiple residual blocks.
The residual block is encoded (S140). The residual block is encoded by the transform unit 140, the quantization unit 150, the scan unit 160 and the unit
IMPI
MEXICAN INSTITUTE OF INDUSTRIAL PROPERTY
170 encoding code of FIG. 1.
The movement information is encoded (S150). The movement information can be predictively encoded using the spatial candidates and a temporal candidate from the current block. Motion information is encoded in skip mode, combination mode, or AMVP mode. In the skip mode, the prediction units are the size of the encoding units and the movement information is encoded using the same method as that of the combining mode. In the combination mode, the current prediction unit movement information is equal to the movement information of a candidate. In AMVP mode, the motion vector of motion information is predictively encoded using one or more motion vector candidates.
FIG. 4 is a flow chart illustrating a method for encoding motion information in combination mode in accordance with the present invention.
The candidates of the spatial combination are derived (S210). FIG. 5 is a conceptual diagram illustrating the positions of the candidate blocks of the spatial combination according to the present invention.
As shown in FIG. 5, the candidate blocks of the combination are the block on the left (block A), the upper block (block B), the block on the upper right
IMPI
MEXICAN INSTITUTE
OF THE INDUSTRIAL PROPERTY (block C), the block at the bottom left (block D) or the block at the top left (block E) of the current block. The blocks are prediction blocks. The upper left block (block E) is set as the candidate block of the combination when one or more of the blocks A, B, C and D are not available. The motion information of a candidate block of the available combination N is set as the candidate of spatial combination N. N is A, B, C, D, or E.
The candidate of the spatial combination can be set as unavailable according to the shape of the current block and the position of the current block. For example, if the coding units are divided into two prediction units (P0 block and Pl block) using asymmetric division, it is likely that the movement information of the PO block is not equal to the movement information of the Pl block. Therefore, if the current block is the asymmetric block Pl, the PO block is set as the unavailable candidate block as shown in FIGS. 6 to 9.
FIG. 6 is a conceptual diagram illustrating the positions of spatial combining candidate blocks in an asymmetric division mode in accordance with the present invention.
As shown in FIG. 6, the coding units are divided into two asymmetric prediction blocks PO and Pl and the division mode is an nLx2N mode. he
INSTITUTO MEXICANO DE LA MOHEDA »INDUSTRIAL size of the PO block is nHx2N and the size of the Pl block is (2h) Nx2N. The value of h is 1/2. The current block is the Pl block. Blocks A, B, C, D, and E are candidate blocks of the spatial combination. The PO block is the candidate block of the spatial combination A.
In the present invention, the spatial pool candidate A is set as unavailable so as not to be listed in the pool candidate list. Also, the candidate blocks of the space combination B, C, D and E that have the same information as the candidate block of the space combination A are set as unavailable.
FIG. 7 is another conceptual diagram illustrating the positions of the candidate blocks of spatial combination in the asymmetric division mode in accordance with the present invention.
As shown in FIG. 7, the coding units are divided into two asymmetric prediction blocks PO and Pl and the division mode is the nRx2N mode. The PO block size is (2-h) Nx2N and the Pl block size is hNx2N. The value of h is 1/2. The current block is the Pl block. Blocks A, B, C, D, and E are spatial combination blocks. The PO block is the candidate block of the spatial combination A.
In the present invention, the candidate of spatial combination A is set as unavailable so as not to be listed
IMPI
MEXICAN PHOTO OF INDUSTRIAL PROPERTY on the list of combination candidates. Also, candidate blocks of space combination B, C, D, or E that have the same information as the candidate block of space combination A are set as unavailable.
FIG. 8 is another conceptual diagram illustrating the positions of candidate blocks of spatial combining in another symmetric division mode in accordance with the present invention.
As shown in FIG. 8, the coding units are divided into two asymmetric prediction blocks P0 and P1 and the division mode is the 2NxnU mode. The size of the PO block is 2NxhN and the size of the P1 block is 2Nx (2-h) N. The value of h is 1/2. The current block is the Pl block. Blocks a, B, C, D, and E are candidate blocks of the spatial combination. The PO block is the candidate block of spatial combination B.
In the present invention, the candidate of the spatial combination B is set as unavailable so as not to be listed in the list of combination candidates. Also, candidate blocks of spatial combination C, D, or E that have the same motion information of the candidate block of spatial combination B is set as unavailable.
FIG. 9 is another conceptual diagram illustrating the positions of the candidate blocks of the spatial combination in another mode of asymmetric division according to the
INSTITUTO MEXICANO DE LA PROPIEDAD INDUSTRIAL,,, Z 1 UAL present invention.
As shown in FIG. 9, the units - "ζΓ5" encoding are divided into two asymmetric prediction blocks P0 and Pl and the division mode is 2NxnD mode. The size of the P0 block is 2Nx (2-h) N and the size of the Pl block is 2NxhN. The value of h is 1/2. The current block is the Pl block. Blocks A, B, C, D and E are candidate blocks of the spatial combination. Block P0 is the candidate block of spatial combination B.
In the present invention, the candidate of spatial combination B is set as unavailable so as not to be listed in the -combination candidate list. Also, candidate blocks of spatial combination C, D or E that have the same motion information as candidate block of spatial combination B are set as unavailable.
Candidates for spatial join can also be set to unavailable based on join area. If the current block and the candidate blocks of the spatial join belong to the same join area, the candidate blocks of spatial join are set as unavailable. The combining area is a unit area in which motion estimation is carried out and the information specifying the combining area is included in a bit stream.
A candidate is derived from the temporal combination (S220).
INSTITUTO MEXICANO I »LA MIOEIEDAP INDUSTRIAL
The candidate of the temporary combination includes the index of the reference image, and a motion vector of the candidate of the temporary combination.
The index of the reference image of the temporal combining candidate can be derived using one or more indices of the reference images of the adjacent blocks. For example, one of the indexes of the reference images of the adjacent block on the left, the upper adjacent block and the adjacent corner block, is set as the index of the reference image of the candidate of the temporary combination. The adjacent corner block is one of the adjacent block at the top right, the adjacent block at the bottom left, and the adjacent block at the top left. Alternatively, the index of the candidate reference image of the temporal combination may be set to zero to reduce complexity.
The motion vector of the temporal combination candidate can be derived in the following way.
First, the candidate image of the temporary combination is determined. The temporary join candidate image includes a temporary join candidate block. A candidate image of the temporary combination is used within a section. The index of the reference image of the candidate image of the temporal combination may be set to zero.
<img file="MX347162B_D0012.tif" />
IMPI
INSTITUTO MEXICANO ot INDUSTRIAL property
If the current section is section P, ima d<sup>p</sup> The reference images in the reference image list 0 is set as the candidate image of the temporary combination. If the current section is a B section, one of the reference images in the reference image lists 0 and 1 is set as the candidate image for the temporal combination. A list indicator, indicating whether the candidate image of the temporary combination belongs to reference image lists 0 or 1 is included in the section heading if the current section is section B. The index of the image of Reference that specifies the candidate image of the temporary combination can be included in the section heading.
Next, the candidate block of the temporary combination is determined. FIG. 10 is a conceptual diagram illustrating the position of the candidate block of the temporal combination in accordance with the present invention. As shown in FIG. 10, a first candidate block can be the block in the lower right corner (block H) of block C. of block C has the same size and location as the current block and is located within the candidate image of the temporary combination. A second candidate block is the block that covers the upper left pixel of the center of block C.
The candidate block of the temporary combination can be the first candidate block or the second candidate block. Yes
Mexican INSTITUTE OF INDUSTRIAL MOFIEDAD the first candidate block is available, the first candidate block is established as the candidate block of the temporary combination. If the first candidate block is not available, the second candidate block is set as the temporary join candidate block. If the second candidate block is not available, the candidate block of the temporary join is set to unavailable.
The candidate block of the temporary combination is determined based on the position of the current block. For example, if the current block is adjacent to a lower LCU (that is, if the first candidate block belongs to a lower LCU), the first candidate block can be changed to a block within the current LCU or set to unavailable. .
Also, the first and second candidate blocks can be changed to another block based on the position of the candidate block within a motion vector storage unit. The motion vector storage unit is a basic unit that stores the motion information of the reference images.
FIG. 11 is a conceptual diagram illustrating a method for storing motion information in accordance with the present invention. As shown in FIG. 11, the movement storage unit can be a 16x16 block. The motion vector storage unit can be divided into sixteen 4x4 blocks. If the motion vector storage unit is a 16x16 block, the motion information is stored according to the motion vector storage unit. If the motion vector storage unit includes multiple reference image prediction units, the motion information of a predetermined prediction unit of multiple prediction units is stored in memory, to reduce the amount of the motion information to be stored in memory. The default prediction unit can be a block that covers one of sixteen 4x4 blocks. The predetermined prediction unit may be a block covering block C3, block BR. Or the default prediction unit can be a block that covers the UL block.
Therefore, if the candidate block does not include the predetermined block, the candidate block is changed to a block that includes the predetermined block.
If the candidate block of the time join is determined, the motion vector of the candidate block of the time join is set as the motion vector of the candidate of the time join.
The combination candidate list is constructed (S230). Available space candidates and available temporary candidates are listed in the default order.
MEXICAN INSTITUTE
OF THE PROPERTY
INDUSTRIAL
The candidates of the spatial combination are listed up to four in the order of A, B, C, D, and E. The candidate of the temporal combination can be listed between B and C or after the spatial candidates.
It is determined whether one or more candidates of the combination are generated or not (S240). The determination is carried out by comparing the number of combination candidates listed in the combination candidate list with a predetermined number of combination candidates. The default number can be determined by image or section.
If the number of combination candidates listed in the combination candidate list is less than a predetermined number of combination candidates, one or more combination candidates is generated (S250). The generated join candidates are listed after the last available join candidate.
If the number of available combination candidates is equal to or greater than 2, one or two available combination candidates has the movement information from list 0 and the other has the movement information from list 1, the combination candidates can be generated by combining the move information from list 0 and the move information from list 1. Multiple combination candidates can be generated if there are multiple combinations.
One or more zero combination candidates can be τ
IMPI Mexican iNSTmrro DE LA MIOPIEDAP INDUSTRIAL ·
<img file="MX347162B_D0013.tif" />
added to the list. If the type of the section is P, the zero-join candidate has only the movement information from list 0. If the section type is B, the zero-join candidate has the information for the list 0 movement and the list movement information 1.
A join predictor is selected from the join candidates from the join list. The combination index that the combination predictor specifies is encoded (S260).
FIG. 12 is a flowchart that an image decoding apparatus 200 according to the present invention.
The image decoding apparatus 200 according to the present invention includes an entropic decoding unit 210, a reverse scan unit 220, an inverse quantization unit 230, an inverse transform unit 240, an intra prediction unit 250, a inter-prediction unit 260, a post-processing unit 270, an image storage unit 280, and an adder 290.
The entropic decoding unit 210 extracts the intra prediction information, the inter prediction information, and the components of the quantized coefficients from a received bit stream using the decoding method.
IMPI INSTITUTO MEXICANO DE LA PROPERTY INDUSTRIAL arithmetic, binary, adaptive to the context.
The reverse scan unit 220 applies a reverse scan pattern to the components of the quantized coefficient to generate the quantized block. In intra prediction, the reverse scan pattern is the diagonal scan. The components of the quantized coefficient include the significant flags, the signs of the coefficient, and the levels of the coefficients.
When the size of the transformation unit is larger than the predetermined size, the significant flags, coefficient signs, and coefficient levels are scanned inversely into units of subsets using diagonal scan, to generate subsets, and subsets are scanned scan in reverse using diagonal scan to generate the quantized block. The default size is equal to the size of the subset. The subset is a 4x4 block that includes 16 transformation coefficients. Significant flags, coefficient signs, and coefficient levels are scanned inversely in the reverse direction. The subsets are also scanned inversely in the reverse direction.
A parameter indicating the position of the last nonzero coefficient and the flags of the
Λ.
nonzero subsets are extracted from the bit stream. The number of encoded subsets is determined based on
<img file="MX347162B_D0014.tif" />
<img file="MX347162B_D0015.tif" />
the parameter indicating the position of the last nonzero coefficient. The nonzero subset flag is used to determine if the corresponding subset has at least one nonzero coefficient. If the nonzero subset flag equals 1, the subset is generated using diagonal scanning. The first subset and the last subset are generated using the reverse scan pattern. '
The inverse quantization unit 230 receives the differential quantization parameter from the entropy decoding unit 210 and generates the quantization parameter predictor to generate the coding unit quantization parameter. The operation for generating the quantization parameter predictor is the same as the operation of the quantization unit 150 of FIG. 1. Then, the quantization parameter of the current coding unit is generated by adding the differential quantization parameter and the predictor of the quantization parameter. If the differential quantization parameter for the current coding unit is not transmitted from the coding side, the differential quantization parameter is set to zero.
The inverse quantization unit 230 inverse quantizes the quantized block.
The inverse transform unit 240 transforms from
<img file="MX347162B_D0016.tif" />
IMPI
INSTITUTO M'TlCANa DS LA PROPIEDAD INDUSTRIAL inversely forms the inversely quantized block to generate a residual block. An inverse transformation matrix is adaptively determined according to the prediction mode and the size of the transformation unit. The inverse transform matrix is a DCT-based integer transform matrix or a DST-based integer transform matrix. In inter prediction, DCT-based integer transformations are used.
The intra prediction unit 250 derives the intra prediction mode from the current prediction using the received intra prediction information, and generates a prediction block in accordance with the intra derived prediction mode.
The inter-prediction unit 260 derives the movement information from the current prediction unit using the received inter-prediction information, and generates a prediction blog using the movement information.
Post-processing unit 270 operates the same as post-processing unit 180 of FIG. 1.
The image storage unit 280 receives the post-processed images from the post-processing unit 270, and stores the images in image units. Images can be frames or fields.
Adder 290 sums the restored residual block and the prediction block to generate a reconstructed block.
FIG. 13 is a flow chart illustrating a method
<img file="MX347162B_D0017.tif" />
for decoding an image in the inter prediction mode according to the present invention.
The movement information of a current block is derived (S310). The current block is a prediction unit. The size of the current block is determined by the size of the encoding unit and the mode of the division.
The information of the movement varies according to the type of prediction. If the type of the prediction is one-way prediction, the motion information includes a reference index that specifies an image from reference list 0 and a motion vector. If the prediction type is bidirectional prediction, the motion information includes a reference index that specifies an image from reference list 0, a reference index that specifies an image from reference list 1, and a motion vector from list 0 and a motion vector from list 1.
Motion information is adaptively decoded according to the motion information encoding mode. The encoding mode of the movement information is determined by means of a skip flag and a combination flag. If the skip flag is equal to 1, the combination flag does not exist and the encoding mode is the skip mode. If the skip flag is equal to 0 and the combination flag is equal to
IMPI
INSTITUTO MEXICANO DE LA PEOF1EOAD INDUSTRIAL
1, the encoding mode is the skip flag mode and the combination flag are equal to 0, the encoding mode is the AMVP mode.
A prediction blog of the current block is generated using the movement information (S320).
If the motion vector indicates an integer pixel location, the prediction block is generated by copying a block from the reference image specified by the motion vector. If the motion vector indicates a sub-pixel location, the prediction block is generated by interpolating the pixels in the reference image. The motion vector is given in quarter pixel units.
As shown in FIG. 3, the pixels labeled with LO, RO, Rl, Ll, Ά0 and B0 are integer position pixels of the reference image and the pixels labeled with a<sub>L</sub>oar<sub>L0</sub> at sub-pixel locations, they are fractional pixels to be interpolated using an interpolation filter which is selected based on the motion vector.
If a pixel to be interpolated is located at a location from subpixel to<sub>L</sub>ooc<sub>L0</sub>, the pixel tagged with a<sub>L0</sub>, b<sub>L</sub>ooc<sub>L</sub>or it is generated by applying an interpolation filter for the pixels of the closest integer position horizontally. If a pixel to be interpolated is located at a subpixel location d<sub>L</sub>o, h<sub>L</sub>oon<sub>L0</sub>, the pixel labeled d<sub>L0</sub>, h<sub>L0</sub> on<sub>L0</sub> is generated by applying an interpolation filter to the pixels of the
IMPI
INSTITUTO MEXICANO DE LA PEOWEDAD INDUSTRIAL whole position vertically closer. If a pixel to be interpolated is located at a subpixel location and<sub>L</sub>oz Ilo or Plo <the skin tagged with e<sub>L</sub>or, Ílo or Plo is generated by applying an interpolation filter to the vertically closest interpolated pixels, each of which includes an 'a' character within its label. If a pixel to be interpolated is located at the subpixel location g<sub>L</sub>oz kro or<sub>L0</sub>, the pixel labeled with g<sub>THE</sub>zk<sub>L0</sub> or<sub>L0</sub> it is generated by applying an interpolation filter to the vertically closest interpolated pixels, each of which includes a 'c' character within its label. If a pixel to be interpolated is located at the subpixel location f<sub>THE</sub>z jto oq<sub>L</sub>oz the pixel tagged with f<sub>L</sub>oz LÍlo oq<sub>L</sub>oz is generated by applying an interpolation filter to vertically adjacent interpolated pixels each of which includes a 'c' character within its label.
The interpolation filter is determined based on the location of the subpixels of the pixels to be interpolated or based on a prediction mode and the location of the subpixels of the pixels to be interpolated.
As shown in Table 1, in the unidirectional unit, the 6-lead symmetric filter can be used to interpolate the pixels of the half-pixel H location, and the 5-lead asymmetric filter can be used to interpolate the pixels of the H location.
<img file="MX347162B_D0018.tif" />
IMPI
MEXICAN INSTITUTE
DE LA PROFIEDAP industrial quarter pixel FL or FR. In bidirectional prediction, the 8-lead symmetric filter can be used for the half-pixel location H, and the 8-lead asymmetric filter can be used for the quarter-pixel location FL and FR.
Alternatively, the filter can be determined only by the location of the sub-pixel of the pixel to be interpolated. In unidirectional prediction, the 8-lead symmetric filter can be used to interpolate the pixels from the half-pixel locations and the 7-lead or 6-lead asymmetric filter can be used to interpolate the pixels from the quarter-pixel locations. . In bidirectional prediction, the same filter and another filter that has a smaller number of taps can be used to interpolate the pixels from the subpixel locations.
A residual block is generated (S330). The residual block is generated by the entropic decoding unit 210, the reverse scan unit 220, the inverse quantization unit 230, and the inverse transform unit 240 of FIG. 12.
A reconstructed block is generated using the prediction block and the residual block (S340).
The prediction block has the same size as the prediction unit, and the residual block has the same size
INSTITUTO MEXICANO DE LA PROPERTY industrial size of the transformation unit. Therefore, the residual signals and the prediction signals of the same size are added to generate the reconstructed signals.
FIG. 14 is a flow chart illustrating a method for deriving motion information in the combining mode.
A combination index is extracted from the bit stream (S410). If the join index does not exist, the number of join candidates is set to one.
The spatial combination candidates are derived (S420). The available spatial combination candidates are the same as described in S210 of FIG. Four.
A temporary combination candidate is derived (S430). The temporal merge candidate includes an index of the reference image and a motion vector of the temporal merge candidate. The reference index and motion vector of the time combining candidate are the same as described in S220 of FIG. Four.
A join candidate list is constructed (S440). The combination list is the same as that described in S230 of FIG. Four.
It is determined whether or not one or more combination candidates are generated (S450). The determination is carried out by comparing the number of combination candidates listed in the combination candidate list with a predetermined number of combination candidates. The default number is determined by image or section.
If the number of combination candidates listed in the combination candidate list is less than the predetermined number of combination candidates, one or more combination candidates is generated (S460). The generated join candidates are listed after the last available join candidate. The combination candidate is generated as the same method described in S250 of FIG. Four.
The join candidate specified by the join index is set as the current block move information (S470).
FIG. 15 is a flow chart illustrating the procedure for generating a residual block in intra prediction mode according to the present invention.
The quantized coefficient components are generated by the entropic decoding unit (S510).
A quantized block is generated by inverse scanning the coefficient components according to the diagonal scan (S520). The components of the coefficient include the significant flags, the signs of the coefficient, and the levels of the coefficients.
When the size of the transformation units is greater than a predetermined size, the significant flags, the coefficient signs, and the levels of
INSTITUTO MEXICANO DE LA PROPIEDAD INDUSTRIAL coefficients are scanned inversely in units of subsets using the diagonal scan to generate the subsets, and the subsets are scanned inversely using the diagonal scan to generate the quantized block. The default size is equal to the size of the subsets. The subsets are 4x4 blocks that include 16 transformation coefficients. Significant flags, coefficient signs, and coefficient levels are scanned inversely in the reverse direction. The subsets are also scanned inversely in the reverse direction.
The parameter indicating the position of the last non-zero coefficient and the non-zero subset flags are extracted from the bit stream. The number of coded subsets is determined based on the parameter indicating the position of the last non-zero coefficient. Nonzero subset flags are used to determine if the subset has at least one nonzero coefficient. If the nonzero subset flag equals 1, the subset is generated using diagonal scanning. The first subset and the last subset are generated using the reverse scan pattern.
The quantized block is inverse quantized using an inverse quantization matrix and a quantization parameter (S530).
IMPI?
MEXICAN INSTITUTE OF INDUSTRIAL PROPERTY
The minimum size of '”TdS uillCtróteB' is determined.<sup>1</sup> of "" quantification. The cu_qp_delta_habilizadas_info parameter that specifies the minimum size is extracted from a bit stream, and the minimum size of the quantization units by means of the following equation.
Log2 (MinQUTsize) = Log2 (MaxCUTsize) -cu_qp_delta_enabled_ info
MinQUTsize indicates the minimum size of the quantization units, MaxCUTsize indicates the size of the LCUs. The cu_qp_delta_habilitado_info parameter is extracted from a set of image parameters.
A differential quantization parameter is derived from the current coding unit. The difference quantization parameter is included per quantization unit. Therefore, if the size of the current coding unit is equal to or greater than the minimum size of the quantization units, the quantization parameter for the current coding unit is reset. If there is no differential quantization parameter, the differential quantization parameter · is set to zero. If multiple coding units belong to a quantization unit, the first coding unit containing at least one non-zero coefficient in the order of d-echoing contains the differential quantization unit.
IMPI
INSTITUTO MEXICANO O £ La nOMEDAO INDUSTRIAL
The encoded differential quantization parameter is arithmetically decoded to generate a binary string indicating the absolute value of the differential quantization parameter and a binary string indicating the sign of the differential quantization parameter. The binary string can be a bound unary code. If the absolute value of the differential quantization parameter is zero, the binary string indicating the sign does not exist. The differential quantization parameter is derived using the binary string
<img file="MX347162B_D0019.tif" />
A that indicates the absolute value of the binary string that the sign indicates.
A predictor of the quantization parameter is derived from the current coding unit. The predictor of the quantization parameter is generated using the quantization parameters of the adjacent coding units and the quantization parameter of the previous coding unit as follows.
One quantization parameter to the left, the upper quantization parameter, and the previous quantization parameter are retrieved sequentially in this order. The average of the first two available quantization parameters retrieved in that order is set as the predictor of the quantization parameter when two or more quantization parameters are available, and when only one quantization parameter is available.
INSTITUTO MEXICANC,. , OF THE quantification PROPERTY, the quantification parameter ΰΐ '^ ι ^ ηισι ^ -states as the predictor of the parameter -do cnant4fioac4p & »^ That is, if the quantization parameter is available on the left and above, the average of the parameter of quantization on the left and top is set as the predictor of the quantization parameter. If only one of the quantization parameter is available to the left and above, the average of the available quantization parameter and the previous quantization parameter is set as the predictor of the quantization parameter. If both the left and top quantization parameter are not available, the previous quantization parameter is set as the predictor of the quantization parameter.
If multiple coding units belong to a minimum size quantization unit, the quantization parameter predictor for the first unit in the order of decoding is derived and used for the other coding units.
The quantization parameter of the current coding unit is generated using the differential quantization parameter and the predictor of the quantization parameter.
A residual block is generated by inverse transforming the block with inverse quantization (S450). One-dimensional DCT-based inverse transformation is used.
ΙΜΡΓ
MEXICAN INSTITUTE OF PROPERTY V \> - JΏ
INDUSTRIAL _____ horizontal and vertical
<img file="MX347162B_D0020.tif" />
Although the invention has been shown and described with reference to certain exemplary embodiments thereof, those skilled in the art will understand that various changes can be made in the form and details thereof without departing from the spirit and scope of the invention. as defined by the appended claims.
<img file="MX347162B_D0021.tif" />
Contents35
36 sheets
Sheet 1 Sheet 2 Sheet 3 Sheet 4 Sheet 5 Sheet 6 Sheet 7 Sheet 8 Sheet 9 Sheet 10 Sheet 11 Sheet 12 Sheet 13 Sheet 14 Sheet 15 Sheet 16 Sheet 17 Sheet 18 Sheet 19 Sheet 20 Sheet 21 Sheet 22 Sheet 23 Sheet 24 Sheet 25 Sheet 26 Sheet 27 Sheet 28 Sheet 29 Sheet 30 Sheet 31 Sheet 32 Sheet 33 Sheet 34 Sheet 35 Sheet 36
152 members in 16 offices
Priority claims9
| Document | Office | Kind | Date |
|---|---|---|---|
| 1020110115348 | Republic of Korea | – | |
| 20110115348 | Republic of Korea | A | |
| 20110115348 | Republic of Korea | A | |
| 2012084018 | China | W | |
| 2012084018 | China | W | |
| 1020110115348 | – | – | – |
| KR20110115348 | – | – | – |
| PCTCN2012084018 | – | – | – |
| WO2012CN84018 | – | – | – |
Members152
| Document | Office | Kind | |
|---|---|---|---|
| KR20130050149A | Republic of Korea | A | |
| CA2849029A1 | Canada | A1 | |
| CA2931709A1 | Canada | A1 | |
| CA2931745A1 | Canada | A1 | |
| CA3039421A1 | Canada | A1 | |
| CA3092392A1 | Canada | A1 | |
| WO2013067903A1 | World Intellectual Property Organization (WIPO) | A1 | |
| AU2012334553A1 | Australia | A1 | |
| SG11201400667QA | Singapore | A | |
| IL231707A0 | Israel | A0 | |
| IL231707D0 | Israel | D0 | |
| KR20140074949A | Republic of Korea | A | |
| KR20140077919A | Republic of Korea | A | |
| KR20140082702A | Republic of Korea | A | |
| EP2752007A1 | European Patent Office (EPO) | A1 | |
| MX2014003962A | Mexico | A | |
| CN104012094A | China | A | |
| US2014269926A1 | United States of America | A1 | |
| JP2014534753A | Japan | A | |
| US8982957B2 | United States of America | B2 | |
| US2015156510A1 | United States of America | A1 | |
| EP2752007A4 | European Patent Office (EPO) | A4 | |
| SG10201503437RA | Singapore | A | |
| SG10201503439UA | Singapore | A | |
| SG10201503443YA | Singapore | A | |
| SG10201503444WA | Singapore | A | |
| HK1199588A | Hong Kong, China | A | |
| HK1199588A1 | Hong Kong, China | A1 | |
| AU2012334553B2 | Australia | B2 | |
| US2015229950A1 | United States of America | A1 | |
| US2015229951A1 | United States of America | A1 | |
| US2015229952A1 | United States of America | A1 | |
| US2015229953A1 | United States of America | A1 | |
| CN104869400A | China | A | |
| CN104869401A | China | A | |
| CN104869402A | China | A | |
| IL239725A0 | Israel | A0 | |
| IL239725D0 | Israel | D0 | |
| IL239726A0 | Israel | A0 | |
| IL239726D0 | Israel | D0 | |
| IL239727A0 | Israel | A0 | |
| IL239727D0 | Israel | D0 | |
| IL239728A0 | Israel | A0 | |
| IL239728D0 | Israel | D0 | |
| CN104883568A | China | A | |
| CN104967847A | China | A | |
| AU2015249102A1 | Australia | A1 | |
| AU2015249103A1 | Australia | A1 | |
| AU2015249104A1 | Australia | A1 | |
| AU2015249105A1 | Australia | A1 | |
| RU2015131333A | Russian Federation | A | |
| JP5827412B2 | Japan | B2 | |
| RU2014113082A | Russian Federation | A | |
| JP2016028516A | Japan | A | |
| JP2016028517A | Japan | A | |
| JP2016028518A | Japan | A | |
| JP2016028519A | Japan | A | |
| US9351012B2 | United States of America | B2 | |
| MX339984B | Mexico | B | |
| RU2589866C2 | Russian Federation | C2 | |
| HK1214442A | Hong Kong, China | A | |
| HK1214442A1 | Hong Kong, China | A1 | |
| HK1214443A | Hong Kong, China | A | |
| HK1214443A1 | Hong Kong, China | A1 | |
| HK1214444A | Hong Kong, China | A | |
| HK1214444A1 | Hong Kong, China | A1 | |
| HK1214701A | Hong Kong, China | A | |
| HK1214701A1 | Hong Kong, China | A1 | |
| HK1214703A | Hong Kong, China | A | |
| HK1214703A1 | Hong Kong, China | A1 | |
| CA2849029C | Canada | C | |
| PH12015500839A1 | Philippines | A1 | |
| PH12015500839B1 | Philippines | B1 | |
| PH12015500841A1 | Philippines | A1 | |
| PH12015500841B1 | Philippines | B1 | |
| PH12015500843A1 | Philippines | A1 | |
| PH12015500843B1 | Philippines | B1 | |
| PH12015500844A1 | Philippines | A1 | |
| PH12015500844B1 | Philippines | B1 | |
| IL231707A | Israel | A | |
| JP6074475B2 | Japan | B2 | |
| JP6074476B2 | Japan | B2 | |
| JP6074477B2 | Japan | B2 | |
| JP6076438B2 | Japan | B2 | |
| US9615106B2 | United States of America | B2 | |
| BR112014007593A2 | Brazil | A2 | |
| MX347162BThis record | Mexico | B | |
| MX347163B | Mexico | B | |
| US9635384B2 | United States of America | B2 | |
| US9641860B2 | United States of America | B2 | |
| US9648343B2 | United States of America | B2 | |
| JP2017085646A | Japan | A | |
| RU2621966C1 | Russian Federation | C1 | |
| RU2621967C1 | Russian Federation | C1 | |
| RU2621970C1 | Russian Federation | C1 | |
| RU2621972C2 | Russian Federation | C2 | |
| US2017214934A1 | United States of America | A1 | |
| IL239725A | Israel | A | |
| AU2015249102B2 | Australia | B2 | |
| AU2015249103B2 | Australia | B2 |
Numbers
- Publication
- 347162
- Publication, DOCDB
- 347162
- Publication, EPODOC
- MX347162
- Application
- 2015004214
- Application, DOCDB
- 2015004214
- Application, EPODOC
- MX20150004214
Titles2
- Spanish
- MÉTODO PARA DECODIFICACIÓN DE DATOS DE VIDEO.
- English
- METHOD FOR DECODING VIDEO DATA.
Classification
- CPC, 29
- H04N19/52
- H04N19/117
- H04N19/523
- H04N19/105
- H04N19/122
- H04N19/126
- H04N19/13
- H04N19/139
- H04N19/159
- H04N19/176
- H04N19/182
- H04N19/196
- H04N19/463
- H04N19/521
- H04N19/587
- H04N19/61
- H04N19/615
- H04N19/635
- H04N19/80
- H04N19/172
- H04N19/517
- H04N19/56
- H04N19/513
- H04N19/119
- H04N19/157
- H04N19/82
- H04N19/59
- H04N19/124
- H04N19/129
- IPC, 7
- H04N19 105
- H04N19 117
- H04N19 122
- H04N19 126
- H04N19 159
- H04N19 176
- H04N19 182