WO2021060830A1 - 비디오 신호 처리 방법 및 장치 - Google Patents

비디오 신호 처리 방법 및 장치 Download PDF

Info

Publication number
WO2021060830A1
WO2021060830A1 PCT/KR2020/012866 KR2020012866W WO2021060830A1 WO 2021060830 A1 WO2021060830 A1 WO 2021060830A1 KR 2020012866 W KR2020012866 W KR 2020012866W WO 2021060830 A1 WO2021060830 A1 WO 2021060830A1
Authority
WO
WIPO (PCT)
Prior art keywords
block
motion information
current block
partition
prediction
Prior art date
Application number
PCT/KR2020/012866
Other languages
English (en)
French (fr)
Inventor
임성원
Original Assignee
주식회사 케이티
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 주식회사 케이티 filed Critical 주식회사 케이티
Publication of WO2021060830A1 publication Critical patent/WO2021060830A1/ko

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/119Adaptive subdivision aspects, e.g. subdivision of a picture into rectangular or non-rectangular coding blocks
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/103Selection of coding mode or of prediction mode
    • H04N19/105Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/17Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
    • H04N19/176Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/503Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
    • H04N19/51Motion estimation or motion compensation
    • H04N19/513Processing of motion vectors
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/593Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving spatial prediction techniques

Definitions

  • the present disclosure relates to a video signal processing method and apparatus.
  • High-resolution and high-quality images such as high definition (HD) images and ultra high definition (UHD) images is increasing in various application fields.
  • the higher the resolution and quality of the video data the higher the amount of data is compared to the existing video data. Therefore, when the video data is transmitted using a medium such as an existing wired/wireless broadband line or stored using an existing storage medium, the transmission cost and The storage cost will increase.
  • High-efficiency image compression techniques can be used to solve these problems that occur as image data becomes high-resolution and high-quality.
  • Inter-screen prediction technology that predicts pixel values included in the current picture from a picture before or after the current picture with image compression technology
  • intra prediction technology that predicts pixel values included in the current picture using pixel information in the current picture
  • Various technologies exist such as an entropy encoding technology that allocates a short code to a value with a high frequency of appearance and a long code to a value with a low frequency of appearance, and it is possible to effectively compress and transmit or store image data using this image compression technology.
  • An object of the present disclosure is to provide a method and apparatus for dividing a coding block into a plurality of prediction units in encoding/decoding a video signal.
  • An object of the present disclosure is to provide a method and apparatus for deriving a prediction sample based on a weighted sum operation after performing inter prediction on each of a plurality of prediction units in encoding/decoding a video signal.
  • the index may specify one of candidates in which an angle index specifying one of angle candidates of a dividing line dividing the current block and a distance index indicating a position of the dividing line in the current block are combined.
  • the video signal encoding method includes dividing a current block into a first partition and a second partition, determining first motion information for the first partition, and second motion information for the second partition. Determining, and deriving a final prediction sample of the current block based on a first prediction sample derived based on the first motion information and a second prediction sample derived based on the second motion information It may include.
  • an index indicating the partition type of the current block is encoded in the bitstream, and the index is an angular index specifying one of angle candidates of a dividing line dividing the current block and a position of the dividing line in the current block.
  • One of the candidates for which the distance index is combined may be specified.
  • the first motion information is obtained by a first merge index specifying one of N candidates included in a merge candidate list
  • the second motion information is It may be obtained by a second merge index specifying one of N-1 candidates excluding the merge candidate corresponding to the first merge index in the candidate list.
  • the final prediction sample is obtained based on a weighted sum operation between the first prediction sample and the second prediction sample, and a first weight applied to the first prediction sample and A second weight applied to the second prediction sample may be determined based on the angle index and the distance index.
  • the video signal decoding method further comprises storing motion information in units of sub-blocks of a predefined size in the current block, wherein the storage type of the sub-block is a representative position sample of the sub-block It can be determined based on the weight of.
  • the representative position sample may be located at the center of the sub-block.
  • the storage type includes: a first storage type for allocating the first motion information to the sub-block, a second storage type for allocating the second motion information to the sub-block, and It may be one of the third storage types in which bidirectional information derived by combining the first motion information and the second motion information is allocated to the sub-block.
  • coding/decoding efficiency can be improved by dividing a coding block into a plurality of prediction units.
  • the present disclosure improves prediction accuracy by performing inter prediction for each of a plurality of prediction units and then inducing prediction samples based on their weighted sum operation. I can.
  • FIG. 1 is a block diagram illustrating an image encoding apparatus according to an embodiment of the present disclosure.
  • FIG. 2 is a block diagram illustrating an image decoding apparatus according to an embodiment of the present disclosure.
  • 3 and 4 show an example in which inter prediction is performed.
  • FIG. 5 is a flowchart illustrating an inter prediction method according to an embodiment to which the present invention is applied.
  • FIG. 6 is a diagram for explaining a location of a spatial neighboring block referenced when inducing motion information of a current block.
  • FIG. 7 is an example for explaining a temporal neighboring block referenced to induce motion information of a current block.
  • FIG. 8 is a diagram illustrating a process of inducing motion information of a current block when a merge mode is applied to a current block.
  • FIG. 9 is a diagram illustrating a process of inducing motion information of a current block when the AMVP mode is applied to the current block.
  • 11 illustrates a case where there are 8 angle candidates.
  • 17 shows an aspect of applying weights according to angleIdx and distanceIdx.
  • 18 and 19 are diagrams for explaining a storage aspect of motion information.
  • 20 is a diagram for explaining a representative weight.
  • 23 and 24 illustrate an example of determining a prediction technique applied to each partition by referring to an encoding mode of a neighboring block.
  • 26 shows an example in which intra prediction information or inter prediction information is derived with reference to a block encoded by a mixed prediction method.
  • first and second may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another component. For example, without departing from the scope of the present disclosure, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element.
  • the term and/or includes a combination of a plurality of related listed items or any of a plurality of related listed items.
  • FIG. 1 is a block diagram illustrating an image encoding apparatus according to an embodiment of the present disclosure.
  • the image encoding apparatus 100 includes a picture splitter 110, a prediction unit 120, 125, a transform unit 130, a quantization unit 135, a rearrangement unit 160, and an entropy encoder ( 165, an inverse quantization unit 140, an inverse transform unit 145, a filter unit 150, and a memory 155.
  • each of the components shown in FIG. 1 is shown independently to represent different characteristic functions in an image encoding apparatus, and does not mean that each component is formed of separate hardware or a single software component. That is, each constituent part is listed and included as a constituent part for convenience of explanation, and at least two constituent parts of each constituent part are combined to form one constituent part, or one constituent part is divided into a plurality of constituent parts to perform functions Integrated embodiments and separate embodiments of the components are also included in the scope of the present disclosure unless departing from the essence of the present disclosure.
  • the components are not essential components that perform essential functions in the present disclosure, but may be optional components only for improving performance.
  • the present disclosure may be implemented by including only components essential to implement the essence of the present disclosure excluding components used for performance improvement, and a structure including only essential components excluding optional components used for performance improvement Also included in the scope of the present disclosure.
  • the picture dividing unit 110 may divide the input picture into at least one processing unit.
  • the processing unit may be a prediction unit (PU), a transform unit (TU), or a coding unit (CU).
  • the picture splitter 110 divides one picture into a combination of a plurality of coding units, prediction units, and transformation units, and combines one coding unit, a prediction unit, and a transformation unit based on a predetermined criterion (for example, a cost function). You can select to encode the picture.
  • a predetermined criterion for example, a cost function
  • one picture may be divided into a plurality of coding units.
  • a recursive tree structure such as a quad tree structure can be used. Encoding that is split into other coding units based on one image or the largest coding unit as a root. A unit may be divided with as many child nodes as the number of divided coding units. Coding units that are no longer split according to certain restrictions become leaf nodes. That is, when it is assumed that only square splitting is possible for one coding unit, one coding unit may be split into up to four different coding units.
  • a coding unit may be used as a unit that performs encoding or a unit that performs decoding.
  • the prediction unit may be split in a shape such as at least one square or rectangle of the same size within one coding unit, or one prediction unit among the prediction units split within one coding unit is another prediction. It may be divided to have a different shape and/or size from the unit.
  • intra prediction may be performed without dividing into a plurality of prediction units NxN.
  • the prediction units 120 and 125 may include an inter prediction unit 120 that performs inter prediction and an intra prediction unit 125 that performs intra prediction. It is possible to determine whether to use inter prediction or to perform intra prediction for the prediction unit, and determine specific information (eg, intra prediction mode, motion vector, reference picture, etc.) according to each prediction method.
  • a processing unit in which prediction is performed may be different from a processing unit in which a prediction method and specific content are determined. For example, a prediction method and a prediction mode are determined in a prediction unit, and prediction may be performed in a transformation unit. A residual value (residual block) between the generated prediction block and the original block may be input to the transform unit 130.
  • prediction mode information, motion vector information, and the like used for prediction may be encoded by the entropy encoder 165 together with a residual value and transmitted to a decoding apparatus.
  • prediction mode information, motion vector information, and the like used for prediction may be encoded by the entropy encoder 165 together with a residual value and transmitted to a decoding apparatus.
  • the inter prediction unit 120 may predict a prediction unit based on information of at least one picture of a picture before or after the current picture, and in some cases, predict based on information of a partial region in the current picture that has been encoded. You can also predict the unit.
  • the inter prediction unit 120 may include a reference picture interpolation unit, a motion prediction unit, and a motion compensation unit.
  • the reference picture interpolation unit may receive reference picture information from the memory 155 and may generate pixel information of an integer number of pixels or less from the reference picture.
  • a DCT-based 8-tap interpolation filter with different filter coefficients may be used to generate pixel information of an integer number of pixels or less in units of 1/4 pixels.
  • a DCT-based interpolation filter with different filter coefficients may be used to generate pixel information of an integer number of pixels or less in units of 1/8 pixels.
  • the motion prediction unit may perform motion prediction based on the reference picture interpolated by the reference picture interpolation unit.
  • Various methods such as a full search-based block matching algorithm (FBMA), a three step search (TSS), and a new three-step search algorithm (NTS), can be used as a method for calculating a motion vector.
  • the motion vector may have a motion vector value in units of 1/2 or 1/4 pixels based on the interpolated pixels.
  • the motion prediction unit may predict the current prediction unit by differently predicting the motion.
  • Various methods such as a skip method, a merge method, an advanced motion vector prediction (AMVP) method, and an intra block copy method may be used as a motion prediction method.
  • AMVP advanced motion vector prediction
  • the intra predictor 125 may generate a prediction unit based on reference pixel information around a current block, which is pixel information in the current picture. If the neighboring block of the current prediction unit is a block that has performed inter prediction and the reference pixel is a pixel that has performed inter prediction, a reference pixel included in the block that has performed inter prediction is a reference pixel of a block that has performed intra prediction around it. Can be used as a substitute for information. That is, when the reference pixel is not available, information about the reference pixel that is not available may be replaced with at least one reference pixel among the available reference pixels.
  • the prediction mode may have a directional prediction mode in which reference pixel information is used according to a prediction direction, and a non-directional mode in which directional information is not used when prediction is performed.
  • a mode for predicting luminance information and a mode for predicting color difference information may be different, and intra prediction mode information or predicted luminance signal information used to predict luminance information may be used to predict chrominance information.
  • intra prediction When performing intra prediction, if the size of the prediction unit and the size of the transformation unit are the same, intra prediction for the prediction unit is based on a pixel on the left, a pixel on the top left, and a pixel on the top of the prediction unit. You can do it. However, when the size of the prediction unit and the size of the transformation unit are different when performing intra prediction, intra prediction may be performed using a reference pixel based on the transformation unit. In addition, intra prediction using NxN splitting may be used for only the smallest coding unit.
  • a prediction block may be generated after applying an adaptive intra smoothing (AIS) filter to a reference pixel according to a prediction mode.
  • AIS adaptive intra smoothing
  • the type of AIS filter applied to the reference pixel may be different.
  • the intra prediction mode of the current prediction unit may be predicted from the intra prediction mode of the prediction unit existing around the current prediction unit.
  • mode information predicted from the neighboring prediction units if the intra prediction modes of the current prediction unit and the neighboring prediction units are the same, the current prediction unit and the neighboring prediction units are used using predetermined flag information.
  • Information indicating that the prediction mode of is the same may be transmitted, and if the prediction modes of the current prediction unit and the neighboring prediction units are different, entropy encoding may be performed to encode prediction mode information of the current block.
  • a residual block including a prediction unit that performs prediction based on a prediction unit generated by the prediction units 120 and 125 and residual information that is a difference value from the original block of the prediction unit may be generated.
  • the generated residual block may be input to the transform unit 130.
  • the transform unit 130 converts the original block and the residual block including residual information of the prediction unit generated through the prediction units 120 and 125 into a DCT (Discrete Cosine Transform), DST (Discrete Sine Transform), and KLT. You can convert it using the same conversion method. Whether to apply DCT, DST, or KLT to transform the residual block may be determined based on intra prediction mode information of a prediction unit used to generate the residual block.
  • DCT Discrete Cosine Transform
  • DST Discrete Sine Transform
  • KLT Discrete Sine Transform
  • the quantization unit 135 may quantize values converted by the transform unit 130 into the frequency domain. Quantization coefficients may vary depending on the block or the importance of the image. The value calculated by the quantization unit 135 may be provided to the inverse quantization unit 140 and the rearrangement unit 160.
  • the rearrangement unit 160 may rearrange coefficient values on the quantized residual values.
  • the rearrangement unit 160 may change the 2-dimensional block shape coefficient into a 1-dimensional vector shape through a coefficient scanning method. For example, the rearrangement unit 160 may scan from a DC coefficient to a coefficient in a high frequency region using a Zig-Zag Scan method, and change it into a one-dimensional vector form.
  • a vertical scan that scans a two-dimensional block shape coefficient in a column direction and a horizontal scan that scans a two-dimensional block shape coefficient in a row direction may be used. That is, according to the size of the transform unit and the intra prediction mode, it is possible to determine which scan method is to be used among zig-zag scan, vertical direction scan, and horizontal direction scan.
  • the entropy encoding unit 165 may perform entropy encoding based on values calculated by the rearrangement unit 160.
  • Entropy coding may use various coding methods such as Exponential Golomb, Context-Adaptive Variable Length Coding (CAVLC), and Context-Adaptive Binary Arithmetic Coding (CABAC).
  • the entropy encoding unit 165 includes residual value coefficient information and block type information of a coding unit, prediction mode information, division unit information, prediction unit information and transmission unit information, and motion from the rearrangement unit 160 and the prediction units 120 and 125.
  • Various information such as vector information, reference frame information, block interpolation information, and filtering information may be encoded.
  • the entropy encoder 165 may entropy-encode a coefficient value of a coding unit input from the reordering unit 160.
  • the inverse quantization unit 140 and the inverse transform unit 145 inverse quantize values quantized by the quantization unit 135 and inverse transform the values transformed by the transform unit 130.
  • the residual value generated by the inverse quantization unit 140 and the inverse transform unit 145 is reconstructed by being combined with the prediction units predicted through the motion estimation unit, motion compensation unit, and intra prediction unit included in the prediction units 120 and 125 Blocks (Reconstructed Block) can be created.
  • the filter unit 150 may include at least one of a deblocking filter, an offset correction unit, and an adaptive loop filter (ALF).
  • a deblocking filter may include at least one of a deblocking filter, an offset correction unit, and an adaptive loop filter (ALF).
  • ALF adaptive loop filter
  • the deblocking filter can remove block distortion caused by the boundary between blocks in the reconstructed picture.
  • it may be determined whether to apply the deblocking filter to the current block based on pixels included in several columns or rows included in the block.
  • a strong filter or a weak filter may be applied according to the required deblocking filtering strength.
  • horizontal filtering and vertical filtering may be processed in parallel when performing vertical filtering and horizontal filtering.
  • the offset correction unit may correct an offset from the original image on a pixel-by-pixel basis for the deblocking image.
  • the pixels included in the image are divided into a certain number of areas, and then the area to be offset is determined and the offset is applied to the area, or offset by considering the edge information of each pixel. You can use the method to apply.
  • Adaptive Loop Filtering may be performed based on a value obtained by comparing the filtered reconstructed image and the original image. After dividing the pixels included in the image into predetermined groups, one filter to be applied to the group may be determined, and filtering may be performed differentially for each group. Information related to whether to apply ALF may be transmitted for each coding unit (CU) of the luminance signal, and the shape and filter coefficient of the ALF filter to be applied may vary according to each block. In addition, the same type (fixed type) ALF filter may be applied regardless of the characteristics of the block to be applied.
  • ALF Adaptive Loop Filtering
  • the memory 155 may store the reconstructed block or picture calculated through the filter unit 150, and the stored reconstructed block or picture may be provided to the prediction units 120 and 125 when performing inter prediction.
  • FIG. 2 is a block diagram illustrating an image decoding apparatus according to an embodiment of the present disclosure.
  • the image decoding apparatus 200 includes an entropy decoding unit 210, a rearrangement unit 215, an inverse quantization unit 220, an inverse transform unit 225, prediction units 230 and 235, and a filter unit. 240) and a memory 245 may be included.
  • the input bitstream may be decoded in a procedure opposite to that of the image encoding apparatus.
  • the entropy decoding unit 210 may perform entropy decoding in a procedure opposite to that performed by the entropy encoding unit of the image encoding apparatus. For example, various methods such as Exponential Golomb, Context-Adaptive Variable Length Coding (CAVLC), and Context-Adaptive Binary Arithmetic Coding (CABAC) may be applied in response to the method performed by the image encoding apparatus.
  • various methods such as Exponential Golomb, Context-Adaptive Variable Length Coding (CAVLC), and Context-Adaptive Binary Arithmetic Coding (CABAC) may be applied in response to the method performed by the image encoding apparatus.
  • CAVLC Context-Adaptive Variable Length Coding
  • CABAC Context-Adaptive Binary Arithmetic Coding
  • the entropy decoder 210 may decode information related to intra prediction and inter prediction performed by the encoding apparatus.
  • the rearrangement unit 215 may perform rearrangement based on a method of rearranging the bitstream entropy-decoded by the entropy decoder 210 by the encoder.
  • the coefficients expressed in the form of a one-dimensional vector may be reconstructed into coefficients in the form of a two-dimensional block and rearranged.
  • the reordering unit 215 may perform reordering through a method of receiving information related to coefficient scanning performed by the encoder and performing reverse scanning based on the scanning order performed by the corresponding encoder.
  • the inverse quantization unit 220 may perform inverse quantization based on a quantization parameter provided by an encoding apparatus and a coefficient value of a rearranged block.
  • the inverse transform unit 225 may perform an inverse transform, that is, an inverse DCT, an inverse DST, and an inverse KLT, for transforms, that is, DCT, DST, and KLT, performed by the transform unit on the quantization result performed by the image encoding apparatus.
  • the inverse transformation may be performed based on a transmission unit determined by the image encoding apparatus.
  • the inverse transform unit 225 of the image decoding apparatus may selectively perform a transformation technique (eg, DCT, DST, KLT) according to a plurality of pieces of information such as a prediction method, a size of a current block, and a prediction direction.
  • a transformation technique eg, DCT, DST, KLT
  • the prediction units 230 and 235 may generate a prediction block based on information related to prediction block generation provided from the entropy decoder 210 and previously decoded block or picture information provided from the memory 245.
  • a pixel existing on the left side of the prediction unit, a pixel existing on the top left side, and Intra prediction is performed on a prediction unit based on an existing pixel, but when the size of the prediction unit and the size of the transformation unit are different when performing intra prediction, intra prediction is performed using a reference pixel based on the transformation unit. can do.
  • intra prediction using NxN splitting for only the smallest coding unit may be used.
  • the prediction units 230 and 235 may include a prediction unit determination unit, an inter prediction unit, and an intra prediction unit.
  • the prediction unit determining unit receives various information such as prediction unit information input from the entropy decoder 210, prediction mode information of the intra prediction method, motion prediction related information of the inter prediction method, etc., and classifies the prediction unit from the current coding unit, and makes predictions. It can be determined whether the unit performs inter prediction or intra prediction.
  • the inter prediction unit 230 uses information required for inter prediction of the current prediction unit provided by the video encoding apparatus, and based on information included in at least one picture of a previous picture or a subsequent picture of the current picture containing the current prediction unit, the current prediction unit 230 Inter prediction may be performed on the prediction unit. Alternatively, inter prediction may be performed based on information on a partial region previously-restored in the current picture including the current prediction unit.
  • the motion prediction method of the prediction unit included in the coding unit based on the coding unit is among the skip mode, merge mode, AMVP mode, and intra block copy mode. You can determine whether or not this is any way.
  • the intra prediction unit 235 may generate a prediction block based on pixel information in the current picture. If the prediction unit is a prediction unit that has performed intra prediction, intra prediction may be performed based on intra prediction mode information of the prediction unit provided by the image encoding apparatus.
  • the intra prediction unit 235 may include an adaptive intra smoothing (AIS) filter, a reference pixel interpolation unit, and a DC filter.
  • the AIS filter is a part that performs filtering on a reference pixel of the current block, and may determine whether to apply the filter according to the prediction mode of the current prediction unit and apply it.
  • AIS filtering may be performed on a reference pixel of the current block by using the prediction mode and AIS filter information of the prediction unit provided by the image encoding apparatus. When the prediction mode of the current block is a mode in which AIS filtering is not performed, the AIS filter may not be applied.
  • the reference pixel interpolator may interpolate the reference pixel to generate a reference pixel of a pixel unit having an integer value or less. If the prediction mode of the current prediction unit is a prediction mode in which a prediction block is generated without interpolating a reference pixel, the reference pixel may not be interpolated.
  • the DC filter may generate a prediction block through filtering when the prediction mode of the current block is the DC mode.
  • the reconstructed block or picture may be provided to the filter unit 240.
  • the filter unit 240 may include a deblocking filter, an offset correction unit, and an ALF.
  • the deblocking filter of the image decoding apparatus may receive information related to the deblocking filter provided by the image encoding apparatus, and the image decoding apparatus may perform deblocking filtering on a corresponding block.
  • the offset correction unit may perform offset correction on the reconstructed image based on the type of offset correction applied to the image during encoding and information on the offset value, and the like.
  • the ALF may be applied to a coding unit based on information on whether to apply ALF and information on ALF coefficients provided from an encoding device.
  • Such ALF information may be provided by being included in a specific parameter set.
  • the memory 245 may store the reconstructed picture or block so that it can be used as a reference picture or a reference block, and may also provide the reconstructed picture to an output unit.
  • a coding unit is used as a coding unit, but may be a unit that performs not only encoding but also decoding.
  • the current block represents a block to be encoded/decoded, and according to an encoding/decoding step, a coding tree block (or coding tree unit), a coding block (or coding unit), a transform block (or transform unit), or a prediction block (Or a prediction unit) or the like.
  • a coding tree block or coding tree unit
  • a coding block or coding unit
  • a transform block or transform unit
  • a prediction block Or a prediction unit
  • the image may be encoded/decoded in units of blocks. Coding blocks can be recursively partitioned based on a tree structure. For example, the coding block may be divided by at least one of quad tree division, binary tree division, or ternary tree division.
  • the coding block may be divided into a plurality of prediction blocks or a plurality of transform blocks.
  • the motion information may include at least one of a motion vector, a reference image index, and a prediction direction.
  • a block most similar to the current block may be searched, and the searched block may be set as a prediction block of the current block. Thereafter, a residual block may be generated by differentiating the current block and the prediction block.
  • 3 and 4 show an example in which inter prediction is performed.
  • FIG. 3 shows a result of searching for a prediction block spaced apart by a motion vector (x, y) from a collocated block located in the same location as a current block in a T-1 reference image.
  • a motion vector (x,y), a reference picture index indicating a reference picture T-1, and prediction direction information indicating that L0 direction prediction has been performed may be determined as motion information of the current block.
  • a reference block 0 spaced apart by a motion vector (x0, y0) is specified from a collocated block in the reference picture T-1.
  • a reference block 1 spaced apart by a motion vector (x1, y1) from the collocated block in the reference picture T+1 is specified.
  • the motion vector (x0, y0), the reference picture index indicating the reference picture T-1, and prediction direction information indicating that the L0 direction prediction has been performed may be determined as the L0 motion information of the current block. Further, a motion vector (x1, y1), a reference picture index indicating a reference picture T+1, and prediction direction information indicating that L1 direction prediction has been performed may be determined as the L1 motion information of the current block.
  • the final prediction block of the current block may be generated based on the sum of weights between the L0 prediction block and the L1 prediction block.
  • the L0 direction is the previous direction of the current picture
  • the L1 direction is the subsequent direction of the current picture.
  • both the L0 direction and the L1 direction may be set as a previous direction of the current picture, or both may be set as a direction after the current picture.
  • the L0 direction may be a direction after the current picture
  • the L1 direction may be set as a previous direction of the current picture.
  • a reference picture in a previous direction and a reference picture in a subsequent direction may be mixed in each of the L0 reference picture list and the L1 reference picture list.
  • motion information of a current block must be encoded and transmitted to a decoder.
  • the motion vector may be encoded as it is and transmitted to the decoder.
  • the amount of the motion vector may be reduced and transmitted to the decoder by using the predicted motion vector generated through the motion vector prediction process.
  • FIG. 5 is a flowchart illustrating an inter prediction method according to an embodiment to which the present invention is applied.
  • motion information of a current block may be determined (S510).
  • the motion information of the current block may include at least one of a motion vector for the current block, a reference picture index of the current block, and an inter prediction direction of the current block.
  • the motion information of the current block may be obtained based on at least one of information signaled through a bitstream or motion information of a spatial/temporal neighboring block of the current block.
  • FIG. 6 is a diagram for explaining a location of a spatial neighboring block referenced when inducing motion information of a current block.
  • LB indicates the position of the lower left sample in the current block
  • RT indicates the position of the upper right sample in the current block.
  • AX (X is 0 to 4) represents the reconstructed sample on the left of the current block
  • BY (Y is 0 to 5) represents the reconstructed sample at the top of the current block.
  • the location of at least one of the illustrated plurality of reconstructed samples may be used to determine a spatial neighboring block.
  • at least one of A0, A1, B0, B1, or B5 may be defined as a reference position for determining a spatial neighboring block.
  • FIG. 7 is an example for explaining a temporal neighboring block referenced to induce motion information of a current block.
  • CX (X is 0 to 35) denotes samples in a collocated block and reconstructed pixels around the collocated block.
  • the location of at least one of the illustrated plurality of reconstructed samples may be used to determine a spatial neighboring block.
  • at least one of C21 and C35 may be defined as a reference position for determining a spatial neighboring block.
  • FIG. 8 is a diagram illustrating a process of inducing motion information of a current block when a merge mode is applied to a current block.
  • a spatial merge candidate may be derived from a spatial neighboring block of the current block (S810).
  • the spatial neighboring block may include at least one of blocks adjacent to the top, left, or corner of the current block (eg, at least one of an upper left corner, an upper right corner, or a lower left corner) of the current block.
  • the motion information of the spatial merge candidate may be set the same as the motion information of the spatial neighboring block.
  • a temporal merge candidate may be derived from a temporal neighboring block of the current block (S820).
  • the temporal neighboring block may mean a co-located block (collocated block) included in a collocated picture.
  • the collocated picture has a temporal order (Picture Order Count, POC) different from the current picture including the current block.
  • the collocated picture may be determined as a picture having a predefined index in the reference picture list, or may be determined by an index signaled from a bitstream.
  • the temporal neighboring block may be determined as an arbitrary block in a block having the same position and size as the current block in the collocated picture or a block adjacent to a block having the same position and size as the current block. For example, at least one of a block including a center coordinate of a block having the same position and size as a current block in the collocated picture, or a block adjacent to a lower right boundary of the block may be determined as a temporal neighboring block.
  • Motion information of the temporal merge candidate may be determined based on motion information of a temporal neighboring block.
  • the motion vector of the temporal merge candidate may be determined based on the motion vector of the temporal neighboring block.
  • the inter prediction direction of the temporal merge candidate may be set to be the same as the inter prediction direction of the temporal neighboring block.
  • the reference picture index of the temporal merge candidate may have a fixed value. As an example, the reference picture index of the temporal merge candidate may be set to '0'.
  • a merge candidate list including the spatial merge candidate and the temporal merge candidate may be generated (S830). If the number of merge candidates included in the merge candidate list is less than the maximum number of merge candidates, a combined merge candidate combining two or more merge candidates or a merge candidate having a (0,0) zero motion vector It can be included in the merge candidate list.
  • motion information included in the global motion information buffer may be added to the merge candidate list as a merge candidate.
  • the global motion information buffer may store motion information of a picture including a current block, a slice, a coding tree unit line, or a block encoded/decoded before the current block in the coding tree unit. For example, motion information of blocks encoded by inter prediction from a first block in a predetermined area to a block before the current block may be stored in a global motion information buffer.
  • the number of motion information that can be stored in the motion information buffer may be limited to M pieces.
  • the closer to the current block, the higher priority may be on the wide area motion information buffer, and the farther from the current block, the lower priority may be on the wide area motion information buffer.
  • the distance from the current block may be determined by a difference in encoding/decoding order from the current block.
  • a higher priority may be given to a block encoded with a lower priority.
  • the block with the lowest priority i.e., the encoding/decoding order is the highest. After the motion information derived from the fast block) is deleted, motion information of a specific block can be added.
  • a wide motion information buffer may be generated for each area to be processed in parallel.
  • motion information of a block included in a parallel-processed area may not be added to the wide motion information buffer.
  • a wide area motion information buffer may be configured for each coding tree unit row. That is, the wide motion information buffer may be initialized for each coding tree unit.
  • preset initial motion information may be inserted into the wide motion information buffer.
  • the preset initial motion information may be global motion information applied to the entire picture.
  • Global motion information may be encoded through an upper header.
  • global motion information may be defined in units of tiles, bricks, or slices.
  • At least one of the merge candidates included in the merge candidate list may be specified based on the merge candidate index (S840).
  • the motion information of the current block may be set to be the same as the motion information of the merge candidate specified by the merge candidate index (S850). For example, when a spatial merge candidate is selected by the merge candidate index, motion information of a current block may be set to be the same as motion information of a spatial neighboring block. Alternatively, when a temporal merge candidate is selected by the merge candidate index, the motion information of the current block may be set to be the same as the motion information of the temporal neighboring block.
  • FIG. 9 is a diagram illustrating a process of inducing motion information of a current block when the AMVP mode is applied to the current block.
  • At least one of the inter prediction direction or the reference picture index of the current block may be decoded from the bitstream (S910). That is, when the AMVP mode is applied, at least one of the inter prediction direction or the reference picture index of the current block may be determined based on information encoded through the bitstream.
  • a spatial motion vector candidate may be determined based on a motion vector of a spatial neighboring block of the current block (S920).
  • the spatial motion vector candidate may include at least one of a first spatial motion vector candidate derived from an upper neighboring block of the current block and a second spatial motion vector candidate derived from a left neighboring block of the current block.
  • the upper neighboring block includes at least one of blocks adjacent to the upper or upper right corner of the current block
  • the left neighboring block of the current block includes at least one of the blocks adjacent to the left or lower left corner of the current block.
  • a block adjacent to the upper left corner of the current block may be treated as an upper neighboring block or a left neighboring block.
  • the spatial motion vector may be obtained by scaling the motion vector of the spatial neighboring block.
  • a temporal motion vector candidate may be determined based on a motion vector of a temporal neighboring block of the current block (S930).
  • the temporal motion vector may be obtained by scaling the motion vector of the temporal neighboring block.
  • a motion vector candidate list including a spatial motion vector candidate and a temporal motion vector candidate may be generated (S940).
  • At least one of the motion vector candidates included in the motion vector candidate list may be specified based on information specifying at least one of the motion vector candidate list (S950).
  • a motion vector candidate of the current block may be obtained by setting the motion vector candidate specified by the information as a motion vector predicted value of the current block and adding the motion vector difference value to the motion vector predicted value (S960).
  • the motion vector difference value may be parsed through a bitstream.
  • motion compensation for the current block may be performed based on the obtained motion information (S520). Specifically, motion compensation for the current block may be performed based on the inter prediction direction of the current block, a reference picture index, and a motion vector.
  • the merge candidate list under the merge mode and the motion vector candidate list under the AMVP mode will be referred to as a motion information list. That is, the motion information list may mean a merge candidate list or a motion vector candidate list according to a prediction mode applied to the current block.
  • the current block may be divided into a plurality of partitions (eg, prediction units). At this time, motion information may be derived for each divided partition. Accordingly, different motion information may be applied to each partition. Dividing the current block into two partitions can be referred to as diagonal division or geometric division.
  • the current block can be divided into two triangular partitions of equal size.
  • 10A is an example of division based on a division line at an angle of 45 degrees with respect to the center of the current block.
  • 10B is an example of division based on a division line at an angle of 135 degrees with respect to the center.
  • Information indicating the division direction of the current block may be encoded and signaled.
  • the information may indicate whether the angle of the dividing line dividing the current block is 45 degrees or 135 degrees.
  • a plurality of prediction blocks may be generated using motion information of each partition.
  • a first prediction block may be generated based on motion information of a first partition
  • a second prediction block may be generated based on motion information of a second partition.
  • each prediction unit may be set to use only unidirectional prediction without using bidirectional prediction.
  • each of the partitions can be set to have a different prediction direction. For example, when prediction in the L0 direction is performed in the first partition, it may be set to perform prediction in the L1 direction in the second partition.
  • the motion information of each partition can be encoded and transmitted to a decoder.
  • motion information of each partition may be determined by referring to motion information of an undistorted block around the current block. For example, by using the merge candidate list, an index specifying a candidate used to derive motion information of each partition may be encoded and signaled.
  • a merge candidate list of the first partition and a merge candidate list of the second partition may be separately generated.
  • the merge candidate list for the first partition may be configured to include merge candidates consisting of only L0 motion information
  • the merge candidate list for the second partition may be configured of merge candidates including only L1 motion information.
  • An index specifying one of merge candidates included in the merge candidate list of the first partition and an index specifying one of merge candidates included in the merge candidate list of the second partition may be encoded and signaled, respectively.
  • an integrated merge candidate list may be configured for the first partition and the second partition.
  • an index specifying a merge candidate for each partition may be signaled.
  • each partition when signaling an index for each partition, each partition may be set to refer to a different merge candidate.
  • the merge candidate specified by the first index for the first partition may be set not to be used to induce motion information of the second partition.
  • the unified merge candidate list may include merge candidates derived from spatial/temporal neighboring blocks.
  • the search order of merge candidates is A1, B1, B0, A0, B5, C35, C21, and the maximum number of merge candidates that the merge candidate list can include is N It is assumed to be personal.
  • spatial/temporal neighboring blocks are searched in the order listed above, and motion information of a spatial/temporal neighboring block for which L0 unidirectional prediction has been performed may be preferentially inserted into the merge candidate list. Nevertheless, if the number of merge candidates included in the merge candidate list is less than the maximum number N, the spatial/temporal neighboring blocks are searched again in the order listed above, and L0 among motion information of the spatial/temporal neighboring block for which bidirectional prediction is performed. Only motion information can be taken and inserted into the merge candidate list.
  • the L1 motion vector may be scaled based on the POC between the current picture and the reference picture of L0. That is, a motion vector derived through a weight sum operation between the L0 motion vector and the scaled L1 motion vector may be set as the merge candidate.
  • the prediction direction of the merge candidate may be set to L0, and the reference picture index may be set to a reference picture index present in the L0 motion information.
  • motion information or a zero vector stored in a global motion information buffer (eg, HMVP table) can be added to the merge candidate list.
  • the wide motion information buffer stores motion information of a block encoded/decoded before the current block.
  • a merge candidate list can also be constructed in a simplified way.
  • the simplified method after determining the maximum number of merge candidates that the merge candidate list may include, the maximum number of spatial merge candidates and the maximum number of temporal merge candidates may be set. In this case, the sum of the maximum number of spatial merge candidates and the maximum number of temporal merge candidates cannot exceed the maximum number of merge candidates that can be included in the merge candidate list.
  • the maximum number of merge candidates that can be included in the merge candidate list is six.
  • the spatial neighboring blocks and the temporal neighboring blocks have a search order of A1, B1, B0, A0, B5, C35, and C21.
  • a spatial merge candidate may be derived by referring to spatial neighboring blocks including each position. If at least one of the above four spatial neighboring blocks is not available as a merge candidate, a merge candidate list may be derived with reference to the spatial neighboring block including the B5 position.
  • a merge candidate derived from a temporal neighboring block including the C35 position may be inserted into the merge candidate list. If the temporal neighboring block at the above position is not available as a merge candidate, a merge candidate derived from the temporal neighboring block including the C21 position may be inserted into the merge candidate list.
  • motion information may be filled in the list with reference to the global motion information buffer.
  • a combined merge candidate may be generated by combining motion information of previously inserted merge candidates in the merge candidate list.
  • a combined merge candidate may be derived by averaging motion vectors between two merge candidates among merge candidates existing in the merge candidate list.
  • a combination [i,j] of merge candidates for generating the combined merge candidates may be predefined.
  • i indicates the index of the first merge candidate used to generate the combined merge candidate
  • j indicates the index of the second merge candidate used to generate the combined merge candidate.
  • the priority of the combination is [0,1], [0,2], [1,2], [1,3], [0,3], [1,3], [2,3] It can be defined as
  • a motion vector derived by averaging a motion vector of a merge candidate with an index of 0 in the merge candidate list and a motion vector of a merge candidate with an index of 1 is determined to be a merge candidate.
  • a merge candidate may be generated for each prediction direction. For example, for the [0,1] combination, a first combined merge candidate is generated by averaging the L0 motion vectors of two merge candidates, and a second combined merge candidate is generated by averaging the L1 motion vectors of the two merge candidates. Can be.
  • a case in which at least one of the merge candidates to be combined does not have motion information for a specific direction may occur.
  • bidirectional prediction may be performed on a merge candidate having an index of 0
  • L1-direction prediction may be performed on a merge candidate having an index of 1.
  • L0 information of the merge candidate having an index of 0 may be set as the motion information of the combined merge candidate.
  • whether to use a preset combination may be determined based on whether the merge candidate to be combined is predicted in one-way or in both directions.
  • the number of combinations for generating the combined merge candidate may be adjusted. For example, information indicating the number of combinations may be encoded through an upper header and transmitted to a decoder. Alternatively, a fixed number of combination candidates may be defined in the encoder and the decoder. For example, it may be defined that only one combination can be used in an encoder and a decoder. In this case, a combined merge candidate may be generated using only the [0,1] combination.
  • an additional merge candidate list may be newly generated based on the previously generated merge candidate list. Based on the newly generated additional merge candidate list, motion information of two partitions may be derived.
  • the additionally generated merge candidate list may be generated by extracting unidirectional motion information of merge candidates included in the existing merge candidate list.
  • merge candidates included in the existing merge candidate list may be sequentially scanned, and L0 motion information and L1 right position information may be alternately extracted.
  • L0 motion information of the first merge candidate existing in the existing merge candidate list is set as the first merge candidate in the additional merge candidate list
  • L1 motion information of the second merge candidate existing in the two existing merge candidate lists is added to the merge candidate list.
  • L0 motion information may be extracted from the 2N-1 th merge candidate
  • L1 motion information may be extracted from the 2N th merge candidate.
  • the maximum number of merge candidates that can be included in the additional merge candidate list may be equal to or smaller than the maximum number of merge candidates that can be included in the existing merge candidate list.
  • Information indicating the maximum number difference between the existing merge candidate list and the additional merge candidate list may be signaled through a bitstream.
  • an index specifying motion information of each partition may be encoded and signaled.
  • the merge candidate used for the first partition may be set to be unavailable for the second partition. Accordingly, when there are N merge candidates usable in the first partition, there may be N-1 merge candidates usable in the second partition. Accordingly, the range of the index value for the first partition may be determined as a value between 0 and N-1, and the range of the index value for the second partition may be determined as a value between 0 and N-2.
  • Information for determining the number N may be encoded and transmitted to a decoder.
  • the information may be transmitted through an upper header such as a sequence, a picture header, or a slice header.
  • N the value of N may be fixed in the encoder and the decoder. In the embodiments described below, it is assumed that the number N is 5.
  • the index of the candidate that matches the combination of merge candidates applied to the two partitions may be encoded and signaled.
  • the index of the candidate that matches the combination of the merge candidates applied to the two partitions and the partitioning direction of the current block is encoded and signaled. May be.
  • the merge candidate list includes at most 5 merge candidates, 40 combinations of merge candidates applied to two partitions and a dividing direction of a current block. (That is, the number of merge candidates applicable to the first sub-block 5 x the number of merge candidates applicable to the second sub-block 4 x the number of division direction candidates applicable to the current block 2) Accordingly, from 0 to 39 for the current block An index indicating one of the values between may be encoded and signaled.
  • the partitioning aspect of the current block is not limited to the illustrated example.
  • the angle ⁇ of the dividing line dividing the current block may have a quantized value between 0 degrees and 180 degrees.
  • the value of the angle ⁇ may be one of 0°, 45°, 90°, or 135°.
  • the range between 0 degrees and 180 may be divided into a larger number than this.
  • 11 illustrates a case where there are 8 angle candidates.
  • the eight angle candidates may be derived by dividing the range from 0 to 180 by eight.
  • the partitioning method overlapping the partitioning structure of the coding block may be excluded from the partitioning candidate.
  • the angle ⁇ of the dividing line is 90 degrees or 180 degrees, it results in the same splitting result as applying a binary tree splitting to a coding block. Accordingly, the angle ⁇ of 90 degrees and 180 degrees can be excluded from the segmentation candidate.
  • the partitions may be divided into non-uniform size.
  • represents the distance from the center of the current block to the dividing line.
  • Information specifying one of the plurality of partitioning candidates may be encoded and signaled.
  • an index specifying one of candidates obtained by combining a division angle and a division distance may be encoded and signaled.
  • a lookup table defining an index allocated to each of candidates obtained by combining a division angle and a division distance may be previously stored.
  • Table 1 is an example of a lookup table defining combinations of the split angle candidate and the split distance candidate shown in FIG. 13A and an index allocated to each of them.
  • merge_gpm_partition_idx represents an index allocated to a combination of an angle candidate and a distance candidate.
  • angleIdx is an index assigned to an angle candidate, and distanceIdx is an index assigned to a distance candidate.
  • the index merge_gpm_partition_idx specifying one of the partitioning candidates may be encoded and signaled.
  • An angle candidate and a distance candidate may be determined by the index merge_gpm_partition_idx, and a partition type of the current block may be determined based on the determined angle candidate and distance candidate.
  • distanceIdx indicates at least one of the distance from the center of the current block to the center of the dividing line or the position of the dividing line at a specific boundary.
  • the value of distanceIdx indicates that the distance from the center of the current block to the center of the dividing line is n/4 point, or indicates that the dividing line passes through the n/4 point on a specific boundary.
  • distnaceIdx when distnaceIdx is m, it indicates that the dividing line passes mW/4 or (4-m)W/4 of the upper boundary, or mH/4 or (4-m)H/4 of the left boundary.
  • W represents the width of the current block
  • H represents the height of the current block.
  • a combination of an angle candidate representing 0 degrees or 180 degrees and a distance candidate having a value of 0 may be excluded from the available split candidates in order to exclude a split type overlapping with the binary tree splitting.
  • Information specifying at least one of the number of angle candidates or the number of distance candidates may be encoded and signaled.
  • a quantization variable of the angle ⁇ of the dividing line eg, the number of partitions in a range from 0 to 180
  • the number of division distances may be encoded and signaled through an upper header.
  • the number and/or type of angle candidates and/or distance candidates may be determined with reference to a partitioning aspect of a neighboring block adjacent to the current block.
  • a high index priority can be given to a specific angle.
  • a combination candidate including an angle of 45 degrees or 135 degrees may be given a high priority.
  • the higher the priority the smaller the value of the index assigned to the corresponding combination.
  • the bit length of an index specifying the combination may decrease.
  • Whether it is allowed to divide the current block into two partitions may be determined by comparing at least one of the width and height of the current block with a threshold value. Specifically, in the case of applying the partitioning, if at least one of the width or height of the partition becomes less than or equal to the threshold value, partitioning the current block may not be allowed. For example, in consideration of the size of the partition for the chroma component, it may be determined whether it is allowed to divide the current block into two partitions.
  • an 8x8 luma block corresponds to a 4x4 chroma block.
  • the width or height of one of the partitions generated by dividing the chroma block may be 1.
  • the partition type may be set not to be applied to the current block.
  • the threshold value may be a natural number expressed as an exponent of 2, such as 1, 2, 4, or 8.
  • Information for determining the threshold value may be encoded and signaled. In this case, when the current image is 4:4:4, encoding of the information may be omitted.
  • the threshold value may be pre-stored in the encoder and the decoder.
  • a threshold value may be derived based on at least one of a chroma block division type (eg, a single tree or a dual tree) and/or a color difference format.
  • the number of available partitioning candidates may be adjusted based on at least one of the size or shape of the current block. For example, when the size of the current block is 8x8 or less, a partition candidate having a distance index of distanceIdx 1 to distanceIdx3 may be set to be unavailable. In this case, a lookup table in which the distanceIdx 1 to distanceIdx3dls partitioning candidates are excluded may be used as the distance index.
  • partition candidates having a distance index of distanceIdx 0 to distanceIdx 3 may be used.
  • the type or number of available angular indexes may be adjusted based on at least one of the size or shape of the current block.
  • a division type applicable to the current block may be determined based on at least one of a division type or a color difference format of the chroma block.
  • the luma block and the split form of the chroma block may be independently determined.
  • an index indicating a split type of a luma block and an index indicating an index of a chroma block may be independently encoded and signaled.
  • Information indicating whether the luma block and the split form of the chroma block are independently determined may be signaled through the bitstream.
  • the information may be a 1-bit flag.
  • the prediction sample of the current block may be derived based on a first prediction sample obtained based on motion information of a first partition and a second prediction sample obtained based on motion information of a second partition. Specifically, the prediction sample of the current block may be obtained based on a weighted sum operation of the first prediction sample and the second prediction sample.
  • a weight for each prediction sample may be determined based on the determined boundary and the size of the current block.
  • the weight includes a first weight applied to the first prediction sample and a second weight applied to the second prediction sample.
  • the weight may be an integer including 0.
  • Variables for determining the weight nW, nH, shift1, offset1, hwRatio, displacementX, displacementY, partFlip, shiftHor can be derived as shown in Table 2 below.
  • hwRatio nH / nW(6)
  • (angleIdx% 16 ! 0 && hwRatio> 0 ))? 0: 1
  • the variables nCbW and nCbH represent the width and height of the current block, respectively.
  • the variable cIdx represents a color component. For example, 0 may represent a luma component, and a value greater than 0 may represent a color difference component.
  • angleIdx represents an index specifying an angle candidate.
  • the variables SubWidthC and SubHeightC represent values determined according to the color difference format. For example, when the color difference format is 4:2:0, both SubWidthC and SubHeightC may be set to 2. If the color difference format is 4:4:4, both SubWidthC and SUbHeightC may be set to 1. If the color difference format is 4:2:2, SubWidthC may be set to 2 and SubHeightC may be set to 1.
  • Variables offsetX and offsetY may be determined based on the variables derived through Table 2. In this case, if the value of the variable shiftHor is 0, the values of offsetX and offsetY may be determined based on items (10) and (11) of Table 3 below. On the other hand, if the value of the variable shiftHor is 1, the values of offsetX and offsetY may be determined based on items (12) and (13) of Table 3.
  • variables xL and yL may be derived as shown in Table 4.
  • a temporary weight value, weightIdx may be derived using a preset lookup table and variables xL, offsetX, yL, and offsetY.
  • Table 5 shows an example of deriving the variable weightIdx.
  • weightIdx (((xL + offsetX) ⁇ 1) + 1) * disLut [displacementX] + (((yL + offsetY) ⁇ 1) + 1))
  • disLut represents a lookup table
  • displacementX represents an index specifying one in the lookup table.
  • the variable displacementX can be derived by angleIdx, as illustrated in Table 2.
  • Table 6 illustrates the lookup table disLut.
  • a weight wValue applied to the sample position (x, y) may be determined using the derived variables.
  • Table 7 shows an example in which the weight wValue is derived.
  • the first weight applied to the first prediction sample is set equal to the variable wValue
  • the second weight applied to the second prediction sample may be derived by differentiating the variable wValue from a predetermined constant value, for example, 8.
  • the variable wValue may be set as the second weight
  • the first weight may be derived by differentiating the variable wVlaue from a predetermined constant value, for example, 8.
  • Equation 1 shows an example of deriving a prediction sample of the current block.
  • Equation 1 P1 and P2 represent a first prediction sample and a second prediction sample, respectively.
  • ⁇ 1 and ⁇ 2 represent a first weight and a second weight, respectively.
  • FIGS. 14 to 16(a) show a first weight applied to each prediction sample
  • FIGS. 14 to 16(b) show a second weight applied to each prediction sample.
  • one of the first weight and the second weight may be set to 0, or both the first weight and the second weight may have a value other than 0 according to the position of the prediction target sample.
  • the prediction sample at the position (3, 1) in the current block may be derived by assigning a weight of 6 to the first prediction sample and assigning a weight of 2 to the second prediction sample.
  • a region including prediction samples having a value other than 0 in both the first weight and the second weight may be defined as a filter application region.
  • the size of the filter application area may be set smaller than that shown in FIG. 14. 15 shows an example of this.
  • a prediction sample may be derived without applying a filter within the current block. 16 is an example of this.
  • Information specifying the filter application region in the current block may be encoded and signaled. For example, when the current block is divided by a division line of 45 degrees or 135 degrees, an index specifying that the current block is applied among the examples shown in FIGS. 14 to 16 may be encoded and signaled.
  • information indicating the number of diagonal lines included in the filter application region may be encoded and signaled.
  • the information may be encoded as a value obtained by dividing the number of diagonal lines other than diagonal lines having the same x-axis and y-axis coordinates by two.
  • Information specifying the filter application region for the current block may be encoded and signaled for each color component.
  • an index for specifying a filter application region may be encoded.
  • a filter application area may be predefined for each color component.
  • the filter application region illustrated in FIG. 14 may be used, and for the chroma component, as in the example of FIG. 16, the filter application region may not be set.
  • the size/shape of the filter application area may be determined based on motion information of the first partition and the second partition.
  • the size/shape of the filter application region may be determined based on at least one of a POC difference between a reference picture of a first partition and a current picture or a POC difference between a reference picture of a second partition.
  • the size/shape of the filter application region may be determined based on whether a difference between the motion vector of the first partition and the motion vector of the second partition is greater than or equal to a threshold value.
  • the size/shape of the filter application area may be determined based on at least one of the division angle ⁇ and the division distance ⁇ .
  • 17 shows an aspect of applying weights according to angleIdx and distanceIdx.
  • 17 shows a value of a weight wValue for each prediction sample when angleIdx is 5 and distanceIdx is 0.
  • motion information may be stored in units of blocks having a predefined size.
  • the predefined size may be any one of 4x4, 8x8, 2x8, or 8x2.
  • information indicating the size of a unit in which motion information is stored may be encoded and signaled.
  • a unit in which motion information is stored has a size of 4x4.
  • a 4x4 size region in which motion information is stored will be referred to as a subblock.
  • motion information of a partition to which it belongs may be stored as motion information of the corresponding subblock.
  • a sub-block belongs to a division boundary e.g., a sub-block located on a division line
  • at least one of motion information of the first partition or motion information of the second partition may be stored as motion information of the sub-block.
  • 18 is a diagram for describing a storage aspect of motion information.
  • Motion information of the first partition may be stored in sub-blocks belonging to the first partition
  • motion information of the second partition may be stored in sub-blocks belonging to the second partition.
  • Motion information stored in sub-blocks positioned on a dividing line dividing the first partition and the second partition may be determined based on motion information of the first partition and the second partition.
  • bidirectional motion information obtained by combining motion information of the first partition and motion information of the second partition may be stored in subblocks.
  • either motion information of the first partition or motion information of the second partition may be stored in the subblock.
  • the motion information of the first partition may be fixedly set to be stored in the sub-block, or the motion information of the second partition may be fixedly set to be stored in the sub-block.
  • information indicating whether to store the motion information of the first partition or the motion information of the second partition in a subblock positioned on the dividing line may be encoded and signaled.
  • the division angle or division distance it may be determined whether to store motion information of the first partition or the motion information of the second partition in the sub-block. For example, when the division distance ⁇ is not 0, the sizes of the first partition and the second partition may be different. In this case, motion information of a partition having a larger size among the first partition and the second partition may be stored in a subblock positioned on the dividing line. Conversely, motion information of a partition having a smaller size among the first partition and the second partition may be stored in a subblock positioned on the dividing line.
  • motion information of a partition having a smaller POC difference value is obtained on a dividing line. It can be stored in a sub-block located in. Alternatively, motion information of a partition having a small reference picture index value among the first partition and the second partition may be stored in a subblock positioned on a dividing line.
  • the above-described method of storing motion information of sub-blocks located on the dividing line may be extendedly applied to sub-blocks included in the filter application area.
  • At least one of motion information of a first partition, motion information of a second partition, or a combination thereof may be stored in sub-blocks included in the filter application region.
  • a weight applied to a sample at a specific location within a sub-block may be set as a representative weight.
  • the specific position may be at least one of a center position, an upper left position, an upper right position, a lower left position, or a lower right position.
  • 20 is a diagram for explaining a representative weight.
  • a weight applied to a sample at a center position within a 4x4 subblock may be set as a representative weight.
  • a weight applied to a sample at a position (2, 2) in a sub-block may be set as a representative weight.
  • a variable weightIdx indicating a temporary weight value for a central position in the subblock may be derived. Equation 2 shows an example in which the temporary weight index weightIdx is derived.
  • Equation 2 (xSbIdx, ySbIdx) represents the index of a subblock in the current block.
  • the variable xSbIdx represents the horizontal direction index of the sub-block
  • the variable ySbIdx represents the vertical direction index of the sub-block.
  • the variable xSbIdx and the variable ySbIdx may be set to a value of 0 or 1, respectively.
  • the temporary weight value weightIdx derived based on Equation 2 may indicate the representative weight of the sub-block.
  • the motion information storage type of the sub-block may be determined by using the temporary weight index.
  • a variable sType indicating a storage type of motion information of a sub-block may be derived.
  • variable sType may be derived using Equation 4 below.
  • N may be 32 (that is, 0.5 is expressed when scaled down to a real position).
  • the method of deriving the variable sType may be determined based on at least one of the size or shape of the current block. For example, when the size of the current block is greater than or equal to the threshold value, the variable sType may be derived by Equation 3, and when the size of the current block is smaller than the threshold value, the variable sType may be derived by Equation 4.
  • the variable sType represents the motion information storage type.
  • the value of the variable sType may be set to a value of 0 to 2. For example, when the value of the variable sType is 0 or 1, it means that motion information of one of the two partitions is stored in the storage area. When the value of the variable sType is 2, it means that bidirectional motion information obtained by combining motion information of two partitions is stored in a sub-block.
  • variable sType can be derived using the temporary weight index for the center coordinates in each sub-block.
  • the storage type sType of the upper left sub-block and the lower right sub-block positioned on the dividing line is a value of 2. Accordingly, bidirectional motion information obtained by combining motion information of the first partition and motion information of the second partition may be stored in the upper left sub-block and the upper right sub-block.
  • the sType of the lower left sub-block not included in the boundary area was derived as 0.
  • motion information of a partition including the sub-block ie, a partition located below the current block
  • the sType of the upper right sub-block not included in the boundary area was derived as 1.
  • motion information of a partition including the sub-block ie, a partition positioned at the top of the current block
  • a variable sType of 2 may be defined as storing motion information of either the first partition or the second partition.
  • motion information of a partition having a larger area occupying the subblock may be stored.
  • the representative weight at the center position eg, (2, 2) position
  • the weight of the neighboring position eg, at least one of the left, right, top, or bottom
  • a variable sType of 2 may be set to store motion information of any one of the first partition or the second partition. Accordingly, the sub-block having the variable sType of 2 may store the same motion information as the sub-block having the variable sType of 0 or the same motion information as the sub-block having the variable sType of 1.
  • the motion information storage method may be adaptively determined based on at least one of the size or shape of the current block. For example, when the size of the current block is 8x8 or less, all sub-blocks in the current block may be set to store motion information stored in a sub-block at a preset location.
  • the sub-block at the preset position may be a sub-block having a size of 4x4 including samples at an upper left, upper right, lower left, lower right, or center position.
  • motion information stored in the sub-block at a preset location may be determined based on a variable sType for the sub-block.
  • the subblocks in the current block are set to store the same motion information, while when the size of the current block is larger than the threshold, the motions stored in each of the subblocks Information can be set to be determined individually.
  • a prediction method in which intra prediction and inter prediction are mixed may be applied.
  • intra prediction may be applied to one of the first partition and the second partition
  • inter prediction may be applied to the other. That is, different prediction methods can be applied between the two partitions.
  • the intra prediction mode may be derived with reference to the intra prediction mode of a neighboring block near the current block.
  • MPM candidates may be derived based on an intra prediction mode of a neighboring block.
  • the neighboring block may include at least one of an upper neighboring block or a left neighboring block.
  • the index of the MPM candidate that is the same as the intra prediction mode of the current block may be encoded and signaled.
  • a fixed intra prediction mode may be used.
  • the planar mode may be fixedly used.
  • the one having the highest frequency of use among intra prediction modes of neighboring blocks adjacent to the current block may be used.
  • one of the default modes may be used for intra prediction modes.
  • the default mode may include at least one of DC, planar, horizontal, vertical, or diagonal mode.
  • the configuration of the default mode may be different according to the division direction of the current block. For example, when the dividing line dividing the current block has an angle of 135 degrees, an intra prediction mode in the diagonal direction of the upper left corner may be set as a default mode. On the other hand, when the dividing line for dividing the current block has an angle of 45 degrees, an intra prediction mode in the lower left diagonal direction or the upper right diagonal direction may be set as the default mode.
  • a prediction technique (eg, intra prediction or inter prediction) applied to each partition may be determined in consideration of an encoding mode of a neighboring block.
  • the neighboring block may include at least one of an upper neighboring block or a left neighboring block.
  • 23 and 24 illustrate an example of determining a prediction technique applied to each partition by referring to an encoding mode of a neighboring block.
  • the angle of the dividing line dividing the current block is 45 degrees.
  • both the upper neighboring block and the left neighboring block are in contact with the boundary of the first partition. Accordingly, a prediction technique to be applied to the first partition may be determined by referring to the encoding modes of the upper neighboring block and the left neighboring block.
  • the same prediction technique as the coding mode may be applied to the first partition.
  • intra prediction may also be applied to the first partition.
  • inter prediction may be applied to the first partition.
  • the prediction technique having a high priority among intra prediction or inter prediction is the first. Can be applied to partitions.
  • information indicating a prediction technique applied to the first partition may be encoded and signaled.
  • a prediction technique different from that of the first partition may be applied to the second partition. For example, when intra prediction is applied to the first partition, inter prediction may be applied to the second partition. Conversely, when inter prediction is applied to the first partition, intra prediction may be applied to the second partition.
  • the upper neighboring block touches the boundary of the first partition
  • the left neighboring block touches the boundary of the second partition.
  • a prediction technique applied to the first partition may be derived based on the coding mode of the upper neighboring block, and the prediction technique applied to the second partition may be derived based on the coding mode of the left neighboring block.
  • the same prediction method as the upper neighboring block is used in the first partition.
  • the same prediction technique as the left neighboring block may be applied to the second partition.
  • intra prediction may be applied to a partition having a larger size among the first and second partitions, and inter prediction may be applied to a smaller partition.
  • an index specifying a combination of the prediction technique applied to the first partition and the prediction technique applied to the second partition may be encoded and signaled.
  • a filter may be applied to the boundary between partitions.
  • a method of determining a weight applied to each prediction sample is as described above.
  • the size of the filter area may be adjusted according to the block size.
  • a higher weight may be applied to a prediction sample derived by a prediction technique having a higher priority by setting a priority between inter prediction and intra prediction. For example, when inter prediction has a higher priority than intra prediction, a larger filter coefficient may be allocated to an area in which inter prediction is performed than an area in which intra prediction is performed.
  • a filter may be set not to be applied to the partition boundary (eg, applying the example of FIG. 16 ).
  • a deblocking filter between reconstructed blocks may be applied.
  • the encoding mode applied to the current block is regarded as intra prediction, and a boundary strength may be determined.
  • the mixed prediction method is applied to the current block, it is considered that the coding mode applied to the current block is inter prediction, and the boundary strength may be determined.
  • the strength of the boundary may be determined based on a prediction technique applied to a partition contacting the boundary to which the deblocking filter is applied among the two partitions.
  • the boundary strength may be determined based on a coding mode applied to a partition adjacent to the upper boundary of the current block. For example, in the example shown in (a) of FIG. 25, since intra prediction is applied to a partition adjacent to the upper boundary of the current block, it is considered that the encoding mode of the current block is intra prediction, and the boundary strength is determined. I can.
  • the boundary strength may be determined based on a coding mode applied to a partition adjacent to the right boundary of the current block. For example, in the example shown in (b) of FIG. 25, since inter prediction is applied to a partition adjacent to the right boundary of the current block, it is assumed that the encoding mode of the current block is inter prediction, and the boundary strength is determined. I can.
  • a block encoded by the mixed prediction method may have both intra prediction information and inter prediction information. Accordingly, a block encoded by the mixed prediction method may be referenced when deriving intra prediction information of a block to be encoded/decoded later, as well as when deriving inter prediction information.
  • 26 shows an example in which intra prediction information or inter prediction information is derived with reference to a block encoded by a mixed prediction method.
  • the mixed prediction method is applied to a block adjacent to the top of the current block, and inter prediction is applied to the current block.
  • motion information of the upper neighboring block may be referred to.
  • a merge candidate may be derived by referring to motion information of an upper neighboring block.
  • the intra prediction information of the left neighboring block may be referred to.
  • the MPM candidate of the current block may be derived based on the intra prediction mode of the left neighboring block.
  • the intra prediction mode of the neighboring block is a predefined intra-prediction mode, and an MPM candidate of the current block may be derived.
  • the predefined intra prediction mode may be a flat mode.
  • the mixed prediction method when the mixed prediction method is applied, it may be set to store only the inter prediction information and not to store the intra prediction information. Conversely, when the mixed prediction method is applied, only intra prediction information may be stored and inter prediction information may not be stored.
  • a residual block may be derived by differentiating a prediction block from an original block, and a transform coefficient may be derived by applying a transform to the residual block.
  • the quantized transform coefficients may be encoded.
  • the transformation may be applied only to a partial area within the current block.
  • the partial region is a region including the current block, and may include as many samples as the number expressed by an exponent of 2.
  • variable W represents the height of the current block
  • variable H represents the width of the current block.
  • bold lines indicate dividing lines for dividing the current block.
  • a reduced transform area including samples on a dividing line is set.
  • the shape of the reduced conversion area may be determined according to the position of the dividing line.
  • at least one of the width or height included in the reduced transform region may be reduced by an exponent of 2 from the width or height of the current block.
  • the reduced transform area may have a shape reduced by 1/2 of the width W and height H of the current block.
  • the reduced transform area may have a width of 1/2 of the width W of the current block and the same height as the height H of the current block. have.
  • the reduction ratio represents the size of the transformed area compared to the size of the current block.
  • an index specifying one of a plurality of reduction ratio candidates may be signaled.
  • Information indicating the reduction ratio may be signaled for each of the width and height.
  • first reduction ratio information indicating a reduction ratio of the width of the reduced transformation region and second reduction ratio information indicating a reduction ratio of the height of the reduced transformation region may be encoded and signaled.
  • a reduction ratio of the reduced transform region may be derived based on at least one of the angle ⁇ of the dividing line dividing the current block and the distance ⁇ of the dividing line.
  • the size of the reduced transformed area may be determined based on the row or column included in the area through which the dividing line passes.
  • a reduction ratio of the reduced transform region may be determined based on the size of the current block. For example, when the size of the current block is 8x8 or less, the width and height of the reduced conversion ratio may be reduced by a maximum of 1/2 of the width and height of the current block, respectively. On the other hand, when the size of the current block is larger than 8x8 (eg, 16x16 or 32x32), the width and height of the reduced conversion ratio may be reduced by a maximum of 1/4 of the width and height of the current block, respectively.
  • the transform may be applied to the reduced transform area, while the transform may not be applied to the remaining area except for the reduced transform area.
  • the residual value of the region to which the transformation is not applied may be set to 0.
  • Information indicating whether the reduced transformation region is used may be encoded and signaled.
  • whether to encode the information may be determined based on a value of a flag (eg, a coded block flag (CBF)) indicating whether a non-zero transform coefficient exists in the current block.
  • CBF coded block flag
  • the decoder may additionally parse information related to the reduced transform region.
  • the CBF when diagonal division is applied to the current block, the CBF may be set to indicate whether a reduced transform area is used. For example, when diagonal division is applied to the current block and the CBF value for the current block is 0, it may indicate that non-zero transform coefficients exist only in the reduced transform region.
  • the encoding mode of the current block it may be determined whether the reduced transform region is usable.
  • the encoding mode represents intra prediction or inter prediction.
  • a transformation method using a reduced transformation region may be set to be unavailable.
  • information related to the reduced transform region may be signaled.
  • the reduced transform area may be set only when the sizes of the partitions in the current block are different or when the sizes of the partitions are the same.
  • information indicating whether a conversion method based on a reduced conversion area is available may be signaled through an upper header.
  • the upper header includes at least one of a sequence, a picture header, and a slice header.
  • information related to the reduced transform region at the block level may be signaled.
  • information related to the reduced transformation region at the block level may not be signaled.
  • a transformation method based on the reduced transformation region may be determined whether or not a transformation method based on the reduced transformation region is available.
  • a transformation method based on a reduced transformation region may be used when the number of samples included in the current block is 64 or less. In this case, information on the reduced transformation area may be signaled.
  • a transformation method based on a reduced transformation region may be set to be unavailable. In this case, information on the reduced transformation region may not be encoded.
  • the reduced transform area is set in consideration of the division boundary.
  • second-order transform may not be performed.
  • second-order transformation may be selectively performed on the reduced transformation area.
  • information indicating whether the second-order transformation has been performed in the reduced transformation region may be encoded and signaled.
  • each of the components (eg, units, modules, etc.) constituting the block diagram in the above-described embodiment may be implemented as a hardware device or software, or a plurality of components are combined to form a single hardware device or software. It can also be implemented.
  • the above-described embodiments may be implemented in the form of program instructions that can be executed through various computer components and recorded in a computer-readable recording medium.
  • the computer-readable recording medium may include program instructions, data files, data structures, and the like alone or in combination.
  • Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magnetic-optical media such as floptical disks. media), and a hardware device specially configured to store and execute program instructions such as ROM, RAM, flash memory, and the like.
  • the hardware device may be configured to operate as one or more software modules to perform processing according to the present disclosure, and vice versa.
  • the present invention can be applied to an electronic device capable of encoding/decoding an image.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)

Abstract

본 개시에 따른 영상 복호화 방법은, 현재 블록을 제1 파티션 및 제2 파티션으로 분할되는 것으로 결정된 경우, 상기 현재 블록의 분할 형태를 나타내는 인덱스를 파싱하는 단계, 상기 제1 파티션에 대한 제1 움직임 정보를 결정하는 단계, 상기 제2 파티션에 대한 제2 움직임 정보를 결정하는 단계, 및 상기 제1 움직임 정보를 기초로 유도되는 제1 예측 샘플 및 상기 제2 움직임 정보를 기초로 유도되는 제2 예측 샘플에 기초하여, 상기 현재 블록의 최종 예측 샘플을 유도하는 단계를 포함할 수 있다. 이때, 상기 인덱스는, 상기 현재 블록을 분할하는 분할선의 각도 후보들 중 하나를 특정하는 각도 인덱스와 상기 현재 블록 내 상기 분할선의 위치를 나타내는 거리 인덱스가 조합된 후보들 중 하나를 특정할 수 있다.

Description

비디오 신호 처리 방법 및 장치
본 개시는 비디오 신호 처리 방법 및 장치에 관한 것이다.
최근 HD(High Definition) 영상 및 UHD(Ultra High Definition) 영상과 같은 고해상도, 고품질의 영상에 대한 수요가 다양한 응용 분야에서 증가하고 있다. 영상 데이터가 고해상도, 고품질이 될수록 기존의 영상 데이터에 비해 상대적으로 데이터량이 증가하기 때문에 기존의 유무선 광대역 회선과 같은 매체를 이용하여 영상 데이터를 전송하거나 기존의 저장 매체를 이용해 저장하는 경우, 전송 비용과 저장 비용이 증가하게 된다. 영상 데이터가 고해상도, 고품질화 됨에 따라 발생하는 이러한 문제들을 해결하기 위해서는 고효율의 영상 압축 기술들이 활용될 수 있다.
영상 압축 기술로 현재 픽쳐의 이전 또는 이후 픽쳐로부터 현재 픽쳐에 포함된 화소값을 예측하는 화면 간 예측 기술, 현재 픽쳐 내의 화소 정보를 이용하여 현재 픽쳐에 포함된 화소값을 예측하는 화면 내 예측 기술, 출현 빈도가 높은 값에 짧은 부호를 할당하고 출현 빈도가 낮은 값에 긴 부호를 할당하는 엔트로피 부호화 기술 등 다양한 기술이 존재하고 이러한 영상 압축 기술을 이용해 영상 데이터를 효과적으로 압축하여 전송 또는 저장할 수 있다.
한편, 고해상도 영상에 대한 수요가 증가함과 함께, 새로운 영상 서비스로서 입체 영상 컨텐츠에 대한 수요도 함께 증가하고 있다. 고해상도 및 초고해상도의 입체 영상 콘텐츠를 효과적으로 제공하기 위한 비디오 압축 기술에 대하여 논의가 진행되고 있다.
본 개시는 비디오 신호를 부호화/복호화함에 있어서, 코딩 블록을 복수의 예측 유닛들로 분할하는 방법 및 장치를 제공하는 것을 목적으로 한다.
본 개시는 비디오 신호를 부호화/복호화함에 있어서, 복수의 예측 유닛 각각에 대해 인터 예측을 수행한 뒤, 이들의 가중합 연산에 기초하여 예측 샘플을 유도하는 방법 및 장치를 제공하는 것을 목적으로 한다.
본 개시에서 이루고자 하는 기술적 과제들은 이상에서 언급한 기술적 과제들로 제한되지 않으며, 언급하지 않은 또 다른 기술적 과제들은 아래의 기재로부터 본 개시가 속하는 기술분야에서 통상의 지식을 가진 자에게 명확하게 이해될 수 있을 것이다.
본 개시에 따른 비디오 신호 복호화 방법은, 현재 블록을 제1 파티션 및 제2 파티션으로 분할되는 것으로 결정된 경우, 상기 현재 블록의 분할 형태를 나타내는 인덱스를 파싱하는 단계, 상기 제1 파티션에 대한 제1 움직임 정보를 결정하는 단계, 상기 제2 파티션에 대한 제2 움직임 정보를 결정하는 단계, 및 상기 제1 움직임 정보를 기초로 유도되는 제1 예측 샘플 및 상기 제2 움직임 정보를 기초로 유도되는 제2 예측 샘플에 기초하여, 상기 현재 블록의 최종 예측 샘플을 유도하는 단계를 포함할 수 있다. 이때, 상기 인덱스는, 상기 현재 블록을 분할하는 분할선의 각도 후보들 중 하나를 특정하는 각도 인덱스와 상기 현재 블록 내 상기 분할선의 위치를 나타내는 거리 인덱스가 조합된 후보들 중 하나를 특정할 수 있다.
본 개시에 따른 비디오 신호 부호화 방법은, 현재 블록을 제1 파티션 및 제2 파티션으로 분할하는 단계, 상기 제1 파티션에 대한 제1 움직임 정보를 결정하는 단계, 상기 제2 파티션에 대한 제2 움직임 정보를 결정하는 단계, 및 상기 제1 움직임 정보를 기초로 유도되는 제1 예측 샘플 및 상기 제2 움직임 정보를 기초로 유도되는 제2 예측 샘플에 기초하여, 상기 현재 블록의 최종 예측 샘플을 유도하는 단계를 포함할 수 있다. 이때, 상기 현재 블록의 분할 형태를 나타내는 인덱스가 비트스트림에 부호화되고, 상기 인덱스는, 상기 현재 블록을 분할하는 분할선의 각도 후보들 중 하나를 특정하는 각도 인덱스와 상기 현재 블록 내 상기 분할선의 위치를 나타내는 거리 인덱스가 조합된 후보들 중 하나를 특정할 수 있다.
본 개시에 따른 비디오 신호 복호화 방법에 있어서, 상기 제1 움직임 정보는, 머지 후보 리스트에 포함된 N 개의 후보들 중 하나를 특정하는 제1 머지 인덱스에 의해 획득되고, 상기 제2 움직임 정보는, 상기 머지 후보 리스트 내 상기 제1 머지 인덱스에 대응되는 머지 후보를 제외한 N-1 개의 후보들 중 하나를 특정하는 제2 머지 인덱스에 의해 획득될 수 있다.
본 개시에 따른 비디오 신호 복호화 방법에 있어서, 상기 최종 예측 샘플은, 상기 제1 예측 샘플 및 상기 제2 예측 샘플 간의 가중합 연산에 기초하여 획득되고, 상기 제1 예측 샘플에 적용되는 제1 가중치 및 상기 제2 예측 샘플에 적용되는 제2 가중치는, 상기 각도 인덱스 및 상기 거리 인덱스에 기초하여 결정될 수 있다.
본 개시에 따른 비디오 신호 복호화 방법에 있어서, 상기 현재 블록 내 기 정의된 크기의 서브 블록 단위로, 움직임 정보를 저장하는 단계를 더 포함하고, 서브 블록의 저장 타입은, 상기 서브 블록의 대표 위치 샘플의 가중치를 기초로 결정될 수 있다.
본 개시에 따른 비디오 신호 복호화 방법에 있어서, 상기 대표 위치 샘플은, 상기 서브 블록의 중앙에 위치할 수 있다.
본 개시에 따른 비디오 신호 복호화 방법에 있어서, 상기 저장 타입은, 상기 서브 블록에 상기 제1 움직임 정보를 할당하는 제1 저장 타입, 상기 서브 블록에 상기 제2 움직임 정보를 할당하는 제2 저장 타입 및 상기 서브 블록에 상기 제1 움직임 정보 및 상기 제2 움직임 정보를 조합하여 유도되는 양방향 정보를 할당하는 제3 저장 타입 중 하나일 수 있다.
본 개시에 의하면, 코딩 블록을 복수의 예측 유닛들로 분할함으로써, 부호화/복호화 효율을 향상시킬 수 있다.
본 개시에 의하면, 본 개시는 비디오 신호를 부호화/복호화함에 있어서, 복수의 예측 유닛 각각에 대해 인터 예측을 수행한 뒤, 이들의 가중합 연산에 기초하여 예측 샘플을 유도함으로써, 예측 정확도를 향상시킬 수 있다.
본 개시에서 얻을 수 있는 효과는 이상에서 언급한 효과들로 제한되지 않으며, 언급하지 않은 또 다른 효과들은 아래의 기재로부터 본 개시가 속하는 기술분야에서 통상의 지식을 가진 자에게 명확하게 이해될 수 있을 것이다.
도 1은 본 개시의 일실시예에 따른 영상 부호화 장치를 나타낸 블록도이다.
도 2는 본 개시의 일실시예에 따른 영상 복호화 장치를 나타낸 블록도이다.
도 3 및 도 4는 인터 예측이 수행되는 예를 나타낸 것이다.
도 5는 본 발명이 적용되는 일실시예로서, 인터 예측 방법을 도시한 순서도이다.
도 6은, 현재 블록의 움직임 정보 유도시 참조되는 공간적 이웃 블록의 위치를 설명하기 위한 도면이다.
도 7은 현재 블록의 움직임 정보를 유도하기 위해 참조되는 시간적 이웃 블록을 설명하기 위한 예시이다.
도 8는 현재 블록에 머지 모드가 적용되는 경우, 현재 블록의 움직임 정보를 유도하는 과정을 나타낸 도면이다.
도 9는 현재 블록에 AMVP 모드가 적용되는 경우, 현재 블록의 움직임 정보를 유도하는 과정을 나타낸 도면이다.
도 10은 현재 블록을 두개의 삼각 형태 파티션들로 분할하는 예를 나타낸 것이다.
도 11은 각도 후보가 8개인 경우를 예시한 것이다.
도 12는 비균등 분할이 적용되는 예를 나타낸다.
도 13은 분할 각도 후보 및 분할 거리 후보들을 예시한 것이다.
도 14 내지 도 16은 현재 블록이 135도 각도를 갖는 분할선을 이용하여 분할된 경우, 각 예측 샘플을 유도하기 위해 적용되는 필터 계수를 나타낸다.
도 17은 angleIdx 및 distanceIdx에 따른 가중치 적용 양상을 나타낸다.
도 18 및 도 19는 움직임 정보의 저장 양상을 설명하기 위한 도면이다.
도 20은 대표 가중치를 설명하기 위한 도면이다.
도 21은 각 서브 블록에 대한 변수 sType의 값을 나타낸다.
도 22는 혼합된 예측 방법이 적용되는 예를 나타낸다.
도 23 및 도 24는 이웃 블록의 부호화 모드를 참조하여, 각 파티션에 적용되는 예측 기법을 결정하는 예를 나타낸 것이다.
도 25는 경계 강도를 결정하는 예를 나타낸다.
도 26은 혼합된 예측 방법에 의해 부호화된 블록을 참조하여 화면 내 예측 정보 또는 화면 간 예측 정보가 유도되는 예를 나타낸다.
도 27은 변환이 적용되는 영역을 예시한 것이다.
본 개시는 다양한 변경을 가할 수 있고 여러 가지 실시예를 가질 수 있는 바, 특정 실시예들을 도면에 예시하고 상세한 설명에 상세하게 설명하고자 한다. 그러나, 이는 본 개시를 특정한 실시 형태에 대해 한정하려는 것이 아니며, 본 개시의 사상 및 기술 범위에 포함되는 모든 변경, 균등물 내지 대체물을 포함하는 것으로 이해되어야 한다. 각 도면을 설명하면서 유사한 참조부호를 유사한 구성요소에 대해 사용하였다.
제1, 제2 등의 용어는 다양한 구성요소들을 설명하는데 사용될 수 있지만, 상기 구성요소들은 상기 용어들에 의해 한정되어서는 안 된다. 상기 용어들은 하나의 구성요소를 다른 구성요소로부터 구별하는 목적으로만 사용된다. 예를 들어, 본 개시의 권리 범위를 벗어나지 않으면서 제1 구성요소는 제2 구성요소로 명명될 수 있고, 유사하게 제2 구성요소도 제1 구성요소로 명명될 수 있다. 및/또는 이라는 용어는 복수의 관련된 기재된 항목들의 조합 또는 복수의 관련된 기재된 항목들 중의 어느 항목을 포함한다.
어떤 구성요소가 다른 구성요소에 "연결되어" 있다거나 "접속되어"있다고 언급된 때에는, 그 다른 구성요소에 직접적으로 연결되어 있거나 또는 접속되어 있을 수도 있지만, 중간에 다른 구성요소가 존재할 수도 있다고 이해되어야 할 것이다. 반면에, 어떤 구성요소가 다른 구성요소에 "직접 연결되어"있다거나 "직접 접속되어"있다고 언급된 때에는, 중간에 다른 구성요소가 존재하지 않는 것으로 이해되어야 할 것이다.
본 출원에서 사용한 용어는 단지 특정한 실시예를 설명하기 위해 사용된 것으로, 본 개시를 한정하려는 의도가 아니다. 단수의 표현은 문맥상 명백하게 다르게 뜻하지 않는 한, 복수의 표현을 포함한다. 본 출원에서, "포함하다" 또는 "가지다" 등의 용어는 명세서상에 기재된 특징, 숫자, 단계, 동작, 구성요소, 부품 또는 이들을 조합한 것이 존재함을 지정하려는 것이지, 하나 또는 그 이상의 다른 특징들이나 숫자, 단계, 동작, 구성요소, 부품 또는 이들을 조합한 것들의 존재 또는 부가 가능성을 미리 배제하지 않는 것으로 이해되어야 한다.
이하, 첨부한 도면들을 참조하여, 본 개시의 바람직한 실시예를 보다 상세하게 설명하고자 한다. 이하, 도면상의 동일한 구성요소에 대해서는 동일한 참조부호를 사용하고 동일한 구성요소에 대해서 중복된 설명은 생략한다.
도 1은 본 개시의 일실시예에 따른 영상 부호화 장치를 나타낸 블록도이다.
도 1을 참조하면, 영상 부호화 장치(100)는 픽쳐 분할부(110), 예측부(120, 125), 변환부(130), 양자화부(135), 재정렬부(160), 엔트로피 부호화부(165), 역양자화부(140), 역변환부(145), 필터부(150) 및 메모리(155)를 포함할 수 있다.
도 1에 나타난 각 구성부들은 영상 부호화 장치에서 서로 다른 특징적인 기능들을 나타내기 위해 독립적으로 도시한 것으로, 각 구성부들이 분리된 하드웨어나 하나의 소프트웨어 구성단위로 이루어짐을 의미하지 않는다. 즉, 각 구성부는 설명의 편의상 각각의 구성부로 나열하여 포함한 것으로 각 구성부 중 적어도 두 개의 구성부가 합쳐져 하나의 구성부로 이루어지거나, 하나의 구성부가 복수개의 구성부로 나뉘어져 기능을 수행할 수 있고 이러한 각 구성부의 통합된 실시예 및 분리된 실시예도 본 개시의 본질에서 벗어나지 않는 한 본 개시의 권리범위에 포함된다.
또한, 일부의 구성 요소는 본 개시에서 본질적인 기능을 수행하는 필수적인 구성 요소는 아니고 단지 성능을 향상시키기 위한 선택적 구성 요소일 수 있다. 본 개시는 단지 성능 향상을 위해 사용되는 구성 요소를 제외한 본 개시의 본질을 구현하는데 필수적인 구성부만을 포함하여 구현될 수 있고, 단지 성능 향상을 위해 사용되는 선택적 구성 요소를 제외한 필수 구성 요소만을 포함한 구조도 본 개시의 권리범위에 포함된다.
픽쳐 분할부(110)는 입력된 픽쳐를 적어도 하나의 처리 단위로 분할할 수 있다. 이때, 처리 단위는 예측 단위(Prediction Unit: PU)일 수도 있고, 변환 단위(Transform Unit: TU)일 수도 있으며, 부호화 단위(Coding Unit: CU)일 수도 있다. 픽쳐 분할부(110)에서는 하나의 픽쳐에 대해 복수의 부호화 단위, 예측 단위 및 변환 단위의 조합으로 분할하고 소정의 기준(예를 들어, 비용 함수)으로 하나의 부호화 단위, 예측 단위 및 변환 단위 조합을 선택하여 픽쳐를 부호화 할 수 있다.
예를 들어, 하나의 픽쳐는 복수개의 부호화 단위로 분할될 수 있다. 픽쳐에서 부호화 단위를 분할하기 위해서는 쿼드 트리 구조(Quad Tree Structure)와 같은 재귀적인 트리 구조를 사용할 수 있는데 하나의 영상 또는 최대 크기 부호화 단위(largest coding unit)를 루트로 하여 다른 부호화 단위로 분할되는 부호화 유닛은 분할된 부호화 단위의 개수만큼의 자식 노드를 가지고 분할될 수 있다. 일정한 제한에 따라 더 이상 분할되지 않는 부호화 단위는 리프 노드가 된다. 즉, 하나의 코딩 유닛에 대하여 정방형 분할만이 가능하다고 가정하는 경우, 하나의 부호화 단위는 최대 4개의 다른 부호화 단위로 분할될 수 있다.
이하, 본 개시의 실시예에서는 부호화 단위는 부호화를 수행하는 단위의 의미로 사용할 수도 있고, 복호화를 수행하는 단위의 의미로 사용할 수도 있다.
예측 단위는 하나의 부호화 단위 내에서 동일한 크기의 적어도 하나의 정사각형 또는 직사각형 등의 형태를 가지고 분할된 것일 수도 있고, 하나의 부호화 단위 내에서 분할된 예측 단위 중 어느 하나의 예측 단위가 다른 하나의 예측 단위와 상이한 형태 및/또는 크기를 가지도록 분할된 것일 수도 있다.
부호화 단위를 기초로 인트라 예측을 수행하는 예측 단위를 생성시 최소 부호화 단위가 아닌 경우, 복수의 예측 단위 NxN 으로 분할하지 않고 인트라 예측을 수행할 수 있다.
예측부(120, 125)는 인터 예측을 수행하는 인터 예측부(120)와 인트라 예측을 수행하는 인트라 예측부(125)를 포함할 수 있다. 예측 단위에 대해 인터 예측을 사용할 것인지 또는 인트라 예측을 수행할 것인지를 결정하고, 각 예측 방법에 따른 구체적인 정보(예컨대, 인트라 예측 모드, 모션 벡터, 참조 픽쳐 등)를 결정할 수 있다. 이때, 예측이 수행되는 처리 단위와 예측 방법 및 구체적인 내용이 정해지는 처리 단위는 다를 수 있다. 예컨대, 예측의 방법과 예측 모드 등은 예측 단위로 결정되고, 예측의 수행은 변환 단위로 수행될 수도 있다. 생성된 예측 블록과 원본 블록 사이의 잔차값(잔차 블록)은 변환부(130)로 입력될 수 있다. 또한, 예측을 위해 사용한 예측 모드 정보, 모션 벡터 정보 등은 잔차값과 함께 엔트로피 부호화부(165)에서 부호화되어 복호화 장치에 전달될 수 있다. 특정한 부호화 모드를 사용할 경우, 예측부(120, 125)를 통해 예측 블록을 생성하지 않고, 원본 블록을 그대로 부호화하여 복호화부에 전송하는 것도 가능하다.
인터 예측부(120)는 현재 픽쳐의 이전 픽쳐 또는 이후 픽쳐 중 적어도 하나의 픽쳐의 정보를 기초로 예측 단위를 예측할 수도 있고, 경우에 따라서는 현재 픽쳐 내의 부호화가 완료된 일부 영역의 정보를 기초로 예측 단위를 예측할 수도 있다. 인터 예측부(120)는 참조 픽쳐 보간부, 모션 예측부, 움직임 보상부를 포함할 수 있다.
참조 픽쳐 보간부에서는 메모리(155)로부터 참조 픽쳐 정보를 제공받고 참조 픽쳐에서 정수 화소 이하의 화소 정보를 생성할 수 있다. 휘도 화소의 경우, 1/4 화소 단위로 정수 화소 이하의 화소 정보를 생성하기 위해 필터 계수를 달리하는 DCT 기반의 8탭 보간 필터(DCT-based Interpolation Filter)가 사용될 수 있다. 색차 신호의 경우 1/8 화소 단위로 정수 화소 이하의 화소 정보를 생성하기 위해 필터 계수를 달리하는 DCT 기반의 4탭 보간 필터(DCT-based Interpolation Filter)가 사용될 수 있다.
모션 예측부는 참조 픽쳐 보간부에 의해 보간된 참조 픽쳐를 기초로 모션 예측을 수행할 수 있다. 모션 벡터를 산출하기 위한 방법으로 FBMA(Full search-based Block Matching Algorithm), TSS(Three Step Search), NTS(New Three-Step Search Algorithm) 등 다양한 방법이 사용될 수 있다. 모션 벡터는 보간된 화소를 기초로 1/2 또는 1/4 화소 단위의 모션 벡터값을 가질 수 있다. 모션 예측부에서는 모션 예측 방법을 다르게 하여 현재 예측 단위를 예측할 수 있다. 모션 예측 방법으로 스킵(Skip) 방법, 머지(Merge) 방법, AMVP(Advanced Motion Vector Prediction) 방법, 인트라 블록 카피(Intra Block Copy) 방법 등 다양한 방법이 사용될 수 있다.
인트라 예측부(125)는 현재 픽쳐 내의 화소 정보인 현재 블록 주변의 참조 픽셀 정보를 기초로 예측 단위를 생성할 수 있다. 현재 예측 단위의 주변 블록이 인터 예측을 수행한 블록이어서, 참조 픽셀이 인터 예측을 수행한 픽셀일 경우, 인터 예측을 수행한 블록에 포함되는 참조 픽셀을 주변의 인트라 예측을 수행한 블록의 참조 픽셀 정보로 대체하여 사용할 수 있다. 즉, 참조 픽셀이 가용하지 않는 경우, 가용하지 않은 참조 픽셀 정보를 가용한 참조 픽셀 중 적어도 하나의 참조 픽셀로 대체하여 사용할 수 있다.
인트라 예측에서 예측 모드는 참조 픽셀 정보를 예측 방향에 따라 사용하는 방향성 예측 모드와 예측을 수행시 방향성 정보를 사용하지 않는 비방향성 모드를 가질 수 있다. 휘도 정보를 예측하기 위한 모드와 색차 정보를 예측하기 위한 모드가 상이할 수 있고, 색차 정보를 예측하기 위해 휘도 정보를 예측하기 위해 사용된 인트라 예측 모드 정보 또는 예측된 휘도 신호 정보를 활용할 수 있다.
인트라 예측을 수행할 때 예측 단위의 크기와 변환 단위의 크기가 동일할 경우, 예측 단위의 좌측에 존재하는 픽셀, 좌측 상단에 존재하는 픽셀, 상단에 존재하는 픽셀을 기초로 예측 단위에 대한 인트라 예측을 수행할 수 있다. 그러나 인트라 예측을 수행할 때 예측 단위의 크기와 변환 단위의 크기가 상이할 경우, 변환 단위를 기초로 한 참조 픽셀을 이용하여 인트라 예측을 수행할 수 있다. 또한, 최소 부호화 단위에 대해서만 NxN 분할을 사용하는 인트라 예측을 사용할 수 있다.
인트라 예측 방법은 예측 모드에 따라 참조 화소에 AIS(Adaptive Intra Smoothing) 필터를 적용한 후 예측 블록을 생성할 수 있다. 참조 화소에 적용되는 AIS 필터의 종류는 상이할 수 있다. 인트라 예측 방법을 수행하기 위해 현재 예측 단위의 인트라 예측 모드는 현재 예측 단위의 주변에 존재하는 예측 단위의 인트라 예측 모드로부터 예측할 수 있다. 주변 예측 단위로부터 예측된 모드 정보를 이용하여 현재 예측 단위의 예측 모드를 예측하는 경우, 현재 예측 단위와 주변 예측 단위의 인트라 예측 모드가 동일하면 소정의 플래그 정보를 이용하여 현재 예측 단위와 주변 예측 단위의 예측 모드가 동일하다는 정보를 전송할 수 있고, 만약 현재 예측 단위와 주변 예측 단위의 예측 모드가 상이하면 엔트로피 부호화를 수행하여 현재 블록의 예측 모드 정보를 부호화할 수 있다.
또한, 예측부(120, 125)에서 생성된 예측 단위를 기초로 예측을 수행한 예측 단위와 예측 단위의 원본 블록과 차이값인 잔차값(Residual) 정보를 포함하는 잔차 블록이 생성될 수 있다. 생성된 잔차 블록은 변환부(130)로 입력될 수 있다.
변환부(130)에서는 원본 블록과 예측부(120, 125)를 통해 생성된 예측 단위의 잔차값(residual)정보를 포함한 잔차 블록을 DCT(Discrete Cosine Transform), DST(Discrete Sine Transform), KLT와 같은 변환 방법을 사용하여 변환시킬 수 있다. 잔차 블록을 변환하기 위해 DCT를 적용할지, DST를 적용할지 또는 KLT를 적용할지는 잔차 블록을 생성하기 위해 사용된 예측 단위의 인트라 예측 모드 정보를 기초로 결정할 수 있다.
양자화부(135)는 변환부(130)에서 주파수 영역으로 변환된 값들을 양자화할 수 있다. 블록에 따라 또는 영상의 중요도에 따라 양자화 계수는 변할 수 있다. 양자화부(135)에서 산출된 값은 역양자화부(140)와 재정렬부(160)에 제공될 수 있다.
재정렬부(160)는 양자화된 잔차값에 대해 계수값의 재정렬을 수행할 수 있다.
재정렬부(160)는 계수 스캐닝(Coefficient Scanning) 방법을 통해 2차원의 블록 형태 계수를 1차원의 벡터 형태로 변경할 수 있다. 예를 들어, 재정렬부(160)에서는 지그-재그 스캔(Zig-Zag Scan)방법을 이용하여 DC 계수부터 고주파수 영역의 계수까지 스캔하여 1차원 벡터 형태로 변경시킬 수 있다. 변환 단위의 크기 및 인트라 예측 모드에 따라 지그-재그 스캔 대신 2차원의 블록 형태 계수를 열 방향으로 스캔하는 수직 스캔, 2차원의 블록 형태 계수를 행 방향으로 스캔하는 수평 스캔이 사용될 수도 있다. 즉, 변환 단위의 크기 및 인트라 예측 모드에 따라 지그-재그 스캔, 수직 방향 스캔 및 수평 방향 스캔 중 어떠한 스캔 방법이 사용될지 여부를 결정할 수 있다.
엔트로피 부호화부(165)는 재정렬부(160)에 의해 산출된 값들을 기초로 엔트로피 부호화를 수행할 수 있다. 엔트로피 부호화는 예를 들어, 지수 골롬(Exponential Golomb), CAVLC(Context-Adaptive Variable Length Coding), CABAC(Context-Adaptive Binary Arithmetic Coding)과 같은 다양한 부호화 방법을 사용할 수 있다.
엔트로피 부호화부(165)는 재정렬부(160) 및 예측부(120, 125)로부터 부호화 단위의 잔차값 계수 정보 및 블록 타입 정보, 예측 모드 정보, 분할 단위 정보, 예측 단위 정보 및 전송 단위 정보, 모션 벡터 정보, 참조 프레임 정보, 블록의 보간 정보, 필터링 정보 등 다양한 정보를 부호화할 수 있다.
엔트로피 부호화부(165)에서는 재정렬부(160)에서 입력된 부호화 단위의 계수값을 엔트로피 부호화할 수 있다.
역양자화부(140) 및 역변환부(145)에서는 양자화부(135)에서 양자화된 값들을 역양자화하고 변환부(130)에서 변환된 값들을 역변환한다. 역양자화부(140) 및 역변환부(145)에서 생성된 잔차값(Residual)은 예측부(120, 125)에 포함된 움직임 추정부, 움직임 보상부 및 인트라 예측부를 통해서 예측된 예측 단위와 합쳐져 복원 블록(Reconstructed Block)을 생성할 수 있다.
필터부(150)는 디블록킹 필터, 오프셋 보정부, ALF(Adaptive Loop Filter)중 적어도 하나를 포함할 수 있다.
디블록킹 필터는 복원된 픽쳐에서 블록간의 경계로 인해 생긴 블록 왜곡을 제거할 수 있다. 디블록킹을 수행할지 여부를 판단하기 위해 블록에 포함된 몇 개의 열 또는 행에 포함된 픽셀을 기초로 현재 블록에 디블록킹 필터 적용할지 여부를 판단할 수 있다. 블록에 디블록킹 필터를 적용하는 경우 필요한 디블록킹 필터링 강도에 따라 강한 필터(Strong Filter) 또는 약한 필터(Weak Filter)를 적용할 수 있다. 또한 디블록킹 필터를 적용함에 있어 수직 필터링 및 수평 필터링 수행시 수평 방향 필터링 및 수직 방향 필터링이 병행 처리되도록 할 수 있다.
오프셋 보정부는 디블록킹을 수행한 영상에 대해 픽셀 단위로 원본 영상과의 오프셋을 보정할 수 있다. 특정 픽쳐에 대한 오프셋 보정을 수행하기 위해 영상에 포함된 픽셀을 일정한 수의 영역으로 구분한 후 오프셋을 수행할 영역을 결정하고 해당 영역에 오프셋을 적용하는 방법 또는 각 픽셀의 에지 정보를 고려하여 오프셋을 적용하는 방법을 사용할 수 있다.
ALF(Adaptive Loop Filtering)는 필터링한 복원 영상과 원래의 영상을 비교한 값을 기초로 수행될 수 있다. 영상에 포함된 픽셀을 소정의 그룹으로 나눈 후 해당 그룹에 적용될 하나의 필터를 결정하여 그룹마다 차별적으로 필터링을 수행할 수 있다. ALF를 적용할지 여부에 관련된 정보는 휘도 신호는 부호화 단위(Coding Unit, CU) 별로 전송될 수 있고, 각각의 블록에 따라 적용될 ALF 필터의 모양 및 필터 계수는 달라질 수 있다. 또한, 적용 대상 블록의 특성에 상관없이 동일한 형태(고정된 형태)의 ALF 필터가 적용될 수도 있다.
메모리(155)는 필터부(150)를 통해 산출된 복원 블록 또는 픽쳐를 저장할 수 있고, 저장된 복원 블록 또는 픽쳐는 인터 예측을 수행 시 예측부(120, 125)에 제공될 수 있다.
도 2는 본 개시의 일실시예에 따른 영상 복호화 장치를 나타낸 블록도이다.
도 2를 참조하면, 영상 복호화 장치(200)는 엔트로피 복호화부(210), 재정렬부(215), 역양자화부(220), 역변환부(225), 예측부(230, 235), 필터부(240), 메모리(245)가 포함될 수 있다.
영상 부호화 장치에서 영상 비트스트림이 입력된 경우, 입력된 비트스트림은 영상 부호화 장치와 반대의 절차로 복호화될 수 있다.
엔트로피 복호화부(210)는 영상 부호화 장치의 엔트로피 부호화부에서 엔트로피 부호화를 수행한 것과 반대의 절차로 엔트로피 복호화를 수행할 수 있다. 예를 들어, 영상 부호화 장치에서 수행된 방법에 대응하여 지수 골롬(Exponential Golomb), CAVLC(Context-Adaptive Variable Length Coding), CABAC(Context-Adaptive Binary Arithmetic Coding)과 같은 다양한 방법이 적용될 수 있다.
엔트로피 복호화부(210)에서는 부호화 장치에서 수행된 인트라 예측 및 인터 예측에 관련된 정보를 복호화할 수 있다.
재정렬부(215)는 엔트로피 복호화부(210)에서 엔트로피 복호화된 비트스트림을 부호화부에서 재정렬한 방법을 기초로 재정렬을 수행할 수 있다. 1차원 벡터 형태로 표현된 계수들을 다시 2차원의 블록 형태의 계수로 복원하여 재정렬할 수 있다. 재정렬부(215)에서는 부호화부에서 수행된 계수 스캐닝에 관련된 정보를 제공받고 해당 부호화부에서 수행된 스캐닝 순서에 기초하여 역으로 스캐닝하는 방법을 통해 재정렬을 수행할 수 있다.
역양자화부(220)는 부호화 장치에서 제공된 양자화 파라미터와 재정렬된 블록의 계수값을 기초로 역양자화를 수행할 수 있다.
역변환부(225)는 영상 부호화 장치에서 수행한 양자화 결과에 대해 변환부에서 수행한 변환 즉, DCT, DST, 및 KLT에 대해 역변환 즉, 역 DCT, 역 DST 및 역 KLT를 수행할 수 있다. 역변환은 영상 부호화 장치에서 결정된 전송 단위를 기초로 수행될 수 있다. 영상 복호화 장치의 역변환부(225)에서는 예측 방법, 현재 블록의 크기 및 예측 방향 등 복수의 정보에 따라 변환 기법(예를 들어, DCT, DST, KLT)이 선택적으로 수행될 수 있다.
예측부(230, 235)는 엔트로피 복호화부(210)에서 제공된 예측 블록 생성 관련 정보와 메모리(245)에서 제공된 이전에 복호화된 블록 또는 픽쳐 정보를 기초로 예측 블록을 생성할 수 있다.
전술한 바와 같이 영상 부호화 장치에서의 동작과 동일하게 인트라 예측을 수행시 예측 단위의 크기와 변환 단위의 크기가 동일할 경우, 예측 단위의 좌측에 존재하는 픽셀, 좌측 상단에 존재하는 픽셀, 상단에 존재하는 픽셀을 기초로 예측 단위에 대한 인트라 예측을 수행하지만, 인트라 예측을 수행시 예측 단위의 크기와 변환 단위의 크기가 상이할 경우, 변환 단위를 기초로 한 참조 픽셀을 이용하여 인트라 예측을 수행할 수 있다. 또한, 최소 부호화 단위에 대해서만 NxN 분할을 사용하는 인트라 예측을 사용할 수도 있다.
예측부(230, 235)는 예측 단위 판별부, 인터 예측부 및 인트라 예측부를 포함할 수 있다. 예측 단위 판별부는 엔트로피 복호화부(210)에서 입력되는 예측 단위 정보, 인트라 예측 방법의 예측 모드 정보, 인터 예측 방법의 모션 예측 관련 정보 등 다양한 정보를 입력 받고 현재 부호화 단위에서 예측 단위를 구분하고, 예측 단위가 인터 예측을 수행하는지 아니면 인트라 예측을 수행하는지 여부를 판별할 수 있다. 인터 예측부(230)는 영상 부호화 장치에서 제공된 현재 예측 단위의 인터 예측에 필요한 정보를 이용해 현재 예측 단위가 포함된 현재 픽쳐의 이전 픽쳐 또는 이후 픽쳐 중 적어도 하나의 픽쳐에 포함된 정보를 기초로 현재 예측 단위에 대한 인터 예측을 수행할 수 있다. 또는, 현재 예측 단위가 포함된 현재 픽쳐 내에서 기-복원된 일부 영역의 정보를 기초로 인터 예측을 수행할 수도 있다.
인터 예측을 수행하기 위해 부호화 단위를 기준으로 해당 부호화 단위에 포함된 예측 단위의 모션 예측 방법이 스킵 모드(Skip Mode), 머지 모드(Merge 모드), AMVP 모드(AMVP Mode), 인트라 블록 카피 모드 중 어떠한 방법인지 여부를 판단할 수 있다.
인트라 예측부(235)는 현재 픽쳐 내의 화소 정보를 기초로 예측 블록을 생성할 수 있다. 예측 단위가 인트라 예측을 수행한 예측 단위인 경우, 영상 부호화 장치에서 제공된 예측 단위의 인트라 예측 모드 정보를 기초로 인트라 예측을 수행할 수 있다. 인트라 예측부(235)에는 AIS(Adaptive Intra Smoothing) 필터, 참조 화소 보간부, DC 필터를 포함할 수 있다. AIS 필터는 현재 블록의 참조 화소에 필터링을 수행하는 부분으로써 현재 예측 단위의 예측 모드에 따라 필터의 적용 여부를 결정하여 적용할 수 있다. 영상 부호화 장치에서 제공된 예측 단위의 예측 모드 및 AIS 필터 정보를 이용하여 현재 블록의 참조 화소에 AIS 필터링을 수행할 수 있다. 현재 블록의 예측 모드가 AIS 필터링을 수행하지 않는 모드일 경우, AIS 필터는 적용되지 않을 수 있다.
참조 화소 보간부는 예측 단위의 예측 모드가 참조 화소를 보간한 화소값을 기초로 인트라 예측을 수행하는 예측 단위일 경우, 참조 화소를 보간하여 정수값 이하의 화소 단위의 참조 화소를 생성할 수 있다. 현재 예측 단위의 예측 모드가 참조 화소를 보간하지 않고 예측 블록을 생성하는 예측 모드일 경우 참조 화소는 보간되지 않을 수 있다. DC 필터는 현재 블록의 예측 모드가 DC 모드일 경우 필터링을 통해서 예측 블록을 생성할 수 있다.
복원된 블록 또는 픽쳐는 필터부(240)로 제공될 수 있다. 필터부(240)는 디블록킹 필터, 오프셋 보정부, ALF를 포함할 수 있다.
영상 부호화 장치로부터 해당 블록 또는 픽쳐에 디블록킹 필터를 적용하였는지 여부에 대한 정보 및 디블록킹 필터를 적용하였을 경우, 강한 필터를 적용하였는지 또는 약한 필터를 적용하였는지에 대한 정보를 제공받을 수 있다. 영상 복호화 장치의 디블록킹 필터에서는 영상 부호화 장치에서 제공된 디블록킹 필터 관련 정보를 제공받고 영상 복호화 장치에서 해당 블록에 대한 디블록킹 필터링을 수행할 수 있다.
오프셋 보정부는 부호화시 영상에 적용된 오프셋 보정의 종류 및 오프셋 값 정보 등을 기초로 복원된 영상에 오프셋 보정을 수행할 수 있다.
ALF는 부호화 장치로부터 제공된 ALF 적용 여부 정보, ALF 계수 정보 등을 기초로 부호화 단위에 적용될 수 있다. 이러한 ALF 정보는 특정한 파라메터 셋에 포함되어 제공될 수 있다.
메모리(245)는 복원된 픽쳐 또는 블록을 저장하여 참조 픽쳐 또는 참조 블록으로 사용할 수 있도록 할 수 있고 또한 복원된 픽쳐를 출력부로 제공할 수 있다.
전술한 바와 같이 이하, 본 개시의 실시예에서는 설명의 편의상 코딩 유닛(Coding Unit)을 부호화 단위라는 용어로 사용하지만, 부호화뿐만 아니라 복호화를 수행하는 단위가 될 수도 있다.
또한, 현재 블록은, 부호화/복호화 대상 블록을 나타내는 것으로, 부호화/복호화 단계에 따라, 코딩 트리 블록(또는 코딩 트리 유닛), 부호화 블록(또는 부호화 유닛), 변환 블록(또는 변환 유닛) 또는 예측 블록(또는 예측 유닛) 등을 나타내는 것일 수 있다. 본 명세서에서, '유닛'은 특정 부호화/복호화 프로세스를 수행하기 위한 기본 단위를 나타내고, '블록'은 소정 크기의 화소 어레이를 나타낼 수 있다. 별도의 구분이 없는 한, '블록'과 '유닛'은 동등한 의미로 사용될 수 있다. 예컨대, 후술되는 실시예에서, 부호화 블록(코딩 블록) 및 부호화 유닛(코딩 유닛)은 상호 동등한 의미인 것으로 이해될 수 있다.
영상은 블록 단위로 부호화/복호화될 수 있다. 코딩 블록은 트리 구조에 기반하여 재귀적으로 분할될 수 있다. 일 예로, 쿼드(quad) 트리 분할, 바이너리(binary) 트리 분할 또는 터너리(ternary) 트리 분할 중 적어도 하나에 의해 코딩 블록이 분할될 수 있다.
또한, 코딩 블록을 복수의 예측 블록들 또는 복수의 변환 블록들로 분할할 수 있다.
픽처 간 중복 데이터는 인터 예측을 통해 제거된다. 인터 예측에 의해 예측 블록을 생성하고자 하는 경우, 움직임 정보를 이용할 수 있다. 여기서, 움직임 정보는, 움직임 벡터, 참조 영상 인덱스 또는 예측 방향 중 적어도 하나를 포함할 수 있다.
현재 픽처의 참조 영상으로부터, 현재 블록과 가장 유사한 블록을 탐색하고, 탐색된 블록을 현재 블록의 예측 블록으로 설정할 수 있다. 이후, 현재 블록과 예측 블록을 차분하여 잔차 블록을 생성할 수 있다.
도 3 및 도 4는 인터 예측이 수행되는 예를 나타낸 것이다.
도 3은, T-1번째 참조 영상 내, 현재 블록과 동일한 위치한 콜로케이티드 블록으로부터, 움직임 벡터 (x, y)만큼 이격된 예측 블록을 탐색한 결과를 나타낸다.
도시된 예에서, 움직임 벡터 (x,y), 참조 영상 T-1을 가리키는 참조 영상 인덱스, 및 L0방향 예측이 수행되었음을 나타내는 예측 방향 정보가 현재 블록의 움직임 정보로 결정될 수 있다.
도 4는 양방향 예측이 수행되는 예시를 나타낸다.
도 4에 도시된 예에서, L0 예측을 통해, 참조 영상 T-1 내 콜로케이티드 블록으로부터, 움직임 벡터 (x0, y0)만큼 이격된 참조 블록 0이 특정된다. 또한, L1 예측을 통해, 참조 영상 T+1 내 콜로케이티드 블록으로부터, 움직임 벡터 (x1, y1)만큼 이격된 참조 블록 1이 특정된다.
이에 따라, 움직임 벡터 (x0, y0), 참조 영상 T-1을 가리키는 참조 영상 인덱스, 및 L0방향 예측이 수행되었음을 나타내는 예측 방향 정보가 현재 블록의 L0 움직임 정보로 결정될 수 있다. 또한, 움직임 벡터 (x1, y1), 참조 영상 T+1을 가리키는 참조 영상 인덱스, 및 L1 방향 예측이 수행되었음을 나타내는 예측 방향 정보가 현재 블록의 L1 움직임 정보로 결정될 수 있다.
양방향 예측이 수행된 경우, 현재 블록의 최종 예측 블록은 L0 예측 블록 및 L1 예측 블록 간의 가중치 합에 기초하여 생성될 수 있다.
도시된 예에서는, L0 방향이 현재 픽처의 이전 방향이고, L1 방향이 현재 픽처의 이후 방향인 것으로 예시되었다. 도시된 예와 달리, L0 방향 및 L1 방향이 모두 현재 픽처의 이전 방향으로 설정되거나 모두 현재 픽처의 이후 방향으로 설정될 수 있다. 또는, L0 방향이 현재 픽처의 이후 방향이고, L1 방향이 현재 픽처의 이전 방향으로 설정될 수도 있다.
이에 따라, L0 참조 픽처 리스트 및 L1 참조 픽처 리스트 각각에는, 이전 방향의 참조 픽처와 이후 방향 참조 픽처가 혼재되어 있을 수 있다.
인터 예측이 적용되는 경우, 현재 블록의 움직임 정보를 부호화하여 복호화기로 전송해야 한다. 이때, 움직임 벡터를 그대로 부호화하여 복호화기로 전송할 수 있다. 또는, 움직임 벡터 예측 과정을 통해 생성된 예측 움직임 벡터를 이용하여 움직임 벡터의 양을 줄여 복호화기로 전송할 수 있다. 이하, 이에 대해 상세히 살펴보기로 한다.
도 5는 본 발명이 적용되는 일실시예로서, 인터 예측 방법을 도시한 순서도이다.
도 5를 참조하면, 현재 블록의 움직임 정보를 결정할 수 있다(S510). 현재 블록의 움직임 정보는, 현재 블록에 관한 움직임 벡터, 현재 블록의 참조 픽처 인덱스 또는 현재 블록의 인터 예측 방향 중 적어도 하나를 포함할 수 있다.
현재 블록의 움직임 정보는 비트스트림을 통해 시그널링되는 정보 또는 현재 블록의 공간적/시간적 이웃 블록의 움직임 정보 중 적어도 하나를 기초로 획득될 수 있다.
도 6은, 현재 블록의 움직임 정보 유도시 참조되는 공간적 이웃 블록의 위치를 설명하기 위한 도면이다.
도 6에서, LB는 현재 블록 내에서 좌하단 샘플의 위치를 나타내고, RT는 현재 블록 내에서 우상단 샘플의 위치를 나타낸다. AX(X는 0부터 4)는 현재 블록 좌측의 복원 샘플을 나타내고, BY(Y는 0부터 5)는 현재 블록 상단의 복원 샘플을 나타낸다.
도시된 복수의 복원 샘플들 중 적어도 하나의 위치를, 공간적 이웃 블록을 결정하는데 이용될 수 있다. 일 예로, A0, A1, B0, B1 또는 B5 중 적어도 하나가 공간적 이웃 블록을 결정하기 위한 기준 위치로 정의될 수 있다.
도 7은 현재 블록의 움직임 정보를 유도하기 위해 참조되는 시간적 이웃 블록을 설명하기 위한 예시이다.
도 7에 도시된 예에서, CX(X는 0부터 35)는 콜로케이티드 블록 내 샘플들 및 상기 콜로케이티드 블록 주변의 복원 화소들을 나타낸다.
도시된 복수의 복원 샘플들 중 적어도 하나의 위치를, 공간적 이웃 블록을 결정하는데 이용될 수 있다. 일 예로, C21 또는 C35 중 적어도 하나가 공간적 이웃 블록을 결정하기 위한 기준 위치로 정의될 수 있다.
도 8는 현재 블록에 머지 모드가 적용되는 경우, 현재 블록의 움직임 정보를 유도하는 과정을 나타낸 도면이다.
현재 블록에 머지 모드가 적용되는 경우, 현재 블록의 공간적 이웃 블록으로부터 공간적 머지 후보를 유도할 수 있다(S810). 공간적 이웃 블록은, 현재 블록의 상단, 좌측 또는 현재 블록의 코너(예컨대, 상단 좌측 코너, 우측 상단 코너 또는 좌측 하단 코너 중 적어도 하나)에 인접한 블록 중 적어도 하나를 포함할 수 있다.
공간적 머지 후보의 움직임 정보는, 공간적 이웃 블록의 움직임 정보와 동일하게 설정될 수 있다.
현재 블록의 시간적 이웃 블록으로부터 시간적 머지 후보를 유도할 수 있다(S820). 시간적 이웃 블록은, 콜로케이티드 픽처에 포함된 동일 위치 블록(co-located block, 콜로케이티드 블록)을 의미할 수 있다. 콜로케이티드 픽처는, 현재 블록을 포함하는 현재 픽처와 상이한 시간적 순서(Picture Order Count, POC)를 갖는다. 콜로케이트 픽처는 참조 픽처 리스트 내 기 정의된 인덱스를 갖는 픽처로 결정되거나, 비트스트림으로부터 시그널링되는 인덱스에 의해 결정될 수 있다. 시간적 이웃 블록은, 콜로케이티드 픽처 내 현재 블록과 동일한 위치 및 크기를 갖는 블록 내 임의의 블록 또는 현재 블록과 동일한 위치 및 크기를 갖는 블록에 인접한 블록으로 결정될 수 있다. 일 예로, 콜로케이티드 픽처 내 현재 블록과 동일한 위치 및 크기를 갖는 블록의 중심 좌표를 포함하는 블록, 또는, 상기 블록의 우측 하단 경계에 인접한 블록 중 적어도 하나가 시간적 이웃 블록으로 결정될 수 있다.
시간적 머지 후보의 움직임 정보는, 시간적 이웃 블록의 움직임 정보에 기초하여 결정될 수 있다. 일 예로, 시간적 머지 후보의 움직임 벡터는, 시간적 이웃 블록의 움직임 벡터에 기초하여 결정될 수 있다. 또한, 시간적 머지 후보의 인터 예측 방향은 시간적 이웃 블록의 인터 예측 방향과 동일하게 설정될 수 있다. 단, 시간적 머지 후보의 참조 픽처 인덱스는, 고정된 값을 가질 수 있다. 일 예로, 시간적 머지 후보의 참조 픽처 인덱스는 '0'으로 설정될 수 있다.
이후, 공간적 머지 후보 및 시간적 머지 후보를 포함하는 머지 후보 리스트를 생성할 수 있다(S830). 만약, 머지 후보 리스트에 포함된 머지 후보의 개수가 최대 머지 후보 개수보다 작은 경우, 둘 이상의 머지 후보를 조합한 조합된 머지 후보 또는 (0,0) 움직임 벡터(zero motion vector)를 갖는 머지 후보가 머지 후보 리스트에 포함될 수 있다.
머지 후보 리스트에 포함된 머지 후보들의 개수가 최대 개수에 다다르지 못한 경우, 글로벌 움직임 정보 버퍼에 포함된 움직임 정보를, 머지 후보로서 머지 후보 리스트에 추가할 수 있다.
글로벌 움직임 정보 버퍼는, 현재 블록이 포함된 픽처, 슬라이스, 코딩 트리 유닛 라인, 또는 코딩 트리 유닛 내 현재 블록보다 이전에 부호화/복호화된 블록의 움직임 정보를 저장할 수 있다. 일 예로, 소정 영역 내 첫번째 블록에서부터 현재 블록 이전 블록까지, 인터 예측으로 부호화된 블록들의 움직임 정보가 모두 글로벌 움직임 정보 버퍼에 저장되어 있을 수 있다.
또는, 움직임 정보 버퍼에 저장 가능한 움직임 정보의 개수를 M개로 제한할 수 있다. 이 경우, 현재 블록과 가까울수록 광역 움직임 정보 버퍼 상에서 높은 우선순위를 갖고, 현재 블록과 멀수록, 광역 움직임 정보 버퍼 상에서 낮은 우선 순위를 가질 수 있다. 여기서, 현재 블록과의 거리는, 현재 블록과의 부호화/복호화 순서 차분에 의해 결정될 수 있다.
즉, 부호화/복호화 순서에 따라, 후순위로 부호화된 블록에 높은 우선 순위가 부여될 수 있다. 특정 블록의 움직임 정보를 광역 움직임 정보 버퍼에 추가하고자 할 때, 광역 움직임 정보 버퍼에 이미 최대 개수 M만큼의 움직임 정보가 저장되어 있을 경우, 우선 순위가 가장 낮은 블록(즉, 부호화/복호화 순서가 가장 빠른 블록)으로부터 유도된 움직임 정보를 삭제한 뒤, 특정 블록의 움직임 정보를 추가할 수 있다.
다른 예로, 병렬 처리 구조가 적용되는 경우, 병렬 처리되는 영역마다 광역 움직임 정보 버퍼가 생성될 수 있다.
또는, 병렬 처리되는 영역에 포함된 블록의 움직임 정보는 광역 움직임 정보 버퍼에 추가하지 않을 수 있다.
또는, 코딩 트리 유닛 행마다 광역 움직임 정보 버퍼를 구성할 수 있다. 즉, 광역 움직임 정보 버퍼가 코딩 트리 유닛마다 초기화될 수 있다.
병렬 처리가 수행되는 영역마다 개별적으로 광역 움직임 정보 버퍼가 정의되어 사용되는 경우, 병렬 처리가 수행되는 영역 내 초반부에는 광역 움직임 정보 버퍼에 저장된 움직임 정보의 개수가 매우 적다. 병렬 처리가 수행되는 영역 내 초반부에 해당하는 블록들의 부호화/복호화 효율을 높이기 위해, 광역 움직임 정보 버퍼에 기 설정된 초기 움직임 정보를 삽입할 수도 있다. 일 예로, 기 설정된 초기 움직임 정보는, 픽처 전체에 적용되는 글로벌 움직임 정보일 수 있다. 글로벌 움직임 정보는 상위 헤더를 통해 부호화 될 수 있다.
또는, 타일, 브릭 또는 슬라이스 단위로, 글로벌 움직임 정보를 정의할 수도 있다.
머지 후보 리스트가 생성되면, 머지 후보 인덱스에 기초하여, 머지 후보 리스트에 포함된 머지 후보 중 적어도 하나를 특정할 수 있다(S840).
현재 블록의 움직임 정보는, 머지 후보 인덱스에 의해 특정된 머지 후보의 움직임 정보와 동일하게 설정될 수 있다(S850). 일 예로, 머지 후보 인덱스에 의해, 공간적 머지 후보가 선택된 경우, 현재 블록의 움직임 정보는, 공간적 이웃 블록의 움직임 정보와 동일하게 설정될 수 있다. 또는, 머지 후보 인덱스에 의해, 시간적 머지 후보가 선택된 경우, 현재 블록의 움직임 정보는, 시간적 이웃 블록의 움직임 정보와 동일하게 설정될 수 있다.
도 9는 현재 블록에 AMVP 모드가 적용되는 경우, 현재 블록의 움직임 정보를 유도하는 과정을 나타낸 도면이다.
현재 블록에 AMVP 모드가 적용되는 경우, 비트스트림으로부터, 현재 블록의 인터 예측 방향 또는 참조 픽처 인덱스 중 적어도 하나를 복호화할 수 있다(S910). 즉, AMVP 모드가 적용되는 경우, 현재 블록의 인터 예측 방향 또는 참조 픽처 인덱스 중 적어도 하나는 비트스트림을 통해 부호화된 정보에 기초하여 결정될 수 있다.
현재 블록의 공간적 이웃 블록의 움직임 벡터를 기초로, 공간적 움직임 벡터 후보를 결정할 수 있다(S920). 공간적 움직임 벡터 후보는, 현재 블록의 상단 이웃 블록으로부터 유도된 제1 공간적 움직임 벡터 후보 및 현재 블록의 좌측 이웃 블록으로부터 유도된 제2 공간적 움직임 벡터 후보 중 적어도 하나를 포함할 수 있다. 여기서, 상단 이웃 블록은, 현재 블록의 상단 또는 상단 우측 코너에 인접한 블록 중 적어도 하나를 포함하고, 현재 블록의 좌측 이웃 블록은, 현재 블록의 좌측 또는 좌측 하단 코너에 인접한 블록 중 적어도 하나를 포함할 수 있다. 현재 블록의 좌측 상단 코너에 인접한 블록은, 상단 이웃 블록으로 취급될 수도 있고, 또는 좌측 이웃 블록으로 취급될 수도 있다.
현재 블록과 공간적 이웃 블록 사이의 참조 픽처가 상이한 경우, 공간적 움직임 벡터는 공간적 이웃 블록의 움직임 벡터를 스케일링함으로써 획득될 수도 있다.
현재 블록의 시간적 이웃 블록의 움직임 벡터를 기초로, 시간적 움직임 벡터 후보를 결정할 수 있다(S930). 현재 블록과 시간적 이웃 블록 사이의 참조 픽처가 상이한 경우, 시간적 움직임 벡터는 시간적 이웃 블록의 움직임 벡터를 스케일링함으로써 획득될 수도 있다.
공간적 움직임 벡터 후보 및 시간적 움직임 벡터 후보를 포함하는 움직임 벡터 후보 리스트를 생성할 수 있다(S940).
움직임 벡터 후보 리스트가 생성되면, 움직임 벡터 후보 리스트 중 적어도 하나를 특정하는 정보에 기초하여, 움직임 벡터 후보 리스트에 포함된 움직임 벡터 후보 중 적어도 하나를 특정할 수 있다(S950).
상기 정보에 의해 특정된 움직임 벡터 후보를, 현재 블록의 움직임 벡터 예측값으로 설정하고, 움직임 벡터 예측값에 움직임 벡터 차분값을 합하여, 현재 블록의 움직임 벡터를 획득할 수 있다(S960). 이때, 움직임 벡터 차분값은, 비트스트림을 통해 파싱될 수 있다.
현재 블록의 움직임 정보가 획득되면, 획득된 움직임 정보에 기초하여, 현재 블록에 대한 움직임 보상을 수행할 수 있다(S520). 구체적으로, 현재 블록의 인터 예측 방향, 참조 픽처 인덱스 및 움직임 벡터에 기초하여, 현재 블록에 대한 움직임 보상이 수행될 수 있다.
후술되는 실시예들에서는, 머지 모드 하에서의 머지 후보 리스트 및 AMVP 모드 하에서의 움직임 벡터 후보 리스트를 움직임 정보 리스트라는 용어로 호칭하기로 한다. 즉, 움직임 정보 리스트는, 현재 블록에 적용되는 예측 모드에 따라, 머지 후보 리스트 또는 움직임 벡터 후보 리스트를 의미할 수 있다.
현재 블록을 복수의 파티션들(예컨대, 예측 유닛들)로 분할할 수 있다. 이때, 분할된 파티션별로, 움직임 정보를 유도할 수 있다. 이에 따라, 각 파티션에는 상이한 움직임 정보가 적용될 수 있다. 현재 블록을 두개의 파티션으로 분할하는 것을, 대각 분할 또는 기하(Geometry) 분할이라 호칭할 수 있다.
도 10은 현재 블록을 두개의 삼각 형태 파티션들로 분할하는 예를 나타낸 것이다.
현재 블록을 크기가 균등한 2개의 삼각 형태 파티션들로 분할할 수 있다. 도 10의 (a)는, 현재 블록의 센터를 기준으로 45도 각도의 분할선에 기초한 분할 예시이다. 도 10의 (b)는, 센터를 기준으로 135도 각도의 분할선에 기초한 분할 예시이다.
현재 블록의 분할 방향을 나타내는 정보가 부호화되어 시그날링될 수 있다. 일 예로, 상기 정보는, 현재 블록을 분할하는 분할 선의 각도가 45도 인지 또는 135도인지 여부를 나타낼 수 있다.
각 파티션의 움직임 정보를 이용하여 복수의 예측 블록을 생성할 수 있다. 일 예로, 첫번째 파티션의 움직임 정보에 기초하여, 제1 예측 블록을 생성하고, 두번째 파티션의 움직임 정보에 기초하여, 제2 예측 블록을 생성할 수 있다. 현재 블록이 복수의 예측 유닛들로 분할되는 경우, 각 예측 유닛이 양방향 예측을 사용하지 않고, 단방향 예측만 사용하도록 설정할 수 있다.
또한, 파티션들 각각이 상이한 예측 방향을 갖도록 설정할 수 있다. 일 예로, 첫번쨰 파티션에서 L0 방향의 예측을 수행한 경우, 두번째 파티션에서는 L1 방향의 예측을 수행하도록 설정할 수 있다.
설명의 편의를 위해, 후술되는 실시예들에서는, 각 파티션마다 단방향 예측이 수행되는 것으로 가정한다.
각 파티션의 움직임 정보를 부호화 하여 복호화기로 전송할 수 있다.
또는, 움직임 정보를 부호화하는데 필요한 비트량을 줄이기 위해, 현재 블록 주변의 기복원된 블록의 움직임 정보를 참조하여, 각 파티션의 움직임 정보를 결정할 수 있다. 일 예로, 머지 후보 리스트를 이용하여, 각 파티션의 움직임 정보를 유도하는데 이용되는 후보를 특정하는 인덱스를 부호화하여 시그날링할 수 있다.
이때, 제1 파티션의 머지 후보 리스트와 제2 파티션의 머지 후보 리스트를 개별적으로 생성할 수도 있다. 제1 파티션에 대한 머지 후보 리스트는 L0 움직임 정보만으로 구성된 머지 후보들을 포함하도록 구성하고, 제2 파티션에 대한 머지 후보 리스트는 L1 움직임 정보만을 포함하는 머지 후보들로 구성할 수 있다. 제1 파티션의 머지 후보 리스트에 포함된 머지 후보들 중 하나를 특정하는 인덱스 및 제2 파티션의 머지 후보 리스트에 포함된 머지 후보들 중 하나를 특정하는 인덱스가 각각 부호화되어 시그날링될 수 있다.
다른 예로, 제1 파티션 및 제2 파티션에 대해 통합된 머지 후보 리스트를 구성할 수도 있다. 이 경우, 파티션별로 머지 후보를 특정하는 인덱스가 시그날링될 수 있다. 이때, 각 파티션에 대한 인덱스를 시그날링하는 경우, 각 파티션이 상이한 머지 후보를 참조하도록 설정할 수 있다. 일 예로, 제1 파티션에 대한 제1 인덱스에 의해 특정되는 머지 후보가, 제2 파티션의 움직임 정보를 유도하는데 이용되지 않도록 설정할 수 있다.
통합된 머지 후보 리스트는 공간적/시간적 이웃 블록으로부터 유도된 머지 후보를 포함할 수 있다.
일 예로, 도 6 및 도 7에 도시된 예에서, 머지 후보들의 탐색 순서가 A1, B1, B0, A0, B5, C35, C21이고, 머지 후보 리스트가 포함할 수 있는 최대 머지 후보의 개수가 N개인 것으로 가정한다.
머지 후보 리스트 구성시, 상기 열거된 순서대로 공간적/시간적 이웃 블록들을 탐색하여, L0 단방향 예측이 수행된 공간적/시간적 이웃 블록의 움직임 정보를 우선적으로 머지 후보 리스트에 삽입할 수 있다. 그럼에도, 머지 후보 리스트에 포함된 머지 후보들의 개수가 최대 개수 N보다 작은 경우, 다시 상기 열거된 순서대로 공간적/시간적 이웃 블록들을 탐색하여, 양방향 예측이 수행된 공간적/시간적 이웃 블록의 움직임 정보 중 L0 움직임 정보만을 취하여 머지 후보 리스트에 삽입할 수 있다. 그럼에도 불구하고, 머지 후보 리스트에 포함된 머지 후보들의 개수가 최대 개수 N보다 작은 경우, 다시 상기 열거된 순서대로 공간적/시간적 이웃 블록들을 탐색하여, 양방향 예측이 수행된 공간적/시간적 이웃 블록의 움직임 정보 중 L1 움직임 정보만을 취하여 머지 후보 리스트에 삽입할 수 있다. 그럼에도 불구하고, 머지 후보 리스트에 포함된 머지 후보들의 개수가 최대 개수 N보다 작은 경우, 다시 상기 열거된 순서대로 공간적/시간적 이웃 블록들을 탐색하여, 양방향 예측이 수행된 공간적/시간적 이웃 블록의 L0 움직임 정보의 움직임 벡터와 L1 움직임 정보의 움직임 벡터를 가중치 합하여 유도된 결과를 머지 후보 리스트에 삽입할 수 있다. 이때, L0 참조 픽처 및 L1 참조 픽처가 상이한 경우, 현재 픽처 및 L0의 참조 픽처 사이의 POC에 기초하여 L1 움직임 벡터를 스케일링할 수 있다. 즉, L0 움직임 벡터와 스케일링된 L1 움직임 벡터 사이의 가중치 합 연산을 통해 유도된 움직임 벡터가 머지 후보로 설정될 수 있다. 상기 머지 후보의 예측 방향은 L0으로 설정될 수 있고, 참조 픽처 인덱스는 L0 움직임 정보에 존재하는 참조 픽처 인덱스로 설정할 수 있다.
그럼에도 불구하고, 머지 후보 리스트에 포함된 머지 후보들의 개수가 최대 개수 N보다 작은 경우, 광역(Global) 움직임 정보 버퍼(예컨대, HMVP 테이블)에 저장된 움직임 정보 또는 제로 벡터를 머지 후보 리스트에 추가할 수 있다. 광역 움직임 정보 버퍼는, 현재 블록 이전에 부호화/복호화된 블록의 움직임 정보를 저장한다.
간소화된 방법으로 머지 후보 리스트를 구성할 수도 있다. 간소화된 방법이 적용되는 경우, 머지 후보 리스트가 포함할 수 있는 머지 후보들의 최대 개수를 결정한 뒤, 공간적 머지 후보들의 최대 개수와 시간적 머지 후보들의 최대 개수를 설정할 수 있다. 이때, 공간적 머지 후보들의 최대 개수 및 시간적 머지 후보들의 최대 개수의 합은, 머지 후보 리스트가 포함할 수 있는 머지 후보들의 최대 개수를 초과할 수 없다.
일 예로, 머지 후보 리스트가 포함할 수 있는 머지 후보들의 최대 개수가 6개인 것으로 가정한다. 또한, 도 6 및 도 7에 도시된 예에서, 공간적 이웃 블록들 및 시간적 이웃 블록들의 탐색 순서가 A1, B1, B0, A0, B5, C35, C21이라 가정한다. 이 경우, A1, B1, B0, 및 A0의 순서대로, 각 위치를 포함하는 공간적 이웃 블록들을 참조하여 공간적 머지 후보를 유도할 수 있다. 위 4개의 공간적 이웃 블록들 중 적어도 하나가 머지 후보로서 이용 가능하지 않은 경우, B5 위치를 포함하는 공간적 이웃 블록을 참조하여 머지 후보 리스트를 유도할 수 있다. 이후, C35 위치를 포함하는 시간적 이웃 블록으로부터 유도된 머지 후보를 머지 후보 리스트에 삽입할 수 있다. 만약, 위 위치의 시간적 이웃 블록이 머지 후보로서 이용 불가능한 경우, C21 위치를 포함하는 시간적 이웃 블록으로부터 유도된 머지 후보를 머지 후보 리스트에 삽입할 수 있다.
공간적/시간적 머지 후보들이 머지 후보 리스트에 추가되었음에도, 머지 후보 리스트가 포함하는 머지 후보들의 개수가 최대 개수보다 작은 경우, 글로벌 움직임 정보 버퍼를 참조하여 움직임 정보를 리스트에 채워 넣을 수 있다.
위 과정을 거쳐 생성된 머지 후보 리스트가 포함하는 머지 후보들의 개수가 아직 최대 개수에 다다르지 못한 경우, 머지 후보 리스트 내 기 삽입된 머지 후보들의 움직임 정보를 조합하여 조합된 머지 후보를 생성할 수 있다.
일 예로, 머지 후보 리스트에 존재하는 머지 후보들 중 두 머지 후보들 간의 움직임 벡터들을 평균하여 조합된 머지 후보를 유도할 수 있다. 이때, 조합된 머지 후보들을 생성하기 위한 머지 후보들의 조합 [i,j]가 기 정의되어 있을 수 있다. 여기서, i는, 조합된 머지 후보를 생성하는데 이용되는 첫번째 머지 후보의 인덱스를 가리키고, j는 조합된 머지 후보를 생성하는데 이용되는 두번째 머지 후보의 인덱스를 가리킨다. 일 예로, 조합의 우선 순위는, [0,1], [0,2], [1,2], [1,3], [0,3], [1,3], [2,3]과 같이 정의될 수 있다.
[0,1] 조합이 적용되는 경우, 머지 후보 리스트 내 인덱스가 0인 머지 후보의 움직임 벡터와, 인덱스가 1인 머지 후보의 움직임 벡터를 평균하여 유도되는 움직임 벡터가 조합된 머지 후보의 것으로 결정될 수 있다. 조합된 머지 후보 생성시, 예측 방향별로 머지 후보가 생성될 수 있다. 예컨대, [0,1] 조합에 대해, 두 머지 후보들의 L0 움직임 벡터를 평균하여 제1 조합된 머지 후보가 생성되고, 두 머지 후보들의 L1 움직임 벡터를 평균하여, 제2 조합된 머지 후보가 생성될 수 있다.
이때, 조합되는 머지 후보들 중 적어도 하나가 특정 방향에 대한 움직임 정보를 갖지 않는 경우(즉, L0 또는 L1 방향에 대한 움직임 정보가 0인 경우)가 발생할 수 있다. 후보를 조합할 때일 예로, 인덱스가 0인 머지 후보는 양방향 예측이 수행된 반면, 인덱스가 1인 머지 후보는 L1방향 예측이 수행된 것일 수 있다. 이 경우, L0 방향에 대한 조합된 머지 후보 생성시, 두 머지 후보들의 L0 움직임 정보를 평균하는 대신 인덱스가 0인 머지 후보의 L0 정보를 조합된 머지 후보의 움직임 정보로 설정할 수 있다.
또는, 조합하고자 하는 머지 후보가 단방향 예측되었는지 또는 양방향 예측되었는지 여부에 기초하여, 기 설정된 조합의 이용 여부가 결정될 수 있다.
조합된 머지 후보를 생성하기 위한 조합의 개수를 조절할 수도 있다. 일 예로, 조합의 개수를 나타내는 정보가 상위 헤더를 통해 부호화되어 복호화기로 전송될 수 있다. 또는, 부호화기 및 복호화기에서 고정된 개수의 조합 후보들이 정의되어 있을 수 있다. 일 예로, 부호화기 및 복호화기에서 1개의 조합만이 이용 가능하도록 정의되어 있을 수 있다. 이 경우, [0,1] 조합만을 이용하여, 조합된 머지 후보를 생성할 수 있다.
또는, 상술한 방식에 의해 머지 후보 리스트를 생성한 뒤, 기 생성된 머지 후보 리스트에 기초하여, 추가 머지 후보 리스트를 신규 생성할 수 있다. 신규 생성된 추가 머지 후보 리스트에 기초하여, 두 파티션의 움직임 정보가 유도될 수 있다.
이때, 추가 생성되는 머지 후보 리스트는, 기존 머지 후보 리스트에 포함된 머지 후보들의 단방향 움직임 정보를 추출하여 생성될 수 있다. 이때, 기존 머지 후보 리스트에 포함된 머지 후보들을 순차적으로 스캐닝하여, L0 움직임 정보 및 L1 우직임 정보를 번갈아가며 추출할 수 있다.
일 예로, 기존 머지 후보 리스트에 존재하는 첫번째 머지 후보의 L0 움직임 정보를 추가 머지 후보 리스트의 첫번째 머지 후보로 설정하고, 두기존 머지 후보 리스트에 존재하는 두번째 머지 후보의 L1 움직임 정보를 추가 머지 후보 리스트의 두번째 머지 후보로 설정할 수 있다.
위와 같은 방식으로, 2N-1 번째 머지 후보로부터 L0 움직임 정보를 추출하고, 2N 번쨰 머지 후보로부터 L1 움직임 정보를 추출할 수 있다. 추가 머지 후보 리스트가 포함할 수 있는 머지 후보들의 최대 개수는 기존 머지 후보 리스트가 포함할 수 있는 머지 후보들의 최대 개수와 같거나 작을 수 있다. 기존 머지 후보 리스트와 추가 머지 후보 리스트 사이의 최대 개수 차분을 나타내는 정보가 비트스트림을 통해 시그날링될 수도 있다.
리스트 생성 후, 각 파티션의 움직임 정보를 특정하는 인덱스를 부호화여 시그날링할 수 있다.
단, 두 파티션의 움직임 정보가 동일하다면, 현재 블록을 굳이 두 파티션들로 분할할 필요가 없다. 이에 따라, 두 파티션 간의 움직임 정보는 상이하게 설정되어야 한다.
두 파티션들이 통합된 머지 후보 리스트를 이용하는 경우, 첫번째 파티션에 이용된 머지 후보는 두번째 파티션에 대해 이용 불가능한 것으로 설정될 수 있다. 이에 따라, 첫번째 파티션에서 사용가능한 머지 후보가 N개인 경우, 두번째 파티션에서 사용가능한 머지 후보는 N-1개일 수 있다. 이에 따라, 첫번째 파티션에 대한 인덱스 값의 범위는 0 부터 N-1 사이의 값으로 결정되고, 두번째 파티션에 대한 인덱스 값의 범위는 0 부터 N-2 사이의 값으로 결정될 수 있다.
개수 N을 결정하기 위한 정보가 부호화되어 복호화기로 전송될 수 있다. 상기 정보는, 시퀀스, 픽처 헤더 또는 슬라이스 헤더와 같은 상위 헤더를 통해 전송될 수 있다.
또는, 부호화기 및 복호화기에서 N값이 고정되어 있을 수 있다. 후술되는 실시예들에서는, 개수 N이 5라고 가정한다.
각 파티션별로 인덱스를 시그날링하는 방식 이외에도, 복수의 머지 후보들을 조합한 후보들 각각에 인덱스를 할당한 뒤, 두 파티션에 적용되는 머지 후보들의 조합과 일치하는 후보의 인덱스를 부호화하여 시그날링할 수도 있다.
또는, 머지 후보들 및 현재 블록의 분할 방향을 조합한 후보들 각각에 인덱스를 할당한 뒤, 두 파티션들에 적용되는 머지 후보들 및 현재 블록의 분할 방향의 조합과 일치하는 후보의 인덱스를 부호화하여 시그날링할 수도 있다. 일 예로, 머지 후보 리스트가 최대 5개의 머지 후보들을 포함하는 경우, 두 파티션들에 적용되는 머지 후보들 및 현재 블록의 분할 방향의 조합은 40개이다. (즉, 첫번쨰 서브 블록에 적용 가능한 머지 후보들의 개수 5 x 두번째 서브 블록에 적용 가능한 머지 후보들의 개수 4 x 현재 블록에 적용 가능한 분할 방향 후보의 개수 2) 이에 따라, 현재 블록에 대해 0부터 39 사이의 값들 중 하나를 나타내는 인덱스가 부호화되어 시그날링될 수 있다.
현재 블록의 분할 양상은 도시된 예에 한정되지 않는다. 현재 블록을 분할하는 분할선의 각도 α는 0도 내지 180도 사이 양자화된 값을 가질 수 있다. 일 예로, 각도 α가 0도 내지 180도 범위를 4등분한 것 중 하나인 경우, 각도 α의 값은 0도, 45도, 90도 또는 135도 중 하나가 될 수 있다. 또는, 0도 내지 180 사이의 범위를 이보다 더 많은 개수로 분할할 수도 있다.
도 11은 각도 후보가 8개인 경우를 예시한 것이다.
여기서, 8개의 각도 후보는 0도에서 180 범위를 8등분하여 유도되는 것일 수 있다.
코딩 블록의 분할 구조와 중첩되는 분할 방법은 분할 후보에서 제외할 수 있다. 일 예로, 분할선의 각도 α가 90도 또는 180도인 것은, 코딩 블록에 바이너리 트리(Binary Tree) 분할을 적용한 것과 동일한 분할 결과를 초래한다. 이에 따라, 각도 α가 90도 및 180도 인 것은, 분할 후보에서 제외할 수 있다.
도시된 예에서는, 현재 블록이 균등한 크기의 파티션들로 분할되는 것으로 예시되었으나, 파티션들을 비균등한 크기로 분할할 수도 있다.
도 12는 비균등 분할이 적용되는 예를 나타낸다.
β는 현재 블록의 센터로부터 분할선 까지의 거리를 나타낸다.
복수개의 분할 후보들 중 하나를 특정하는 정보가 부호화되어 시그날링될 수 있다. 일 예로, 분할 각도 및 분할 거리를 조합한 후보들 중 하나를 특정하는 인덱스가 부호화되어 시그날링될 수 있다. 부호화기 및 복호화기에는, 분할 각도 및 분할 거리를 조합한 후보들 각각에 할당되는 인덱스를 정의한 룩업 테이블이 기 저장되어 있을 수 있다.
도 13은 분할 각도 후보 및 분할 거리 후보들을 예시한 것이다.
표 1은 도 13의 (a)에 도시된 분할 각도 후보 및 분할 거리 후보들의 조합들과 이들 각각에 할당되는 인덱스를 정의한 룩업 테이블에 대한 예시이다.
merge_gpm_partition_idx 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
angleIdx 0 0 2 2 2 2 3 3 3 3 4 4 4 4 5 5
distanceIdx 1 3 0 1 2 3 0 1 2 3 0 1 2 3 0 1
merge_gpm_partition_idx 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
angleIdx 5 5 8 8 11 11 11 11 12 12 12 12 13 13 13 13
distanceIdx 2 3 1 3 0 1 2 3 0 1 2 3 0 1 2 3
merge_gpm_partition_idx 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
angleIdx 14 14 14 14 16 16 18 18 18 19 19 19 20 20 20 21
distanceIdx 0 1 2 3 1 3 1 2 3 1 2 3 1 2 3 1
merge_gpm_partition_idx 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
angleIdx 21 21 24 24 27 27 27 28 28 28 29 29 29 30 30 30
distanceIdx 2 3 1 3 1 2 3 1 2 3 1 2 3 1 2 3
표 1에서, merge_gpm_partition_idx는 각도 후보 및 거리 후보의 조합에 할당되는 인덱스를 나타낸다. angleIdx는 각도 후보에 할당되는 인덱스이고, distanceIdx는 거리 후보에 할당되는 인덱스이다.
분할 후보들 중 하나를 특정하는 인덱스 merge_gpm_partition_idx가 부호화되어 시그날링될 수 있다. 상기 인덱스 merge_gpm_partition_idx에 의해, 각도 후보 및 거리 후보가 결정되고, 결정된 각도 후보 및 거리 후보에 의해 현재 블록의 분할 형태가 결정될 수 있다.
도 13의 (b)는, angleIdx가 4인 경우 distanceIdx 값에 따른 분할 양상을 예시한 것이다. distanceIdx는 현재 블록의 중심으로부터 분할선 중심 까지의 거리 또는 특정 경계에서 분할선의 위치 중 적어도 하나를 가리킨다. 일 예로, distanceIdx의 값은, 현재 블록의 중심으로부터 분할선 중심 까지의 거리가 n/4 지점임을 나타내거나, 분할선이 특정 경계상의 n/4 지점을 지남을 나타낸다.
일 예로, distnaceIdx가 m인 것은, 분할선이 상단 경계의 mW/4 또는 (4-m)W/4를 지나거나, 좌측 경계의 mH/4 또는 (4-m)H/4를 지남을 나타낸다. 여기서, W는 현재 블록의 너비를 나타내고, H는 현재 블록의 높이를 나타낸다.
앞서 설명한 바와 같이, 바이너리 트리 분할과 중복되는 분할 형태를 배제하기 위해, 0도 또는 180도를 나타내는 각도 후보와, 값이 0인 거리 후보의 조합은, 이용 가능한 분할 후보에서 제외될 수 있다.
각도 후보의 개수 또는 거리 후보의 개수 중 적어도 하나를 특정하는 정보가 부호화되어 시그날링될 수 있다. 일 예로, 분할선의 각도 α의 양자화 변수(예컨대, 0 내지 180 사이의 범위의 구획 개수)가 상위 헤더를 통해 부호화되어 시그날링될 수 있다.
또는, 분할 거리의 개수가 상위 헤더를 통해 부호화되어 시그날링될 수 있다.
다른 예로, 현재 블록에 인접하는 이웃 블록의 분할 양상을 참조하여, 각도 후보 및/또는 거리 후보의 개수 및/또는 종류가 결정될 수 있다.
조합 후보 구성시, 특정 각도에 높은 인덱스 우선 순위를 부여할 수 잇다. 일 예로, 45도 또는 135도 각도를 포함하는 조합 후보에 높은 우선 순위를 부여할 수 있다. 여기서, 우선 순위가 높을수록, 해당 조합에 할당되는 인덱스의 값이 작아질 수 있다. 특정 조합에 높은 우선 순위를 부여하는 경우, 해당 조합을 특정하는 인덱스의 비트 길이가 감소할 수 있다.
현재 블록을 두개의 파티션들로 분할하는 것이 허용되는지 여부는, 현재 블록의 너비 및 높이 중 적어도 하나를 임계값과 비교하여 결정될 수 있다. 구체적으로, 상기 분할을 적용할 경우, 파티션의 너비 또는 높이 중 적어도 하나가 임계값 이하가 되는 경우, 현재 블록을 분할하는 것이 허용되지 않을 수 있다. 일 예로, 크로마 성분에 대한 파티션의 크기를 고려하여, 현재 블록을 두개의 파티션으로 분할하는 것이 허용되는지 여부가 결정될 수 있다.
일 예로, 색차 포맷이 4:2:0인 경우, 8x8 크기의 루마 블록은 4x4 크기의 크로마 블록에 대응한다. 이때, distanceIdx가 3인 대각 분할이 적용되는 경우, 크로마 블록을 분할함으로써 생성된 파티션들 중 하나의 너비 또는 높이가 1이 될 수 있다. 부호화/복호화 효율을 고려할 때, 너비 또는 높이가 1인 파티션을 생성하는 것은 바람직하지 않은 바, 위 경우, 상기 분할 형태가 현재 블록에 적용되지 않도록 설정될 수 있다.
임계값은, 1, 2, 4 또는 8과 같이 2의 지수승으로 표현되는 자연수일 수 있다. 임계값을 결정하기 위한 정보가 부호화되어 시그날링될 수 있다. 이때, 현재 영상이 4:4:4인 경우, 상기 정보의 부호화가 생략될 수 있다.
또는, 부호화기 및 복호화기에 임계값이 기 저장되어 있을 수 있다. 또는, 크로마 블록 분할 형태(예컨대, 싱글 트리 또는 듀얼 트리) 및/또는 색차 포맷 중 적어도 하나에 기초하여 임계값이 유도될 수 있다.
현재 블록의 크기 또는 형태 중 적어도 하나에 기초하여, 이용 가능한 분할 후보의 개수를 조절할 수도 있다. 일 예로, 현재 블록의 크기가 8x8 이하인 경우, 거리 인덱스가 distanceIdx 1 내지 distanceIdx3인 분할 후보는 이용 불가능한 것으로 설정될 수 있다. 이 경우, 거리 인덱스가 distanceIdx 1 내지 distanceIdx3dls 분할 후보가 제외된 룩업 테이블이 이용될 수 있다.
반면, 현재 블록의 크기가 8x8보다 큰 경우, 거리 인덱스가 distanceIdx 0 내지 distanceIdx 3인 분할 후보들이 이용 가능할 수 있다.
현재 블록의 크기 또는 형태 중 적어도 하나에 기초하여, 이용 가능한 각도 인덱스의 종류 또는 개수가 조절될 수도 있다.
크로마 블록의 분할 형태 또는 색차 포맷 중 적어도 하나에 기초하여, 현재 블록에 적용 가능한 분할 형태가 결정될 수도 있다.
루마 블록의 및 크로마 블록의 분할 형태를 독립적으로 결정할 수도 있다. 일 예로, 루마 블록의 분할 형태를 나타내는 인덱스 및 크로마 블록의 인덱스를 나타내는 인덱스가 독립적으로 부호화되어 시그날링될 수 있다.
루마 블록 및 크로마 블록의 분할 형태가 독립적으로 결정되는지 여부를 나타내는 정보가 비트스트림을 통해 시그날링될 수 있다. 일 예로, 상기 정보는 1비트의 플래그일 수 있다.
또는, 크로마 블록의 크기, 형태 또는 색차 포맷 중 적어도 하나에 기초하여, 루마 블록 및 크로마 블록의 분할 형태가 독립적으로 결정되는지 여부가 결정될 수 있다.
현재 블록의 예측 샘플은, 제1 파티션의 움직임 정보를 기초로 획득되는 제1 예측 샘플과, 제2 파티션의 움직임 정보를 기초로 획득되는 제2 예측 샘플을 기초로 유도될 수 있다. 구체적으로, 현재 블록의 예측 샘플은, 제1 예측 샘플 및 제2 예측 샘플의 가중합 연산에 기초하여 획득될 수 있다.
현재 블록 내 파티션들의 경계가 결정되면, 결정된 경계 및 현재 블록의 크기에 기초하여, 각 예측 샘플에 대한 가중치가 결정될 수 있다. 여기서, 가중치는, 제1 예측 샘플에 적용되는 제1 가중치 및 제2 예측 샘플에 적용되는 제2 가중치를 포함한다. 가중치는 0을 포함하는 정수일 수 있다.
가중치를 결정하기 위한 변수 nW, nH, shift1, offset1, hwRatio, displacementX, displacementY, partFlip, shiftHor는 다음의 표 2와 같이 유도될 수 있다.
(1) nW = ( cIdx == 0 ) ? nCbW : nCbW * SubWidthC(2) nH = ( cIdx == 0 ) ? nCbH : nCbH * SubHeightC(3) shift1 = Max( 5, 17 - BitDepth )(4) offset1 = 1 << ( shift1 - 1 )(5) hwRatio = nH / nW(6) displacementX = angleIdx(7) displacementY = ( angleIdx + 8 ) % 32(8) partFlip = ( angleIdx >= 13 && angleIdx <= 27 ) ? 0 : 1(9) shiftHor = ( angleIdx % 16 == 8 || ( angleIdx % 16 != 0 && hwRatio > 0 )) ? 0 : 1
표 2에서 변수 nCbW 및 nCbH는 각각 현재 블록의 너비와 높이를 나타낸다. 변수 cIdx는, 컬러 성분을 나타내고, 예컨대, 0인 경우는 루마 성분, 0보다 큰 값은 색차 성분을 나타낼 수 있다. angleIdx는 각도 후보를 특정하는 인덱스를 나타낸다. 변수 SubWidthC 및 SubHeightC는 색차 포맷에 따라 결정되는 값을 나타낸다. 일 예로, 색차 포맷이 4:2:0인 경우, SubWidthC와 SubHeightC는 모두 2로 설정될 수 있다. 색차 포맷이 4:4:4라면, SubWidthC와 SUbHeightC는 모두 1로 설정될 수 있다. 색차 포맷이 4:2:2라면, SubWidthC는 2로 설정되고, SubHeightC는 1로 설정될 수 있다.
표 2를 통해 유도된 변수에 기초하여 변수 offsetX와 offsetY를 결정할 수 있다. 이때, 만약 변수 shiftHor의 값이 0이라면, offsetX와 offsetY의 값은 다음 표 3의 (10) 및 (11) 항목에 기초하여 결정될 수 있다. 반면, 변수 shiftHor의 값이 1이라면, offsetX와 offsetY의 값은 표 3의 (12)및 (13) 항목에 기초하여 결정될 수 있다.
(10) offsetX = ( -nW )  >>  1(11) offsetY = ( ( -nH )  >>  1 ) + (angleIdx < 16 ? ( distanceIdx * nH )  >>  3 : -( ( distanceIdx * nH )  >>  3 ) )(12) offsetX = ( ( -nW )  >>  1 ) + ( angleIdx < 16 ? ( distanceIdx * nW )  >>  3 : -( ( distanceIdx * nW )  >>  3 ) )(13) offsetY = ( - nH )  >>  1
이후, 화소 위치 (x, y)에 기초하여, 표 4와 같이 변수 xL과 yL을 유도할 수 있다.
(14) xL = ( cIdx = = 0 ) ? x : x * SubWidthC(15) yL = ( cIdx = = 0 ) ? y : y * SubHeightC
이후, 기설정된 룩업 테이블과 변수 xL, offsetX, yL 및 offsetY를 이용하여 임시 가중치 값인 weightIdx를 유도할 수 있다. 표 5는 변수 weightIdx를 유도하는 예를 나타낸 것이다.
(16) weightIdx = ( ( ( xL + offsetX )  <<  1 ) + 1 ) * disLut[ displacementX ] + ( ( ( yL + offsetY )  <<  1 ) + 1 ) ) * disLut[ displacementY ]
표 5에서, disLut는 룩업 테이블을 나타내고, displacementX는 룩업 테이블 내 하나를 특정하는 인덱스를 나타낸다. 변수 displacementX는 표 2에 예시된 바와 같이, angleIdx에 의해 유도될 수 있다.
표 6은 룩업 테이블 disLut를 예시한 것이다.
Idx 0 2 3 4 5 6 8 10 11 12 13 14
disLut[idx] 8 8 8 4 4 2 0 -2 -4 -4 -8 -8
Idx 16 18 19 20 21 22 24 26 27 28 29 30
disLut[idx] -8 -8 -8 -4 -4 -2 0 2 4 4 8 8
이후, 유도된 변수들을 이용하여, 샘플 위치 (x, y)에 적용되는 가중치 wValue가 결정될 수 있다. 표 7은 가중치 wValue가 유도되는 예를 나타낸다.
(17) weightIdxL = partFlip  ?  32 + weightIdx  :  32 - weightIdx(18) wValue = Clip3( 0, 8, ( weightIdxL + 4 )  >>  3 )
제1 예측 샘플에 적용되는 제1 가중치는 변수 wValue와 동일하게 설정되고, 제2 예측 샘플에 적용되는 제2 가중치는 소정의 상수값, 예컨대, 8에서 변수 wValue를 차분하여 유도될 수 있다. 또는, 이와 반대로, 변수 wValue를 제2 가중치로 설정하고, 소정의 상수값, 예컨대 8에서 변수 wVlaue를 차분하여 제1 가중치를 유도할 수도 있다.
수학식 1은 현재 블록의 예측 샘플을 유도하는 예를 나타낸 것이다.
Figure PCTKR2020012866-appb-M000001
수학식 1에서, P1 및 P2는 각각 제1 예측 샘플 및 제2 예측 샘플을 나타낸다. 그리고, α1 및 α2는 각각 제1 가중치 및 제2 가중치를 나타낸다.
도 14 내지 도 16은 현재 블록이 135도 각도를 갖는 분할선을 이용하여 분할된 경우, 각 예측 샘플을 유도하기 위해 적용되는 필터 계수를 나타낸다.
도 14 내지 도 16의 (a)는 각 예측 샘플에 적용되는 제1 가중치를 나타낸 것이고, 도 14 내지 도 16의 (b)는 각 예측 샘플에 적용되는 제2 가중치를 나타낸 것이다.
도 14에 도시된 예에서와 같이, 예측 대상 샘플의 위치에 따라, 제1 가중치 및 제2 가중치 중 하나가 0으로 설정되거나, 제1 가중치 및 제2 가중치가 모두 0이 아닌 값을 가질 수 있다. 일 예로, 현재 블록 내 (3, 1) 위치의 예측 샘플은, 제1 예측 샘플에 6의 가중치를 할당하고, 제2 예측 샘플에 2의 가중치를 할당하여 유도될 수 있다.
제1 가중치 및 제2 가중치가 모두 0이 아닌 값을 갖는 예측 샘플들을 포함하는 영역을 필터 적용 영역으로 정의할 수 있다. 도 14에 도시된 것보다 필터 적용 영역의 크기를 더 작게 설정할 수 있다. 도 15는 이에 대한 예시를 나타낸다.
또는, 현재 블록 내 필터를 적용함이 없이 예측 샘플을 유도할 수도 있다. 도 16은 이에 대한 예시이다.
현재 블록 내 필터 적용 영역을 특정하는 정보가 부호화되어 시그날링될 수 있다. 일 예로, 현재 블록이 45도 또는 135도의 분할선에 의해 분할되는 경우, 도 14 내지 도 16에 도시된 예시들 중 현재 블록에 적용되는 것을 특정하는 인덱스가 부호화되어 시그날링될 수 있다.
또는, 필터 적용 영역에 포함된 대각 라인의 개수를 나타내는 정보가 부호화되어 시그날링될 수 있다. 이때, 상기 정보는, x축 및 y축 좌표가 동일한 대각 라인 이외의 대각 라인들의 개수를 2로 나눈 값으로 부호화될 수도 있다.
현재 블록에 대한 필터 적용 영역을 특정하는 정보는, 컬러 성분별로 부호화되어 시그날링될 수도 있다. 일 예로, 루마 성분 및 크로마 성분 각각에 대해 필터 적용 영역을 특정하기 위한 인덱스가 부호화될 수 있다.
또는, 컬러 성분별로 필터 적용 영역이 기 정의되어 있을 수 있다. 일 예로, 루마 성분에 대해서는, 도 14에 예시된 필터 적용 영역이 이용되고, 크로마 성분에 대해서는, 도 16의 예시에서와 같이, 필터 적용 영역을 설정하지 않을 수 있다.
또는, 제1 파티션 및 제2 파티션의 움직임 정보에 기초하여, 필터 적용 영역의 크기/형태가 결정될 수 있다. 일 예로, 제1 파티션의 참조 픽처와 현재 픽처 사이의 POC 차분 또는 제2 파티션의 참조 픽처 사이의 POC 차분 중 적어도 하나에 기초하여 필터 적용 영역의 크기/형태가 결정될 수 있다. 또는, 제1 파티션의 움직임 벡터와 제2 파티션의 움직임 벡터 사이의 차분이 문턱값 이상인지 여부에 기초하여, 필터 적용 영역의 크기/형태가 결정될 수 있다.
또는, 분할 각도 α 또는 분할 거리 β 중 적어도 하나에 기초하여, 필터 적용 영역의 크기/형태가 결정될 수 있다.
도 17은 angleIdx 및 distanceIdx에 따른 가중치 적용 양상을 나타낸다.
도 17은, angleIdx가 5이고, distanceIdx가 0인 경우, 예측 샘플들 각각에 대한 가중치 wValue의 값을 나타낸다.
현재 블록이 인터 예측으로 부호화된 경우, 기 정의된 크기의 블록 단위로 움직임 정보를 저장할 수 있다. 여기서, 기 정의된 크기는 4x4, 8x8, 2x8 또는 8x2 중 어느 하나일 수 있다. 또는, 움직임 정보가 저장되는 단위의 크기를 나타내는 정보가 부호화되어 시그날링될 수 있다.
후술되는 실시예에서는, 움직임 정보가 저장되는 단위가 4x4 크기를 갖는 것으로 가정한다. 아울러, 움직임 정보가 저장되는 4x4 크기 영역을 서브 블록이라 호칭하기로 한다.
현재 블록이 복수개의 예측 유닛들로 분할된 경우, 분할 경계에 속하지 않는 서브 블록(예컨대, 분할선이 지나지 않는 서브 블록)은, 속하는 파티션의 움직임 정보가 해당 서브 블록의 움직임 정보로서 저장될 수 있다. 반면, 서브 블록이 분할 경계에 속하는 경우(에컨대, 분할선 상에 위치하는 서브 블록), 제1 파티션의 움직임 정보 또는 제2 파티션의 움직임 정보 중 적어도 하나가 서브 블록의 움직임 정보로서 저장될 수 있다.
도 18은 움직임 정보의 저장 양상을 설명하기 위한 도면이다.
제1 파티션에 속하는 서브 블록들에는 제1 파티션의 움직임 정보가 저장되고, 제2 파티션에 속하는 서브 블록들에는 제2 파티션의 움직임 정보가 저장될 수 있다.
제1 파티션 및 제2 파티션을 분할하는 분할선상에 위치하는 서브 블록들에 저장되는 움직임 정보는, 제1 파티션 및 제2 파티션의 움직임 정보에 기초하여 결정될 수 있다.
일 예로, 제1 파티션 및 제2 파티션 간의 예측 방향이 상이한 경우, 제1 파티션의 움직임 정보 및 제2 파티션의 움직임 정보가 조합된 양방향 움직임 정보가 서브 블록들에 저장될 수 있다.
반면, 제1 파티션 및 제2 파티션의 예측 방향이 동일한 경우, 제1 파티션의 움직임 정보 또는 제2 파티션의 움직임 정보 중 어느 하나가 서브 블록에 저장될 수 있다.
일 예로, 제1 파티션의 움직임 정보 및 제2 파티션의 움직임 정보 중 제1 파티션의 움직임 정보가 서브 블록에 저장되도록 고정적으로 설정되거나, 제2 파티션의 움직임 정보가 서브 블록에 저장되도록 고정적으로 설정될 수 있다. 또는, 분할선 상에 위치하는 서브 블록에 제1 파티션의 움직임 정보를 저장할 것인지 또는 제2 파티션의 움직임 정보를 저장할 것인지를 나타내는 정보가 부호화되어 시그날링될 수 있다.
또는, 분할 각도 또는 분할 거리에 기초하여, 서브 블록에 제1 파티션의 움직임 정보를 저장할 것인지 또는 제2 파티션의 움직임 정보를 저장할 것인지 여부가 결정될 수 있다. 일 예로, 분할 거리 β가 0이 아닌 경우, 제1 파티션 및 제2 파티션의 크기는 상이할 수 있다. 이때, 제1 파티션 및 제2 파티션 중 크기가 큰 파티션의 움직임 정보를 분할선 상에 위치하는 서브 블록에 저장할 수 있다. 반대로, 제1 파티션 및 제2 파티션 중 크기가 작은 파티션의 움직임 정보를 분할선 상에 위치하는 서브 블록에 저장할 수도 있다.
또는, 제1 파티션의 참조 픽처 인덱스와 현재 픽처 사이의 POC 차분 및 제2 파티션의 참조 픽처 인덱스와 현재 픽처 사이의 POC 차분을 비교하여, POC 차분값이 더 작은 파티션의 움직임 정보를, 분할선 상에 위치하는 서브 블록에 저장할 수 있다. 또는, 제1 파티션 및 제2 파티션 중 참조 픽처 인덱스 값이 작은 파티션의 움직임 정보를, 분할선 상에 위치하는 서브 블록에 저장할 수 있다.
상술한 분할선 상에 위치하는 서브 블록들의 움직임 정보 저장 방법을, 필터 적용 영역에 포함되는 서브 블록들에 확장 적용할 수도 있다.
도 19는 이에 대한 예시를 나타낸다.
도 19에 도시된 예에서, 필터 적용 영역에 포함된 서브 블록들에는, 제1 파티션의 움직임 정보, 제2 파티션의 움직임 정보 또는 이들을 조합한 움직임 정보 중 적어도 하나가 저장될 수 있다.
다른 예로, 서브 블록 내 대표 가중치를 이용하여, 서브 블록이 분할 경계에 위치하는지 여부가 결정될 수 있다.
서브 블록 내 특정 위치의 샘플에 적용되는 가중치를 대표 가중치로 설정할 수 있다. 여기서, 특정 위치는, 중심 위치, 좌측 상단 위치, 우측 상단 위치, 좌측 하단 위치 또는 우측 하단 위치 중 적어도 하나일 수 있다.
또는, 서브 블록 내 가중치의 최대값, 최소값 또는 평균값을 이용하여, 해당 서브 블록이 분할 경계에 위치하는지 여부가 결정될 수 있다.
도 20은 대표 가중치를 설명하기 위한 도면이다.
도시된 예에서와 같이, 4x4 크기의 서브 블록 내 중앙 위치의 샘플에 적용되는 가중치를 대표 가중치로 설정할 수 있다. 일 예로, 서브 블록 내 (2, 2) 위치의 샘플에 적용되는 가중치가 대표 가중치로 설정될 수 있다.
표 3에 예시된 (10) 내지 (13) 항목에 기초하여, 각 파티션의 변수 offsetX와 offsetY를 계산한 뒤, 서브 블록 내 중앙 위치에 대해 임시 가중치 값을 나타내는 변수 weightIdx를 유도할 수도 있다. 수학식 2는 임시 가중치 인덱스 weightIdx가 유도되는 예를 나타낸 것이다.
Figure PCTKR2020012866-appb-M000002
수학식 2에서, (xSbIdx, ySbIdx)는 현재 블록 내 서브 블록의 인덱스를 나타낸다. 구체적으로, 변수 xSbIdx는 서브 블록의 수평 방향 인덱스를 나타내고, 변수 ySbIdx는 서브 블록의 수직 방향 인덱스를 나타낸다. 일 예로, 현재 블록이 8x8 크기를 갖는 경우, 변수 xSbIdx 및 변수 ySbIdx는 각각 0 또는 1의 값으로 설정될 수 있다.
수학식 2에 기초하여 유도된 임시 가중치값 weightIdx이 서브 블록의 임시 가중치 인덱스가 상기 서브 블록의 대표 가중치를 가리킬 수 있다.
이후, 상기 임시 가중치 인덱스를 이용하여, 서브 블록의 움직임 정보 저장 타입을 결정할 수 있다. 일 예로, 다음 수학식 3에 기초하여, 서브 블록의 움직임 정보 저장 타입을 나타내는 변수 sType이 유도될 수 있다.
Figure PCTKR2020012866-appb-M000003
또는, 다음의 수학식 4를 이용하여, 변수 sType을 유도할 수 있다.
Figure PCTKR2020012866-appb-M000004
수학식 4에서, N은 32(즉, 실수 위치로 스케일 다운하면 0.5를 표현)일 수 있다.
변수 sType의 유도 방법은 현재 블록의 크기 또는 형태 중 적어도 하나에 기초하여 결정될 수 있다. 일 예로, 현재 블록의 크기가 문턱값 이상인 경우, 상기 수학식 3에 의해 변수 sType이 유도되고, 현재 블록의 크기가 문턱값보다 작은 경우, 상기 수학식 4에 의해 변수 sType이 유도될 수 있다.
변수 sType은 움직임 정보 저장 타입을 나타낸다. 변수 sType의 값은 0 내지 2의 값으로 설정될 수 있다. 일 예로, 변수 sType의 값이 0 또는 1인 것은, 두 파티션들 중 하나의 움직임 정보가 저장 영역에 저장됨을 의미한다. 변수 sType의 값이 2인 것은, 두 파티션의 움직임 정보를 조합한 양방향 움직임 정보가 서브 블록에 저장됨을 의미한다.
도 21은 각 서브 블록에 대한 변수 sType의 값을 나타낸다.
각 서브 블록 내 중심 좌표에 대한 임시 가중치 인덱스를 이용하여, 변수 sType이 유도될 수 있다.
도시된 예에서, 분할선 상에 위치하는 좌측 상단 서브 블록 및 우측 하단 서브 블록의 저장 타입 sType이 2의 값는 것으로 예시되었다. 이에 따라, 좌측 상단 서브 블록 및 우측 상단 서브 블록에는 제1 파티션의 움직임 정보 및 제2 파티션의 움직임 정보가 조합된 양방향 움직임 정보가 저장될 수 있다.
반면, 경계 영역에 포함되지 않는 좌측 하단 서브 블록의 sType은 0으로 유도되었다. 이 경우, 해당 서브 블록에는 해당 서브 블록을 포함하는 파티션(즉, 현재 블록의 하단에 위치하는 파티션)의 움직임 정보가 저장될 수 있다. 또한, 경계 영역에 포함되지 않는 우측 상단 서브 블록의 sType은 1로 유도되었다. 이 경우, 해당 서브 블록에는 해당 서브 블록을 포한하는 파티션(즉, 현재 블록의 상단에 위치하는 파티션)의 움직임 정보가 저장될 수 있다.
또는, 변수 sType이 2인 것을, 제1 파티션 또는 제2 파티션 중 어느 하나의 움직임 정보가 저장되는 것으로 정의할 수도 있다. 일 예로, 변수 sType이 2인 서브 블록에는, 해당 서브 블록을 점유하는 면적이 더 큰 파티션의 움직임 정보가 저장될 수 있다. 또는, 중심 위치(예컨대, (2, 2) 위치)에서의 대표 가중치 및 상기 중심 위치의 이웃 위치(예컨대, 좌측, 우측, 상단 또는 하단 중 적어도 하나)의 가중치를 비교하여, 제1 파티션 또는 제2 파티션 중 하나가 선택될 수도 있다.
또는, 변수 sType이 2인 것을, 제1 파티션 또는 제2 파티션 중 기 정의된 어느 하나의 움직임 정보를 저장하는 것으로 설정할 수도 있다. 이에 따라, 변수 sType이 2인 서브 블록은, 변수 sType이 0인 서브 블록과 동일한 움직임 정보를 저장하거나, 변수 sType이 1인 서브 블록과 동일한 움직임 정보를 저장할 수 있다.
또는, 현재 블록의 크기 또는 형태 중 적어도 하나에 기초하여, 움직임 정보 저장 방법이 적응적으로 결정될 수도 있다. 일 예로, 현재 블록의 크기가 8x8 이하인 경우, 현재 블록 내 전체 서브 블록이, 기 설정된 위치의 서브 블록에 저장되는 움직임 정보를 저장하도록 설정될 수 있다. 여기서, 기 설정된 위치의 서브 블록은, 좌측 상단, 우측 상단, 좌측 하단, 우측 하단 또는 중앙 위치의 샘플을 포함하는 4x4 크기의 서브 블록일 수 있다. 여기서, 기 설정된 위치의 서브 블록에 저장되는 움직임 정보는, 상기 서브 블록에 대한 변수 sType에 기초하여 결정될 수 있다.
또는, 현재 블록의 블록의 크기가 문턱값보다 작은 경우, 현재 블록 내 서브 블록들이 동일한 움직임 정보를 저장하도록 설정하는 한편, 현재 블록의 크기가 문턱값보다 큰 경우, 서브 블록들 각각에 저장되는 움직임 정보가 개별적으로 결정되도록 설정할 수 있다.
현재 블록이 두개의 파티션들로 분할된 경우, 화면 내 에측 및 화면 간 예측이 혼합된 예측 방법을 적용할 수도 있다.
도 22는 혼합된 예측 방법이 적용되는 예를 나타낸다.
현재 블록이 두개의 파티션들로 분할된 경우, 제1 파티션 및 제2 파티션 중 어느 하나에는 화면 내 예측을 적용하고, 다른 하나에는 화면 간 예측을 적용할 수 있다. 즉, 두 파티션 간 상이한 예측 방법을 적용할 수 있다.
화면 내 예측 모드는, 현재 블록 주변 이웃 블록의 화면 내 예측 모드를 참조하여 유도될 수 있다. 일 예로, 이웃 블록의 화면 내 예측 모드를 기초로 MPM 후보들을 유도할 수 있다. 여기서, 이웃 블록은, 상단 이웃 블록 또는 좌측 이웃 블록 중 적어도 하나를 포함할 수 있다. MPM 후보 리스트에 포함된 복수의 MPM 후보들 중 현재 블록의 화면 내 예측 모드와 동일한 MPM 후보의 인덱스를 부호화하여 시그날링할 수 있다.
또는, 조합된 예측 방법이 적용되는 경우, 고정된 화면 내 예측 모드를 사용할 수도 있다. 일 예로, 조합된 예측 방법이 적용되는 경우, 평면 모드를 고정적으로 사용할 수 있다.
또는, 현재 블록에 인접하는 이웃 블록들의 화면 내 예측 모드들 중 사용 빈도가 가장 높은 것을 이용할 수도 있다.
또는, 디폴트 모드들 중 어느 하나의 화면 내 예측 모드들을 사용할 수 있다. 여기서, 디폴트 모드는, DC, 평면, 수평, 수직 또는 대각 방향 모드 중 적어도 하나를 포함할 수 있다. 디폴트 모드의 구성은, 현재 블록의 분할 방향에 따라 상이할 수 있다. 일 예로, 현재 블록을 분할하는 분할선이 135도의 각도를 갖는 경우, 좌상단 대각 방향의 화면 내 예측 모드가 디폴트 모드로 설정될 수 있다. 반면, 현재 블록을 분할하는 분할선이 45도의 각도를 갖는 경우, 좌하단 대각 방향 또는 우상단 대각 방향의 화면 내 예측 모드가 디폴트 모드로 설정될 수 있다.
각 파티션에 적용되는 예측 기법(예컨대, 화면 내 예측 또는 화면 간 예측)은 주변 블록의 부호화 모드를 고려하여 결정될 수 있다. 여기서, 주변 블록은 상단 이웃 블록 또는 좌측 이웃 블록 중 적어도 하나를 포함할 수 있다.
도 23 및 도 24는 이웃 블록의 부호화 모드를 참조하여, 각 파티션에 적용되는 예측 기법을 결정하는 예를 나타낸 것이다.
도 23은 현재 블록을 분할하는 분할선의 각도가 45도인 경우의 예시이다.
도 23에 도시된 예에서, 상단 이웃 블록 및 좌측 이웃 블록은 모두 제1 파티션의 경계에 접한다. 이에 따라, 상단 이웃 블록 및 좌측 이웃 블록의 부호화 모드를 참조하여, 제1 파티션에 적용될 예측 기법을 결정할 수 있다.
구체적으로, 상단 이웃 블록 및 좌측 이웃 블록의 부호화 모드가 동일한 경우, 상기 부호화 모드와 동일한 예측 기법을 제1 파티션에 적용할 수 있다. 일 예로, 도 23의 (a)에 도시된 예에서와 같이, 상단 이웃 블록 및 좌측 이웃 블록이 모두 화면 내 예측으로 부호화된 경우, 제1 파티션에도 화면 내 예측이 적용될 수 있다. 반면, 도 23의 (d)에 도시된 예에서와 같이, 상단 이웃 블록 및 좌측 이웃 블록이 모두 화면 간 예측으로 부호화된 경우, 제1 파티션에도 화면 간 예측이 적용될 수 있다.
도 23의 (b) 및 (c)에 도시된 예에서와 같이, 상단 이웃 블록 및 좌측 이웃 블록의 부호화 모드가 상이한 경우, 화면 내 예측 또는 화면 간 예측 중 높은 우선 순위를 갖는 예측 기법이 제1 파티션에 적용될 수 있다. 또는, 상단 이웃 블록 및 좌측 이웃 블록의 부호화 모드가 상이한 경우, 제1 파티션에 적용되는 예측 기법을 나타내는 정보가 부호화되어 시그날링될 수 있다.
제2 파티션에는 제1 파티션과 상이한 예측 기법이 적용될 수 있다. 일 예로, 제1 파티션에 화면 내 예측이 적용되는 경우, 제2 파티션에는 화면 간 예측이 적용될 수 있다. 반대로, 제1 파티션에 화면 간 예측이 적용되는 경우, 제2 파티션에는 화면 내 예측이 적용될 수 있다.
도 24는 현재 블록을 분할하는 분할선의 각도가 135도인 경우의 예시이다.
도 24에 도시된 예에서, 상단 이웃 블록은 제1 파티션의 경계에 접하고, 좌측 이웃 블록은 제2 파티션의 경계에 접한다.
이에 따라, 제1 파티션에 적용되는 예측 기법은 상단 이웃 블록의 부호화 모드를 기초로 유도되고, 제2 파티션에 적용되는 예측 기법은 좌측 이웃 블록의 부호화 모드를 기초로 유도될 수 있다.
일 예로, 도 24의 (b) 및 (c)에 도시된 예에서와 같이, 상단 이웃 블록의 부호화 모드와 좌측 이웃 블록의 부호화 모드가 상이한 경우, 제1 파티션에는 상단 이웃 블록과 동일한 예측 기법이 적용되고, 제2 파티션에는 좌측 이웃 블록과 동일한 예측 기법이 적용될 수 있다.
도 24의 (a) 및 (d)에 도시된 예에서와 같이, 상단 이웃 블록 및 좌측 이웃 블록의 부호화 모드가 동일한 경우, 제1 파티션 또는 제2 파티션에 적용되는 예측 기법을 나타내는 정보가 부호화되어 시그날링될 수 있다.
또는, 제1 파티션 및 제2 파티션이 상이한 크기를 갖는 경우, 제1 파티션 및 제2 파티션 중 크기가 더 큰 파티션에 화면 내 예측을 적용하고, 더 작은 파티션에 화면 간 예측을 적용할 수 있다.
또는, 제1 파티션에 적용되는 예측 기법 및 제2 파티션에 적용되는 예측 기법의 조합을 특정하는 인덱스가 부호화되어 시그날링될 수도 있다.
혼합된 예측 방법이 적용되는 경우에도, 파티션 간 경계에 필터가 적용될 수 있다. 이때, 각 예측 샘플에 적용되는 가중치를 결정하는 방법은, 상술한 설명과 같다.
또는, 블록 크기에 따라, 필터 영역의 크기를 조절할 수도 있다.
또는, 필터 적용시, 화면 간 예측 및 화면 내 예측 간 우선 순위를 설정하여, 우선 순위가 높은 예측 기법에 의해 유도된 예측 샘플에 보다 높은 가중치가 적용되도록 설정할 수도 있다. 일 예로, 화면 간 예측이 화면 내 예측 보다 더 높은 우선 순위를 가질 경우, 화면 간 예측이 수행된 영역에 화면 내 예측이 수행된 영역보다 더 큰 필터 계수를 할당할 수 있다.
또는, 화면 간 예측과 화면 내 예측이 혼합된 예측 방법이 적용되는 경우에는, 파티션 경계에 필터가 적용되지 않도록 설정할 수도 있다 (예컨대, 도 16의 예시 적용).
블록이 복원되면, 복원된 블록 간 디블록킹 필터가 적용될 수 있다.
이때, 현재 블록에 혼합된 예측 방법이 적용된 경우, 현재 블록에 적용된 부호화 모드가 화면 내 예측인 것으로 간주하여, 경계 강도(Boundary Strength)를 결정할 수 있다. 또는, 현재 블록에 혼합된 에측 방법이 적용된 경우, 현재 블록에 적용된 부호화 모드가 화면 간 예측인 것으로, 간주하여, 경계 강도를 결정할 수도 있다.
또는, 두 파티션들 중, 디블록킹 필터가 적용되는 경계에 접하는 파티션에 적용된 예측 기법에 기초하여, 경계 강도가 결정될 수도 있다.
도 25는 경계 강도를 결정하는 예를 나타낸다.
현재 블록의 상단에 디블록킹 필터가 적용되는 경우, 현재 블록의 상단 경계에 인접하는 파티션에 적용된 부호화 모드에 기초하여, 경계 강도가 결정될 수 있다. 일 예로, 도 25의 (a)에 도시된 예에서, 현재 블록의 상단 경계에 인접하는 파티션에는 화면 내 예측이 적용되었으므로, 현재 블록의 부호화 모드가 화면 내 예측인 것으로 간주하고, 경계 강도를 결정할 수 있다.
현재 블록의 우측에 디블록킹 필터가 적용되는 경우, 현재 블록의 우측 경계에 인접하는 파티션에 적용된 부호화 모드에 기초하여, 경계 강도가 결정될 수 있다. 일 예로, 도 25의 (b)에 도시된 예에서, 현재 블록의 우측 경계에 인접하는 파티션에는 화면 간 예측이 적용되었으므로, 현재 블록의 부호화 모드가 화면 간 예측인 것으로 간주하고, 경계 강도를 결정할 수 있다.
혼합된 예측 방법에 의해 부호화된 블록은, 화면 내 예측 정보와 화면 간 예측 정보를 모두 가질 수 있다. 이에 따라, 혼합된 예측 방법에 의해 부호화된 블록은, 차후 부호화/복호화될 블록의 화면 내 예측 정보 유도시 참조 가능할 뿐만 아니라, 화면 간 예측 정보 유도시 참조 가능할 수 있다.
도 26은 혼합된 예측 방법에 의해 부호화된 블록을 참조하여 화면 내 예측 정보 또는 화면 간 예측 정보가 유도되는 예를 나타낸다.
도 26의 (a)에 도시된 예에서, 현재 블록의 상단에 이웃하는 블록에 혼합된 예측 방법이 적용되었고, 현재 블록에는 화면 간 예측이 적용되는 것으로 가정한다. 이 경우, 현재 블록의 화면 간 예측 정보 유도시, 상단 이웃 블록의 움직임 정보를 참조할 수 있다. 일 예로, 상단 이웃 블록의 움직임 정보를 참조하여 머지 후보를 유도할 수 있다.
도 26의 (b)에 도시된 예에서, 현재 블록의 좌측에 이웃하는 블록에 혼합된 예측 방법이 적용되었고, 현재 블록에는 화면 내 예측이 적용되는 것으로 가정한다. 이 경우, 현재 블록의 화면 내 예측 정보 유도시, 좌측 이웃 블록의 화면 내 예측 정보를 참조할 수 있다. 일 예로, 좌측 이웃 블록의 인트라 예측 모드에 기초하여, 현재 블록의 MPM 후보를 유도할 수 있다.
또는, 이웃 블록에 혼합된 예측 방법이 적용된 경우, 이웃 블록의 인트라 예측 모드가 기 정의된 화면 내 예측 모드인 것으로 간주하고, 현재 블록의 MPM 후보를 유도할 수도 있다. 여기서, 기 정의된 화면 내 예측 모드는 평면 모드일 수 있다.
다른 예로, 혼합된 예측 방법이 적용된 경우, 화면 간 예측 정보만을 저장하고, 화면 내 예측 정보는 저장하지 않도록 설정될 수도 있다. 이와 반대로, 혼합된 예측 방법이 적용된 경우, 화면 내 예측 정보만을 저장하고, 화면 간 예측 정보는 저장하지 않도록 설정될 수도 있다.
원본 블록에서 예측 블록을 차분하여 잔차 블록을 유도하고, 잔차 블록에 변환을 적용하여 변환 계수가 유도될 수 있다. 아울러, 유도된 변환 계수를 양자화한 뒤, 양자화된 변환 계수를 부호화할 수 있다.
이때, 현재 블록이 두개의 파티션들로 분할되는 경우, 현재 블록 내 일부 영역에만 변환을 적용할 수 있다. 여기서, 일부 영역은, 현재 블록을 분할하는 포함하는 영역이고, 2의 지수승으로 표현되는 개수 만큼의 샘플들을 포함할 수 있다.
도 27은 변환이 적용되는 영역을 예시한 것이다.
도 27의 예에서, 변수 W는 현재 블록이 높이를 나타내고, 변수 H는 현재 블록의 너비를 나타낸다. 도 27에서 굵은선으로 표시된 것은, 현재 블록을 분할하는 분할선을 나타낸다.
도 27에 예시된 바와 같이, 분할선 상의 샘플들을 포함하는 축소된 변환 영역을 설정한다. 축소된 변환 영역의 형태는, 분할선의 위치에 따라 결정될 수 있다. 또한, 축소된 변환 영역이 포함하는 너비 또는 높이 중 적어도 하나는, 현재 블록의 너비 또는 높이보다 2의 지수승 만큼 감소된 상태일 수 있다. 일 예로, 도 27의 (a)에 도시된 예에서와 같이, 축소된 변환 영역은 현재 블록의 너비 W 및 높이 H 대비 1/2 만큼 축소된 형태를 가질 수 있다. 또는, 도 27의 (b) 또는 (c)에 도시된 예에서와 같이, 축소된 변환 영역은 현재 블록의 너비 W 대비 1/2 크기의 너비와, 현재 블록의 높이 H와 동일한 높이를 가질 수 있다.
변환 영역의 축소 비율을 나타내는 정보가 부호화되어 시그날링될 수 있다. 여기서, 축소 비율은, 현재 블록 크기 대비 변환 영역 크기를 나타낸다. 일 예로, 복수의 축소 비율 후보 중 하나를 특정하는 인덱스가 부호화됭 시그날링될 수 있다.
상기 축소 비율을 나타내는 정보가 너비 및 높이 각각에 대해 시그날링될 수 있다. 일 예로, 축소된 변환 영역의 너비의 축소 비율을 나타내는 제1 축소 비율 정보 정보와, 축소된 변환 영역의 높이의 축소 비율을 나타내는 제2 축소 비율 정보가 각각 부호화되어 시그날링될 수 있다.
또는, 현재 블록을 분할하는 분할 선의 각도 α와 분할 선의 거리 β 중 적어도 하나에 기초하여, 축소된 변환 영역의 축소 비율이 유도될 수 있다.
또는, 분할선이 통과하는 영역에 포함된 행 또는 열에 기초하여, 축소된 변환 영역이 크기가 결정될 수 있다.
또는, 현재 블록의 크기에 기초하여, 축소된 변환 영역의 축소 비율이 결정될 수 있다. 일 예로, 현재 블록의 크기가 8x8 이하인 경우, 축소된 변환 비율의 너비 및 높이는 각각 현재 블록의 너비 및 높이 대비 최대 1/2 만큼 축소될 수 있다. 반면, 현재 블록의 크기가 8x8 보다 큰 경우(예컨대, 16x16 또는 32x32인 경우), 축소된 변환 비율의 너비 및 높이는 각각 현재 블록의 너비 및 높이 대비 최대 1/4 만큼 축소될 수 있다.
축소된 변환 영역에는 변환이 적용되는 한편, 축소된 변환 영역을 제외한 잔여 영역에는 변환이 적용되지 않을 수 있다. 이 경우, 변환이 적용되지 않는 영역의 잔차값은 0으로 설정될 수 있다.
축소된 변환 영역이 이용되는지 여부를 나타내는 정보가 부호화되어 시그날링될 수 있다. 이때, 현재 블록 내 0이 아닌 변환 계수가 존재하는지 여부를 나타내는 플래그(예컨대, CBF(coded block flag))의 값에 기초하여, 상기 정보의 부호화 여부가 결정될 수 있다. 일 예로, 복호화기는 CBF의 값이 0이 아닌 경우(즉, 현재 블록에 0이 아닌 변환 계수가 존재하는 것으로 결정된 경우), 추가적으로 축소된 변환 영역과 관련된 정보를 파싱할 수 있다.
다른 예로, 현재 블록에 대각 분할이 적용된 경우에는, CBF가 축소된 변환 영역이 이용되었는지 여부를 나타내도록 설정할 수도 있다. 일 예로, 현재 블록에 대각 분할이 적용되었고, 상기 현재 블록에 대한 CBF의 값이 0인 것은, 축소된 변환 영역에만 0이 아닌 변환 계수가 존재함을 나타낼 수 있다.
다른 예로, 현재 블록의 부호화 모드에 따라, 축소된 변환 영역이 사용 가능한지 여부가 결정될 수 있다. 여기서, 부호화 모드는, 화면 내 예측 또는 화면 간 에측을 나타낸다. 일 예로, 현재 블록이 화면 내 예측으로 부호화된 경우, 축소된 변환 영역을 이용한 변환 방식은 이용 불가능한 것으로 설정될 수 있다. 반면, 현재 블록이 화면 간 예측으로 부호화된 경우, 축소된 변환 영역과 관련된 정보가 시그날링될 수 있다.
또는, 현재 블록 내 파티션들의 크기가 상이한 경우 또는 파티션들의 크기가 동일한 경우에만, 축소된 변환 영역이 설정될 수도 있다.
다른 예로, 축소된 변환 영역에 기초한 변환 방법이 이용 가능한지 여부를 나타내는 정보가 상위 헤더를 통해 시그날링될 수 있다. 여기서, 상위 헤더는, 시퀀스, 픽처 헤더 또는 슬라이스 헤더 중 적어도 하나를 포함한다. 상기 정보가, 축소된 변환 영역에 기초한 변환 방법이 이용 가능함을 나타내는 경우에 한하여, 블록 레벨에서 축소된 변환 영역과 관련한 정보가 시그날링될 수 있다. 반면, 상기 정보가, 축소된 변환 영역에 기초한 변환 방법이 이용 불가능함을 나타내는 경우, 블록 레벨에서 축소된 변환 영역과 관련한 정보가 시그날링되지 않을 수 있다.
또는, 현재 블록의 크기 또는 형태 중 적어도 하나에 기초하여, 축소된 변환 영역에 기초한 변환 방법이 이용 가능한지 여부가 결정될 수 있다. 일 예로, 현재 블록이 포함하는 샘플의 개수가 64개 이하인 경우, 축소된 변환 영역에 기초한 변환 방법이 이용 가능한 것으로 설정될 수 있다. 이 경우, 축소된 변환 영역에 관한 정보가 시그날링될 수 있다. 반면, 현재 블록이 포함하는 샘플의 개수가 64개를 초과하는 경우, 축소된 변환 영역에 기초한 변환 방법이 이용 불가능한 것으로 설정될 수 있다. 이 경우, 축소된 변환 영역에 관한 정보도 부호화되지 않을 수 있다.
상술한 예에서는, 분할 경계를 고려하여, 축소된 변환 영역이 설정되는 것으로 예시되었다. 다른 예로, 분할 경계와 관계없이, 축소된 변환 영역을 설정할 수도 있다.
축소된 변환 영역에는, 2차 변환이 수행되지 않을 수 있다.
또는, 축소된 변환 영역에 2차 변환을 선택적으로 수행할 수 있다. 이 경우, 축소된 변환 영역에 2차 변환이 수행되었는지 여부를 나타내는 정보가 부호화되어 시그날링될 수 있다.
상술한 실시예들에서 사용된 신택스들의 명칭은, 설명의 편의를 위해 명명된 것에 불과하다.
복호화 과정 또는 부호화 과정을 중심으로 설명된 실시예들을, 부호화 과정 또는 복호화 과정에 적용하는 것은, 본 개시의 범주에 포함되는 것이다. 소정의 순서로 설명된 실시예들을, 설명된 것과 상이한 순서로 변경하는 것 역시, 본 개시의 범주에 포함되는 것이다.
상술한 실시예는 일련의 단계 또는 순서도를 기초로 설명되고 있으나, 이는 발명의 시계열적 순서를 한정한 것은 아니며, 필요에 따라 동시에 수행되거나 다른 순서로 수행될 수 있다. 또한, 상술한 실시예에서 블록도를 구성하는 구성요소(예를 들어, 유닛, 모듈 등) 각각은 하드웨어 장치 또는 소프트웨어로 구현될 수도 있고, 복수의 구성요소가 결합하여 하나의 하드웨어 장치 또는 소프트웨어로 구현될 수도 있다. 상술한 실시예는 다양한 컴퓨터 구성요소를 통하여 수행될 수 있는 프로그램 명령어의 형태로 구현되어 컴퓨터 판독 가능한 기록 매체에 기록될 수 있다. 상기 컴퓨터 판독 가능한 기록 매체는 프로그램 명령어, 데이터 파일, 데이터 구조 등을 단독으로 또는 조합하여 포함할 수 있다. 컴퓨터 판독 가능한 기록 매체의 예에는, 하드 디스크, 플로피 디스크 및 자기 테이프와 같은 자기 매체, CD-ROM, DVD와 같은 광기록 매체, 플롭티컬 디스크(floptical disk)와 같은 자기-광 매체(magneto-optical media), 및 ROM, RAM, 플래시 메모리 등과 같은 프로그램 명령어를 저장하고 수행하도록 특별히 구성된 하드웨어 장치가 포함된다. 상기 하드웨어 장치는 본 개시에 따른 처리를 수행하기 위해 하나 이상의 소프트웨어 모듈로서 작동하도록 구성될 수 있으며, 그 역도 마찬가지이다.
본 발명은 영상을 부호화/복호화할 수 있는 전자 장치에 적용될 수 있다.

Claims (13)

  1. 현재 블록을 제1 파티션 및 제2 파티션으로 분할되는 것으로 결정된 경우, 상기 현재 블록의 분할 형태를 나타내는 인덱스를 파싱하는 단계;
    상기 제1 파티션에 대한 제1 움직임 정보를 결정하는 단계;
    상기 제2 파티션에 대한 제2 움직임 정보를 결정하는 단계; 및
    상기 제1 움직임 정보를 기초로 유도되는 제1 예측 샘플 및 상기 제2 움직임 정보를 기초로 유도되는 제2 예측 샘플에 기초하여, 상기 현재 블록의 최종 예측 샘플을 유도하는 단계를 포함하되,
    상기 인덱스는, 상기 현재 블록을 분할하는 분할선의 각도 후보들 중 하나를 특정하는 각도 인덱스와 상기 현재 블록 내 상기 분할선의 위치를 나타내는 거리 인덱스가 조합된 후보들 중 하나를 특정하는 것을 특징으로 하는, 영상 복호화 방법.
  2. 제1 항에 있어서,
    상기 제1 움직임 정보는, 머지 후보 리스트에 포함된 N 개의 후보들 중 하나를 특정하는 제1 머지 인덱스에 의해 획득되고,
    상기 제2 움직임 정보는, 상기 머지 후보 리스트 내 상기 제1 머지 인덱스에 대응되는 머지 후보를 제외한 N-1 개의 후보들 중 하나를 특정하는 제2 머지 인덱스에 의해 획득되는 것을 특징으로 하는, 영상 복호화 방법.
  3. 제1 항에 있어서,
    상기 최종 예측 샘플은, 상기 제1 예측 샘플 및 상기 제2 예측 샘플 간의 가중합 연산에 기초하여 획득되고,
    상기 제1 예측 샘플에 적용되는 제1 가중치 및 상기 제2 예측 샘플에 적용되는 제2 가중치는, 상기 각도 인덱스 및 상기 거리 인덱스에 기초하여 결정되는 것을 특징으로 하는, 영상 복호화 방법.
  4. 제1 항에 있어서,
    상기 현재 블록 내 기 정의된 크기의 서브 블록 단위로, 움직임 정보를 저장하는 단계를 더 포함하고,
    서브 블록의 저장 타입은, 상기 서브 블록의 대표 위치 샘플의 가중치를 기초로 결정되는 것을 특징으로 하는, 영상 복호화 방법.
  5. 제4 항에 있어서,
    상기 대표 위치 샘플은, 상기 서브 블록의 중앙에 위치하는 것을 특징으로 하는, 영상 복호화 방법.
  6. 제4 항에 있어서,
    상기 저장 타입은, 상기 서브 블록에 상기 제1 움직임 정보를 할당하는 제1 저장 타입, 상기 서브 블록에 상기 제2 움직임 정보를 할당하는 제2 저장 타입 및 상기 서브 블록에 상기 제1 움직임 정보 및 상기 제2 움직임 정보를 조합하여 유도되는 양방향 정보를 할당하는 제3 저장 타입 중 하나인 것을 특징으로 하는, 영상 복호화 방법.
  7. 현재 블록을 제1 파티션 및 제2 파티션으로 분할하는 단계, 상기 현재 블록의 분할 형태를 나타내는 인덱스가 비트스트림에 부호화됨;
    상기 제1 파티션에 대한 제1 움직임 정보를 결정하는 단계;
    상기 제2 파티션에 대한 제2 움직임 정보를 결정하는 단계; 및
    상기 제1 움직임 정보를 기초로 유도되는 제1 예측 샘플 및 상기 제2 움직임 정보를 기초로 유도되는 제2 예측 샘플에 기초하여, 상기 현재 블록의 최종 예측 샘플을 유도하는 단계를 포함하되,
    상기 인덱스는, 상기 현재 블록을 분할하는 분할선의 각도 후보들 중 하나를 특정하는 각도 인덱스와 상기 현재 블록 내 상기 분할선의 위치를 나타내는 거리 인덱스가 조합된 후보들 중 하나를 특정하는 것을 특징으로 하는, 영상 부호화 방법.
  8. 제7 항에 있어서,
    머지 후보 리스트에 포함된 N 개의 후보들 중 상기 제1 움직임 정보와 동일한 제1 머지 후보를 나타내는 제1 머지 인덱스가 부호화된 이후,
    상기 머지 후보 리스트 내 상기 제1 머지 후보를 제외한 N-1 개의 후보들 중 상기 제2 움직임 정보와 동일한 제2 머지 후보를 나타내는 제2 머지 인덱스가 부호화되는 것을 특징으로 하는, 영상 부호화 방법.
  9. 제7 항에 있어서,
    상기 최종 예측 샘플은, 상기 제1 예측 샘플 및 상기 제2 예측 샘플 간의 가중합 연산에 기초하여 획득되고,
    상기 제1 예측 샘플에 적용되는 제1 가중치 및 상기 제2 예측 샘플에 적용되는 제2 가중치는, 상기 각도 인덱스 및 상기 거리 인덱스에 기초하여 결정되는 것을 특징으로 하는, 영상 부호화 방법.
  10. 제7 항에 있어서,
    상기 현재 블록 내 기 정의된 크기의 서브 블록 단위로, 움직임 정보를 저장하는 단계를 더 포함하고,
    서브 블록의 저장 타입은, 상기 서브 블록의 대표 위치 샘플의 가중치를 기초로 결정되는 것을 특징으로 하는, 영상 부호화 방법.
  11. 제10 항에 있어서,
    상기 대표 위치 샘플은, 상기 서브 블록의 중앙에 위치하는 것을 특징으로 하는, 영상 부호화 방법.
  12. 제10 항에 있어서,
    상기 저장 타입은, 상기 서브 블록에 상기 제1 움직임 정보를 할당하는 제1 저장 타입, 상기 서브 블록에 상기 제2 움직임 정보를 할당하는 제2 저장 타입 및 상기 서브 블록에 상기 제1 움직임 정보 및 상기 제2 움직임 정보를 조합하여 유도되는 양방향 정보를 할당하는 제3 저장 타입 중 하나인 것을 특징으로 하는, 영상 부호화 방법.
  13. 비디오 부호화 방법에 의해 부호화되는 비트스트림을 저장하는 컴퓨터로 판독가능한 기록 매체에 있어서,
    상기 비디오 부호화 방법은,
    현재 블록을 제1 파티션 및 제2 파티션으로 분할하는 단계, 상기 현재 블록의 분할 형태를 나타내는 인덱스가 비트스트림에 부호화됨;
    상기 제1 파티션에 대한 제1 움직임 정보를 결정하는 단계;
    상기 제2 파티션에 대한 제2 움직임 정보를 결정하는 단계; 및
    상기 제1 움직임 정보를 기초로 유도되는 제1 예측 샘플 및 상기 제2 움직임 정보를 기초로 유도되는 제2 예측 샘플에 기초하여, 상기 현재 블록의 최종 예측 샘플을 유도하는 단계를 포함하되,
    상기 인덱스는, 상기 현재 블록을 분할하는 분할선의 각도 후보들 중 하나를 특정하는 각도 인덱스와 상기 현재 블록 내 상기 분할선의 위치를 나타내는 거리 인덱스가 조합된 후보들 중 하나를 특정하는 것을 특징으로 하는, 컴퓨터로 판독가능한 기록 매체.
PCT/KR2020/012866 2019-09-23 2020-09-23 비디오 신호 처리 방법 및 장치 WO2021060830A1 (ko)

Applications Claiming Priority (10)

Application Number Priority Date Filing Date Title
KR10-2019-0116839 2019-09-23
KR20190116841 2019-09-23
KR20190116838 2019-09-23
KR10-2019-0116838 2019-09-23
KR10-2019-0116841 2019-09-23
KR20190116839 2019-09-23
KR20190116840 2019-09-23
KR10-2019-0116840 2019-09-23
KR20200038384 2020-03-30
KR10-2020-0038384 2020-03-30

Publications (1)

Publication Number Publication Date
WO2021060830A1 true WO2021060830A1 (ko) 2021-04-01

Family

ID=75165875

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2020/012866 WO2021060830A1 (ko) 2019-09-23 2020-09-23 비디오 신호 처리 방법 및 장치

Country Status (2)

Country Link
KR (1) KR20210035069A (ko)
WO (1) WO2021060830A1 (ko)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20170023921A (ko) * 2017-02-24 2017-03-06 에스케이텔레콤 주식회사 다양한 형태의 블록 분할을 이용하는 영상 부호화/복호화 장치 및 방법
KR20180025285A (ko) * 2016-08-31 2018-03-08 주식회사 케이티 비디오 신호 처리 방법 및 장치
KR20180087880A (ko) * 2012-10-12 2018-08-02 한국전자통신연구원 영상의 부호화/복호화 방법 및 이를 이용하는 장치
WO2019039322A1 (en) * 2017-08-22 2019-02-28 Panasonic Intellectual Property Corporation Of America IMAGE ENCODER, IMAGE DECODER, IMAGE ENCODING METHOD, AND IMAGE DECODING METHOD
WO2019164031A1 (ko) * 2018-02-22 2019-08-29 엘지전자 주식회사 영상 코딩 시스템에서 블록 분할 구조에 따른 영상 디코딩 방법 및 장치

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20180087880A (ko) * 2012-10-12 2018-08-02 한국전자통신연구원 영상의 부호화/복호화 방법 및 이를 이용하는 장치
KR20180025285A (ko) * 2016-08-31 2018-03-08 주식회사 케이티 비디오 신호 처리 방법 및 장치
KR20170023921A (ko) * 2017-02-24 2017-03-06 에스케이텔레콤 주식회사 다양한 형태의 블록 분할을 이용하는 영상 부호화/복호화 장치 및 방법
WO2019039322A1 (en) * 2017-08-22 2019-02-28 Panasonic Intellectual Property Corporation Of America IMAGE ENCODER, IMAGE DECODER, IMAGE ENCODING METHOD, AND IMAGE DECODING METHOD
WO2019164031A1 (ko) * 2018-02-22 2019-08-29 엘지전자 주식회사 영상 코딩 시스템에서 블록 분할 구조에 따른 영상 디코딩 방법 및 장치

Also Published As

Publication number Publication date
KR20210035069A (ko) 2021-03-31

Similar Documents

Publication Publication Date Title
WO2018026219A1 (ko) 비디오 신호 처리 방법 및 장치
WO2017171370A1 (ko) 비디오 신호 처리 방법 및 장치
WO2017222326A1 (ko) 비디오 신호 처리 방법 및 장치
WO2017204532A1 (ko) 영상 부호화/복호화 방법 및 이를 위한 기록 매체
WO2017176030A1 (ko) 비디오 신호 처리 방법 및 장치
WO2018155986A2 (ko) 비디오 신호 처리 방법 및 장치
WO2018026118A1 (ko) 영상 부호화/복호화 방법
WO2017018664A1 (ko) 인트라 예측 모드 기반 영상 처리 방법 및 이를 위한 장치
WO2018030599A1 (ko) 인트라 예측 모드 기반 영상 처리 방법 및 이를 위한 장치
WO2020111785A1 (ko) 영상 신호 부호화/복호화 방법 및 이를 위한 장치
WO2019225993A1 (ko) 비디오 신호 처리 방법 및 장치
WO2018097626A1 (ko) 비디오 신호 처리 방법 및 장치
WO2018047995A1 (ko) 인트라 예측 모드 기반 영상 처리 방법 및 이를 위한 장치
WO2012005520A2 (en) Method and apparatus for encoding video by using block merging, and method and apparatus for decoding video by using block merging
WO2019050292A1 (ko) 비디오 신호 처리 방법 및 장치
WO2019190201A1 (ko) 비디오 신호 처리 방법 및 장치
WO2018212579A1 (ko) 비디오 신호 처리 방법 및 장치
WO2018124333A1 (ko) 인트라 예측 모드 기반 영상 처리 방법 및 이를 위한 장치
WO2019182295A1 (ko) 비디오 신호 처리 방법 및 장치
WO2020096428A1 (ko) 영상 신호 부호화/복호화 방법 및 이를 위한 장치
WO2018066958A1 (ko) 비디오 신호 처리 방법 및 장치
WO2019050291A1 (ko) 비디오 신호 처리 방법 및 장치
WO2018056701A1 (ko) 비디오 신호 처리 방법 및 장치
WO2019190199A1 (ko) 비디오 신호 처리 방법 및 장치
WO2020096427A1 (ko) 영상 신호 부호화/복호화 방법 및 이를 위한 장치

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20870136

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20870136

Country of ref document: EP

Kind code of ref document: A1