KR20210035069A - Method and apparatus for processing a video - Google Patents

Abstract

A video decoding method according to the present disclosure may comprise the steps of: when it is determined that a current block is divided into a first partition and a second partition, parsing an index indicating the partition type of the current block; determining first motion information for the first partition; determining second motion information for the second partition; and deriving a final prediction sample of the current block on the basis of a first prediction sample derived on the basis of the first motion information and a second prediction sample derived on the basis of the second motion information. At this time, the index may specify one candidate among candidates in which an angle index specifying one angle candidate among angle candidates of a dividing line dividing the current block and a distance index indicating the position of the dividing line in the current block are combined. Accordingly, encoding/decoding efficiency can be improved.

Description

Video signal processing method and apparatus {METHOD AND APPARATUS FOR PROCESSING A VIDEO}

The present disclosure relates to a video signal processing method and apparatus.

Recently, demand for 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.

Meanwhile, as the demand for high-resolution images increases, the demand for 3D image contents as a new image service is also increasing. Discussion is underway on a video compression technology for effectively providing 3D image content of high resolution and ultra high resolution.

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 technical problems to be achieved in the present disclosure are not limited to the technical problems mentioned above, and other technical problems that are not mentioned will be clearly understood by those of ordinary skill in the technical field to which the present disclosure belongs from the following description. I will be able to.

In the video signal decoding method according to the present disclosure, when it is determined that the current block is divided into a first partition and a second partition, parsing an index indicating a partition type of the current block, a first motion for the first partition Determining information, determining second motion information for the second partition, and a first prediction sample derived based on the first motion information and a second prediction derived based on the second motion information Based on the sample, it may include the step of deriving a final prediction sample of the current block. In this case, 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 according to the present disclosure 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. In this case, 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.

In the video signal decoding method according to the present disclosure, the first motion information is obtained by a first merge index specifying one of N candidates included in a merge candidate list, and 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.

In the video signal decoding method according to the present disclosure, 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 according to the present disclosure 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.

In the video signal decoding method according to the present disclosure, the representative position sample may be located at the center of the sub-block.

In the video signal decoding method according to the present disclosure, 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.

According to the present disclosure, coding/decoding efficiency can be improved by dividing a coding block into a plurality of prediction units.

According to the present disclosure, in encoding/decoding a video signal, 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.

The effects obtainable in the present disclosure are not limited to the above-mentioned effects, and other effects not mentioned may be clearly understood by those of ordinary skill in the art from the following description. will be.

1 is a block diagram illustrating an image encoding apparatus according to an embodiment of the present disclosure.
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.
5 is a flowchart illustrating an inter prediction method according to an embodiment to which the present invention is applied.
6 is a diagram for explaining a location of a spatial neighboring block referenced when inducing motion information of a current block.
7 is an example for explaining a temporal neighboring block referenced to induce motion information of a current block.
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.
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.
10 shows an example of dividing the current block into two triangular partitions.
11 illustrates a case where there are 8 angle candidates.
12 shows an example in which non-uniform division is applied.
13 exemplifies division angle candidates and division distance candidates.
14 to 16 show filter coefficients applied to derive each prediction sample when the current block is divided using a dividing line having an angle of 135 degrees.
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.
21 shows the value of the variable sType for each sub-block.
22 shows an example in which a mixed prediction method is applied.
23 and 24 illustrate an example of determining a prediction technique applied to each partition with reference to an encoding mode of a neighboring block.
25 shows an example of determining the boundary strength.
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.
27 illustrates an area to which a transformation is applied.

In the present disclosure, various modifications may be made and various embodiments may be provided, and specific embodiments will be illustrated in the drawings and will be described in detail in the detailed description. However, this is not intended to limit the present disclosure to a specific embodiment, it should be understood to include all changes, equivalents, and substitutes included in the spirit and scope of the present disclosure. In describing each drawing, similar reference numerals have been used for similar elements.

Terms such as 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.

When a component is referred to as being "connected" or "connected" to another component, it is understood that it may be directly connected or connected to the other component, but other components may exist in the middle. It should be. On the other hand, when a component is referred to as being "directly connected" or "directly connected" to another component, it should be understood that there is no other component in the middle.

The terms used in the present application are used only to describe specific embodiments, and are not intended to limit the present disclosure. Singular expressions include plural expressions unless the context clearly indicates otherwise. In the present application, terms such as "comprise" or "have" are intended to designate the presence of features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, but one or more other features. It is to be understood that the presence or addition of elements or numbers, steps, actions, components, parts, or combinations thereof does not preclude in advance.

Hereinafter, a preferred embodiment of the present disclosure will be described in more detail with reference to the accompanying drawings. Hereinafter, the same reference numerals are used for the same elements in the drawings, and duplicate descriptions for the same elements are omitted.

1 is a block diagram illustrating an image encoding apparatus according to an embodiment of the present disclosure.

Referring to FIG. 1, 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.

In addition, some of 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. In this case, 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.

For example, one picture may be divided into a plurality of coding units. In order to split the coding units in a picture, 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.

Hereinafter, in an embodiment of the present disclosure, 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.

When a prediction unit that performs intra prediction based on a coding unit is not a minimum coding 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. In this case, 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. In addition, 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. In the case of using a specific encoding mode, it is possible to encode an original block as it is and transmit it to a decoder without generating a prediction block through the prediction units 120 and 125.

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. In the case of a luminance pixel, 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. In the case of a color difference signal, 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.

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.

In intra prediction, 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.

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.

In the intra prediction method, a prediction block may be generated after applying an adaptive intra smoothing (AIS) filter to a reference pixel according to a prediction mode. The type of AIS filter applied to the reference pixel may be different. To perform the intra prediction method, 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. When predicting the prediction mode of the current prediction unit using 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.

In addition, 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.

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. Depending on the size of the transform unit and the intra prediction mode, instead of zig-zag scan, 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).

The deblocking filter can remove block distortion caused by the boundary between blocks in the reconstructed picture. In order to determine whether to perform deblocking, 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. When applying a deblocking filter to a block, a strong filter or a weak filter may be applied according to the required deblocking filtering strength. In addition, in applying the deblocking filter, 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. To perform offset correction for a specific picture, 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 (ALF) 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.

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.

2 is a block diagram illustrating an image decoding apparatus according to an embodiment of the present disclosure.

Referring to FIG. 2, 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.

When an image bitstream is input by the image encoding apparatus, 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.

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.

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.

As described above, if the size of the prediction unit and the size of the transformation unit are the same when intra prediction is performed in the same manner as the operation of the video encoding apparatus, 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. In addition, 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.

In order to perform inter prediction, 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.

When the prediction mode of the prediction unit is a prediction unit that performs intra prediction based on a pixel value obtained by interpolating a reference pixel, 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.

Information on whether a deblocking filter is applied to a corresponding block or picture from the video encoding apparatus, and when a deblocking filter is applied, information on whether a strong filter or a weak filter is applied may be provided. 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.

As described above, in the embodiment of the present disclosure hereinafter, for convenience of description, a coding unit is used as a coding unit, but may be a unit that performs not only encoding but also decoding.

In addition, 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. In this specification,'unit' denotes a basic unit for performing a specific encoding/decoding process, and'block' may denote a pixel array having a predetermined size. Unless otherwise specified,'block' and'unit' may be used interchangeably. For example, in an embodiment to be described later, it may be understood that the coding block (coding block) and the coding unit (coding unit) have the same meaning as each other.

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.

In addition, the coding block may be divided into a plurality of prediction blocks or a plurality of transform blocks.

Redundant data between pictures is removed through inter prediction. When a prediction block is to be generated by inter prediction, motion information can be used. Here, the motion information may include at least one of a motion vector, a reference image index, and a prediction direction.

From the reference image of the current picture, 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.

In the illustrated example, 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.

4 shows an example in which bidirectional prediction is performed.

In the example shown in FIG. 4, through L0 prediction, a reference block 0 spaced apart by a motion vector (x0, y0) is specified from a collocated block in the reference picture T-1. Also, through L1 prediction, a reference block 1 spaced apart by a motion vector (x1, y1) from the collocated block in the reference picture T+1 is specified.

Accordingly, 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.

When bi-directional prediction is performed, 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.

In the illustrated example, it is illustrated that the L0 direction is the previous direction of the current picture, and the L1 direction is the subsequent direction of the current picture. Unlike the illustrated example, 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. Alternatively, the L0 direction may be a direction after the current picture, and the L1 direction may be set as a previous direction of the current picture.

Accordingly, 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.

When inter prediction is applied, motion information of a current block must be encoded and transmitted to a decoder. In this case, the motion vector may be encoded as it is and transmitted to the decoder. Alternatively, 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. Hereinafter, this will be described in detail.

5 is a flowchart illustrating an inter prediction method according to an embodiment to which the present invention is applied.

Referring to FIG. 5, 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.

6 is a diagram for explaining a location of a spatial neighboring block referenced when inducing motion information of a current block.

In FIG. 6, LB indicates the position of the lower left sample in the current block, and 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, and 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. For example, at least one of A0, A1, B0, B1, or B5 may be defined as a reference position for determining a spatial neighboring block.

7 is an example for explaining a temporal neighboring block referenced to induce motion information of a current block.

In the example shown in FIG. 7, 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. For example, at least one of C21 and C35 may be defined as a reference position for determining a spatial neighboring block.

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.

When the merge mode is applied to the 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. For example, the motion vector of the temporal merge candidate may be determined based on the motion vector of the temporal neighboring block. Also, 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. However, 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'.

Thereafter, 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.

When the number of merge candidates included in the merge candidate list does not reach the maximum number, 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.

Alternatively, the number of motion information that can be stored in the motion information buffer may be limited to M pieces. In this case, 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. Here, the distance from the current block may be determined by a difference in encoding/decoding order from the current block.

That is, according to the encoding/decoding order, a higher priority may be given to a block encoded with a lower priority. When adding motion information of a specific block to the wide motion information buffer, if the maximum number of motion information is already stored in the wide motion information buffer, 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.

As another example, when a parallel processing structure is applied, a wide motion information buffer may be generated for each area to be processed in parallel.

Alternatively, motion information of a block included in a parallel-processed area may not be added to the wide motion information buffer.

Alternatively, 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.

When a wide-area motion information buffer is individually defined and used for each area in which parallel processing is performed, the number of motion information stored in the wide-area motion information buffer is very small at the beginning of the area in which parallel processing is performed. In order to increase the encoding/decoding efficiency of blocks corresponding to the first half of a region where parallel processing is performed, preset initial motion information may be inserted into the wide motion information buffer. For example, 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.

Alternatively, global motion information may be defined in units of tiles, bricks, or slices.

When the merge candidate list is generated, 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.

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.

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. Here, the upper neighboring block includes at least one of blocks adjacent to the upper or upper right corner of the current block, and 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. I can. 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.

When the reference pictures between the current block and the spatial neighboring block are different, 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). When the reference pictures between the current block and the temporal neighboring block are different, 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).

When the motion vector candidate list is generated, 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). In this case, the motion vector difference value may be parsed through a bitstream.

When motion information of the current block is obtained, 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.

In embodiments to be described later, 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.

10 shows an example of dividing the current block into two triangular partitions.

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. For example, 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. As an example, a first prediction block may be generated based on motion information of a first partition, and a second prediction block may be generated based on motion information of a second partition. When the current block is divided into a plurality of prediction units, each prediction unit may be set to use only unidirectional prediction without using bidirectional prediction.

In addition, each of the partitions can be set to have a different prediction direction. For example, the first?? When prediction in the L0 direction is performed in the partition, it may be set to perform prediction in the L1 direction in the second partition.

For convenience of explanation, in embodiments to be described later, it is assumed that unidirectional prediction is performed for each partition.

The motion information of each partition can be encoded and transmitted to a decoder.

Alternatively, in order to reduce the amount of bits required to encode the motion information, 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.

In this case, 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, and 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.

As another example, an integrated merge candidate list may be configured for the first partition and the second partition. In this case, an index specifying a merge candidate for each partition may be signaled. In this case, when signaling an index for each partition, each partition may be set to refer to a different merge candidate. For example, 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.

For example, in the examples shown in FIGS. 6 and 7, 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.

When constructing a merge candidate list, 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. Nevertheless, when 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 motion information of the spatial/temporal neighboring block for which bidirectional prediction is performed. Among them, only the L1 motion information can be taken and inserted into the merge candidate list. Nevertheless, when 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 the L0 motion of the spatial/temporal neighboring block for which bidirectional prediction is performed. A result derived by weighted summation of the motion vector of the information and the motion vector of the L1 motion information may be inserted into the merge candidate list. In this case, when the L0 reference picture and the L1 reference picture are different, 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.

Nevertheless, when the number of merge candidates included in the merge candidate list is less than the maximum number N, motion information or a zero vector stored in a global motion information buffer (eg, HMVP table) can be added to the merge candidate list. have. 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. When the simplified method is applied, 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.

As an example, it is assumed that the maximum number of merge candidates that can be included in the merge candidate list is six. In addition, in the examples shown in FIGS. 6 and 7, it is assumed that the spatial neighboring blocks and the temporal neighboring blocks have a search order of A1, B1, B0, A0, B5, C35, and C21. In this case, in the order of A1, B1, B0, and A0, 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. Thereafter, 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.

Even though spatial/temporal merge candidates are added to the merge candidate list, when the number of merge candidates included in the merge candidate list is less than the maximum number, motion information may be filled in the list with reference to the global motion information buffer.

When the number of merge candidates included in the merge candidate list generated through the above process has not yet reached the maximum number, a combined merge candidate may be generated by combining motion information of previously inserted merge candidates in the merge candidate list. .

For example, a combined merge candidate may be derived by averaging motion vectors between two merge candidates among merge candidates existing in the merge candidate list. In this case, a combination [i,j] of merge candidates for generating the combined merge candidates may be predefined. Here, i indicates the index of the first merge candidate used to generate the combined merge candidate, and j indicates the index of the second merge candidate used to generate the combined merge candidate. For example, 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

When the [0,1] combination is applied, 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. I can. When generating the combined 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.

In this case, a case in which at least one of the merge candidates to be combined does not have motion information for a specific direction (ie, motion information for the L0 or L1 direction is 0) may occur. For example, when candidates are combined, bidirectional prediction may be performed on a merge candidate having an index of 0, whereas L1-direction prediction may be performed on a merge candidate having an index of 1. In this case, when generating the combined merge candidate for the L0 direction, instead of averaging the L0 motion information of the two merge candidates, L0 information of the merge candidate having an index of 0 may be set as the motion information of the combined merge candidate.

Alternatively, 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.

Alternatively, after the merge candidate list is generated by the above-described method, 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.

In this case, the additionally generated merge candidate list may be generated by extracting unidirectional motion information of merge candidates included in the existing merge candidate list. At this time, 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.

As an example, 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, and L1 motion information of the second merge candidate existing in the two existing merge candidate lists is added to the merge candidate list. Can be set as the second merge candidate of

In the same manner as above, L0 motion information is extracted from the 2N-1 th merge candidate, and 2N times?? L1 motion information can be extracted from the 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.

After the list is generated, an index specifying motion information of each partition may be encoded and signaled.

However, if the motion information of the two partitions is the same, there is no need to divide the current block into two partitions. Accordingly, motion information between the two partitions must be set differently.

When the merge candidate list in which two partitions are integrated is used, 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.

Alternatively, 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.

In addition to the method of signaling the index for each partition, after allocating an index to each of the candidates combining a plurality of merge candidates, the index of the candidate that matches the combination of merge candidates applied to the two partitions may be encoded and signaled. .

Alternatively, after allocating an index to each of the merge candidates and the candidates that combine the partitioning direction of the current block, 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. As an example, when 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?? subblock 5 x the number of merge candidates applicable to the second subblock 4 x the number of partition direction candidates applicable to the current block 2) Accordingly, from 0 for the current block An index representing one of the values between 39 may be encoded and signaled.

The partitioning aspect of the current block is not limited to the illustrated example. The angle α of the dividing line dividing the current block may have a quantized value between 0 degrees and 180 degrees. As an example, when the angle α is one of four divisions in the range of 0° to 180°, the value of the angle α may be one of 0°, 45°, 90°, or 135°. Alternatively, 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.

Here, 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. For example, when the angle α 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 α of 90 degrees and 180 degrees can be excluded from the segmentation candidate.

In the illustrated example, although it is illustrated that the current block is divided into partitions of equal size, the partitions may be divided into non-uniform size.

12 shows an example in which non-uniform division is applied.

β 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. For example, an index specifying one of candidates obtained by combining a division angle and a division distance may be encoded and signaled. In the encoder and the decoder, 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.

13 exemplifies division angle candidates and division distance candidates.

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 0 One 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 One 3 0 One 2 3 0 One 2 3 0 One 2 3 0 One 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 One 3 0 One 2 3 0 One 2 3 0 One 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 One 2 3 One 3 One 2 3 One 2 3 One 2 3 One 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 One 3 One 2 3 One 2 3 One 2 3 One 2 3

In Table 1, 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.

13B illustrates a division pattern according to a distanceIdx value when angleIdx is 4. 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. As an example, 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.

For example, 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. . Here, W represents the width of the current block, and H represents the height of the current block.

As described above, 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. For example, a quantization variable of the angle α of the dividing line (eg, the number of partitions in a range from 0 to 180) may be encoded and signaled through an upper header.

Alternatively, the number of division distances may be encoded and signaled through an upper header.

As another example, 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.

When constructing a combination candidate, a high index priority can be given to a specific angle. For example, a combination candidate including an angle of 45 degrees or 135 degrees may be given a high priority. Here, the higher the priority, the smaller the value of the index assigned to the corresponding combination. When a specific combination is given a high priority, 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.

For example, when the color difference format is 4:2:0, an 8x8 luma block corresponds to a 4x4 chroma block. In this case, when diagonal division with distanceIdx of 3 is applied, the width or height of one of the partitions generated by dividing the chroma block may be 1. When considering encoding/decoding efficiency, it is not desirable to create a partition having a width or height of 1, and in the above case, 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.

Alternatively, the threshold value may be pre-stored in the encoder and the decoder. Alternatively, 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.

On the other hand, when the size of the current block is larger than 8x8, 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. For example, 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. For example, the information may be a 1-bit flag.

Alternatively, it may be determined whether the luma block and the division type of the chroma block are independently determined based on at least one of the size, shape, or color difference format of the chroma block.

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.

When the boundaries of the partitions in the current block are determined, a weight for each prediction sample may be determined based on the determined boundary and the size of the current block. Here, 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.

(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

In Table 2, 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. Based on the variables derived through Table 2, the variables offsetX and offsetY may be determined. 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.

(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

Thereafter, based on the pixel positions (x, y), variables xL and yL may be derived as shown in Table 4.

(14) xL = (cIdx = = 0)? x: x * SubWidthC
(15) yL = (cIdx = = 0)? y: y * SubHeightC

Thereafter, 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.

(16) weightIdx = (((xL + offsetX) << 1) + 1) * disLut [displacementX] + (((yL + offsetY) << 1) + 1)) * disLut [displacementY]

In Table 5, disLut represents a lookup table, and 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.

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

Thereafter, 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.

(17) weightIdxL = partFlip? 32 + weightIdx: 32-weightIdx
(18) wValue = Clip3( 0, 8, (weightIdxL + 4) >> 3)

The first weight applied to the first prediction sample is set equal to the variable wValue, and 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. Alternatively, on the contrary, the variable wValue may be set as the second weight, and 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.

In Equation 1, P1 and P2 represent a first prediction sample and a second prediction sample, respectively. In addition, α1 and α2 represent a first weight and a second weight, respectively.

14 to 16 show filter coefficients applied to derive each prediction sample when the current block is divided using a dividing line having an angle of 135 degrees.

14 to 16(a) show a first weight applied to each prediction sample, and FIGS. 14 to 16(b) show a second weight applied to each prediction sample.

As in the example illustrated in FIG. 14, 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. . For example, 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.

Alternatively, 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.

Alternatively, information indicating the number of diagonal lines included in the filter application region may be encoded and signaled. In this case, 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. For example, for each of the luma component and the chroma component, an index for specifying a filter application region may be encoded.

Alternatively, a filter application area may be predefined for each color component. As an example, for the luma 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.

Alternatively, the size/shape of the filter application area may be determined based on motion information of the first partition and the second partition. For example, 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. Alternatively, 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.

Alternatively, the size/shape of the filter application area may be determined based on at least one of the division angle α and the division distance β.

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.

When the current block is encoded by inter prediction, motion information may be stored in units of blocks having a predefined size. Here, the predefined size may be any one of 4x4, 8x8, 2x8, or 8x2. Alternatively, information indicating the size of a unit in which motion information is stored may be encoded and signaled.

In an embodiment to be described later, it is assumed that a unit in which motion information is stored has a size of 4x4. In addition, a 4x4 size region in which motion information is stored will be referred to as a subblock.

When the current block is divided into a plurality of prediction units, for a subblock that does not belong to a division boundary (eg, a subblock that does not pass a division line), motion information of a partition to which it belongs may be stored as motion information of the corresponding subblock. . On the other hand, when 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. have.

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, and 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.

For example, when the prediction directions between the first partition and the second partition are different, bidirectional motion information obtained by combining motion information of the first partition and motion information of the second partition may be stored in subblocks.

On the other hand, when the prediction directions of the first partition and the second partition are the same, either motion information of the first partition or motion information of the second partition may be stored in the subblock.

For example, among the motion information of the first partition and the motion information of the second partition, 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. I can. Alternatively, 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.

Alternatively, based on 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 β 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.

Alternatively, by comparing the POC difference between the reference picture index of the first partition and the current picture and the POC difference between the reference picture index of the second partition and the current picture, 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.

19 shows an example of this.

In the example illustrated in FIG. 19, 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.

As another example, it may be determined whether the sub-block is located at the division boundary using the representative weight within the sub-block.

A weight applied to a sample at a specific location within a sub-block may be set as a representative weight. Here, 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.

Alternatively, it may be determined whether a corresponding sub-block is located at a division boundary using a maximum value, a minimum value, or an average value of the weight in the sub-block.

20 is a diagram for explaining a representative weight.

As in the illustrated example, a weight applied to a sample at a center position within a 4x4 subblock may be set as a representative weight. For example, a weight applied to a sample at a position (2, 2) in a sub-block may be set as a representative weight.

Based on items (10) to (13) illustrated in Table 3, after calculating the variables offsetX and offsetY of each partition, 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.

In Equation 2, (xSbIdx, ySbIdx) represents the index of a subblock in the current block. Specifically, the variable xSbIdx represents the horizontal direction index of the sub-block, and the variable ySbIdx represents the vertical direction index of the sub-block. For example, when the current block has a size of 8x8, 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.

Thereafter, the motion information storage type of the sub-block may be determined by using the temporary weight index. As an example, based on Equation 3 below, a variable sType indicating a storage type of motion information of a sub-block may be derived.

Alternatively, the variable sType may be derived using Equation 4 below.

In Equation 4, 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.

21 shows the value of the variable sType for each sub-block.

The variable sType can be derived using the temporary weight index for the center coordinates in each sub-block.

In the illustrated example, it is illustrated that 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.

On the other hand, the sType of the lower left sub-block not included in the boundary area was derived as 0. In this case, motion information of a partition including the sub-block (ie, a partition located below the current block) may be stored in the sub-block. In addition, the sType of the upper right sub-block not included in the boundary area was derived as 1. In this case, motion information of a partition including the sub-block (ie, a partition positioned at the top of the current block) may be stored in the sub-block.

Alternatively, a variable sType of 2 may be defined as storing motion information of either the first partition or the second partition. For example, in a subblock having a variable sType of 2, motion information of a partition having a larger area occupying the subblock may be stored. Alternatively, by comparing the representative weight at the center position (eg, (2, 2) position) and the weight of the neighboring position (eg, at least one of the left, right, top, or bottom) of the center location, the first partition or the first partition One of the 2 partitions may be selected.

Alternatively, 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.

Alternatively, 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. Here, 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. Here, motion information stored in the sub-block at a preset location may be determined based on a variable sType for the sub-block.

Alternatively, when the size of the block of the current block is smaller than the threshold, 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.

When the current block is divided into two partitions, a prediction method in which intra prediction and inter prediction are mixed may be applied.

22 shows an example in which a mixed prediction method is applied.

When the current block is divided into two partitions, intra prediction may be applied to one of the first partition and the second partition, and 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. As an example, MPM candidates may be derived based on an intra prediction mode of a neighboring block. Here, the neighboring block may include at least one of an upper neighboring block or a left neighboring block. Among a plurality of MPM candidates included in the MPM candidate list, the index of the MPM candidate that is the same as the intra prediction mode of the current block may be encoded and signaled.

Alternatively, when a combined prediction method is applied, a fixed intra prediction mode may be used. For example, when the combined prediction method is applied, the planar mode may be fixedly used.

Alternatively, the one having the highest frequency of use among intra prediction modes of neighboring blocks adjacent to the current block may be used.

Alternatively, one of the default modes may be used for intra prediction modes. Here, 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. Here, 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.

23 is an example of a case where the angle of the dividing line dividing the current block is 45 degrees.

In the example shown in FIG. 23, 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.

Specifically, when the upper neighboring block and the left neighboring block have the same coding mode, the same prediction technique as the coding mode may be applied to the first partition. As an example, as in the example shown in (a) of FIG. 23, when both the upper neighboring block and the left neighboring block are encoded by intra prediction, intra prediction may also be applied to the first partition. On the other hand, as in the example shown in (d) of FIG. 23, when both the upper neighboring block and the left neighboring block are encoded by inter prediction, inter prediction may be applied to the first partition.

As in the examples shown in FIGS. 23B and 23C, when the coding modes of the upper neighboring block and the left neighboring block are different, the prediction technique having a high priority among intra prediction or inter prediction is the first. Can be applied to partitions. Alternatively, when encoding modes of the upper neighboring block and the left neighboring block are different, 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.

24 is an example of a case where the angle of the dividing line dividing the current block is 135 degrees.

In the example shown in FIG. 24, the upper neighboring block touches the boundary of the first partition, and the left neighboring block touches the boundary of the second partition.

Accordingly, 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.

For example, as in the examples shown in (b) and (c) of FIG. 24, when the coding mode of the upper neighboring block and the coding mode of the left neighboring block are different, the same prediction method as the upper neighboring block is used in the first partition. Applied, and the same prediction technique as the left neighboring block may be applied to the second partition.

As in the examples shown in FIGS. 24A and 24D, when the upper neighboring block and the left neighboring block have the same encoding mode, information indicating a prediction technique applied to the first partition or the second partition is encoded. Can be signaled.

Alternatively, when the first partition and the second partition have different sizes, 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.

Alternatively, 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.

Even when the mixed prediction method is applied, a filter may be applied to the boundary between partitions. In this case, a method of determining a weight applied to each prediction sample is as described above.

Alternatively, the size of the filter area may be adjusted according to the block size.

Alternatively, when a filter is applied, 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.

Alternatively, when a prediction method in which inter prediction and intra prediction are mixed is applied, a filter may be set not to be applied to the partition boundary (eg, applying the example of FIG. 16 ).

When a block is reconstructed, a deblocking filter between reconstructed blocks may be applied.

In this case, when the mixed prediction method is applied to the current block, the encoding mode applied to the current block is regarded as intra prediction, and a boundary strength may be determined. Alternatively, when 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.

Alternatively, 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.

25 shows an example of determining the boundary strength.

When the deblocking filter is applied to the upper edge of the current block, 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.

When the deblocking filter is applied to the right side of the current block, 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.

In the example shown in (a) of FIG. 26, it is assumed that 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. In this case, when inducing the inter prediction information of the current block, motion information of the upper neighboring block may be referred to. As an example, a merge candidate may be derived by referring to motion information of an upper neighboring block.

In the example shown in (b) of FIG. 26, it is assumed that a mixed prediction method is applied to a block adjacent to the left of the current block, and intra prediction is applied to the current block. In this case, when inducing the intra prediction information of the current block, the intra prediction information of the left neighboring block may be referred to. As an example, the MPM candidate of the current block may be derived based on the intra prediction mode of the left neighboring block.

Alternatively, when the mixed prediction method is applied to the neighboring block, it is assumed that 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. Here, the predefined intra prediction mode may be a flat mode.

As another example, 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. In addition, after quantizing the derived transform coefficients, the quantized transform coefficients may be encoded.

In this case, when the current block is divided into two partitions, the transformation may be applied only to a partial area within the current block. Here, 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.

27 illustrates an area to which a transformation is applied.

In the example of FIG. 27, the variable W represents the height of the current block, and the variable H represents the width of the current block. In FIG. 27, the bold lines indicate dividing lines for dividing the current block.

As illustrated in FIG. 27, 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. In addition, 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. For example, as in the example shown in (a) of FIG. 27, the reduced transform area may have a shape reduced by 1/2 of the width W and height H of the current block. Alternatively, as in the example shown in (b) or (c) of FIG. 27, 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.

Information indicating the reduction ratio of the transform region may be encoded and signaled. Here, the reduction ratio represents the size of the transformed area compared to the size of the current block. As an example, an index specifying one of a plurality of reduction ratio candidates is encoded?? Can be signaled.

Information indicating the reduction ratio may be signaled for each of the width and height. For example, 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.

Alternatively, a reduction ratio of the reduced transform region may be derived based on at least one of the angle α of the dividing line dividing the current block and the distance β of the dividing line.

Alternatively, 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.

Alternatively, 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. In this case, 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. In this case, 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. For example, when the CBF value is not 0 (that is, when it is determined that a non-zero transform coefficient exists in the current block), the decoder may additionally parse information related to the reduced transform region.

As another example, 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.

As another example, according to the encoding mode of the current block, it may be determined whether the reduced transform region is usable. Here, the encoding mode represents intra prediction or inter prediction. For example, when the current block is encoded by intra prediction, a transformation method using a reduced transformation region may be set to be unavailable. On the other hand, when the current block is encoded by inter prediction, information related to the reduced transform region may be signaled.

Alternatively, 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.

As another example, information indicating whether a conversion method based on a reduced conversion area is available may be signaled through an upper header. Here, the upper header includes at least one of a sequence, a picture header, and a slice header. As long as the information indicates that a transform method based on the reduced transform region is available, information related to the reduced transform region at the block level may be signaled. On the other hand, when the information indicates that the transformation method based on the reduced transformation region is not available, information related to the reduced transformation region at the block level may not be signaled.

Alternatively, based on at least one of the size or shape of the current block, it may be determined whether or not a transformation method based on the reduced transformation region is available. As an example, when the number of samples included in the current block is 64 or less, it may be set that a transformation method based on a reduced transformation region can be used. In this case, information on the reduced transformation area may be signaled. On the other hand, when the number of samples included in the current block exceeds 64, 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.

In the above-described example, it has been exemplified that the reduced transform area is set in consideration of the division boundary. As another example, it is possible to set the reduced transform area irrespective of the division boundary.

In the reduced transform area, second-order transform may not be performed.

Alternatively, second-order transformation may be selectively performed on the reduced transformation area. In this case, information indicating whether the second-order transformation has been performed in the reduced transformation region may be encoded and signaled.

The names of the syntaxes used in the above-described embodiments are only named for convenience of description.

It is included in the scope of the present disclosure to apply the embodiments described centering on the decoding process or the encoding process to the encoding process or the decoding process. It is also included in the scope of the present disclosure to change the embodiments described in a predetermined order in an order different from that described.

The above-described embodiment has been described based on a series of steps or flow charts, but this does not limit the time-series order of the invention, and may be performed simultaneously or in a different order as necessary. In addition, 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.

Claims (13)

When it is determined that the current block is divided into a first partition and a second partition, parsing an index indicating a partition type of the current block;
Determining first motion information for the first partition;
Determining second motion information for the second partition; And
Inducing 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,
The index is characterized in that it specifies 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 Way. The method of claim 1,
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 obtained by a second merge index specifying one of N-1 candidates excluding a merge candidate corresponding to the first merge index in the merge candidate list. . The method of claim 1,
The final prediction sample is obtained based on a weighted sum operation between the first prediction sample and the second prediction sample,
A first weight applied to the first prediction sample and a second weight applied to the second prediction sample are determined based on the angle index and the distance index. The method of claim 1,
Further comprising the step of storing motion information in units of sub-blocks having a predefined size in the current block,
The storage type of the sub-block is determined based on a weight of the representative position sample of the sub-block. The method of claim 4,
The representative position sample is located at the center of the sub-block. The method of claim 4,
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 the first motion information and the One of the third storage types for allocating bidirectional information derived by combining second motion information. Dividing the current block into a first partition and a second partition, wherein an index indicating a division type of the current block is encoded in the bitstream;
Determining first motion information for the first partition;
Determining second motion information for the second partition; And
Inducing 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,
The index is characterized in that to 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 Way. The method of claim 7,
After a first merge index indicating a first merge candidate that is the same as the first motion information among N candidates included in the merge candidate list is encoded,
And a second merge index indicating a second merge candidate identical to the second motion information among N-1 candidates excluding the first merge candidate in the merge candidate list is encoded. The method of claim 7,
The final prediction sample is obtained based on a weighted sum operation between the first prediction sample and the second prediction sample,
A first weight applied to the first prediction sample and a second weight applied to the second prediction sample are determined based on the angle index and the distance index. The method of claim 7,
Further comprising the step of storing motion information in units of sub-blocks having a predefined size in the current block,
The storage type of the sub-block is determined based on a weight of the representative position sample of the sub-block. The method of claim 10,
The representative position sample is located at the center of the sub-block. The method of claim 10,
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 the first motion information and the One of the third storage types for allocating bidirectional information derived by combining second motion information. A computer-readable recording medium storing a bitstream encoded by a video encoding method, comprising:
The video encoding method,
Dividing the current block into a first partition and a second partition, wherein an index indicating a division type of the current block is encoded in the bitstream;
Determining first motion information for the first partition;
Determining second motion information for the second partition; And
Inducing 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,
The index is characterized in that it specifies 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. Readable recording medium.

Images