Image encoding / decoding method and device, and recording medium storing bit stream

Abstract

Disclosed are an image encoding / decoding method and apparatus, and a recording medium storing a bitstream. An image decoding method of the present invention comprises the steps of: determining whether a current block is in a bidirectional optical flow mode; if the current block is in the bidirectional optical flow mode, gradient information about prediction sample points of the current block is calculated; and generating a prediction block of the current block by using the calculated gradient information, in which the step of calculating gradient information on a prediction sample point of the current block calculates gradient information by using at least one adjacent sample point adjacent to the prediction sample point.

Description

[0001] This application is a divisional application of an invention patent application with an application date of December 20, 2019, application number "201980085120.7", and invention name "Image encoding / decoding method and device and recording medium for storing bit stream". Technical Field

[0002] The present invention relates to an image encoding / decoding method and apparatus and a recording medium storing a bit stream. More specifically, the present invention relates to a method and apparatus for encoding / decoding an image on a block basis using bidirectional optical flow (BIO). Background Art

[0003] Recently, in various applications, the demand for high-resolution and high-quality images such as high-definition (HD) or ultra-high-definition (UHD) images has increased. As the resolution and quality of images increase, the amount of data increases accordingly. This is one of the reasons for the increase in transmission cost and storage cost when image data is transmitted through existing transmission media such as wired or wireless broadband channels or when image data is stored. In order to solve these problems of high-resolution and high-quality image data, efficient image encoding / decoding technology is required.

[0004] There are various video compression techniques, such as inter-frame prediction techniques that predict the values ​​of pixels in a current picture from the values ​​of pixels in a previous picture or a subsequent picture, intra-frame prediction techniques that predict the values ​​of pixels in an area of ​​the current picture from the values ​​of pixels in another area of ​​the current picture, transformation and quantization techniques for compressing the energy of residual signals, and entropy coding techniques that assign shorter codes to frequently occurring pixel values ​​and longer codes to less frequently occurring pixel values.

[0005] In a conventional image encoding / decoding method and apparatus using bidirectional optical flow (BIO), since BIO is applicable only to a block having two pieces of motion information, BIO is not applicable to a block having one piece of motion information.

[0006] Furthermore, in a conventional image encoding / decoding method and apparatus using BIO, a gradient value may be calculated using pixel values ​​outside a target block region, thereby increasing memory bandwidth or the amount of calculation. Summary of the invention

[0007] Technical issues

[0008] An object of the present invention is to provide a method and apparatus capable of applying bidirectional optical flow (BIO) by deriving second motion information when an encoding / decoding target block has only one piece of motion information under a condition that bidirectional prediction is enabled.

[0009] Another object of the present invention is to provide a method / device capable of variably providing a unit size of a subgroup for obtaining a BIO offset to reduce complexity, a method / device for calculating a BIO offset in units of subgroups, and a method / device capable of encoding / decoding by selecting whether to apply BIO in units of blocks.

[0010] Another object of the present invention is to provide a method / apparatus for calculating gradients and BIO parameters to reduce memory bandwidth when applying BIO.

[0011] Technical Solution

[0012] A method for decoding an image according to an embodiment of the present invention may include: determining whether a current block is in a bidirectional optical flow (BIO) mode, calculating gradient information of a prediction sample of the current block when the current block is in the BIO mode, and generating a prediction block of the current block using the calculated gradient information. The step of calculating the gradient information of the prediction sample of the current block may include: calculating the gradient information using at least one neighboring sample adjacent to the prediction sample.

[0013] In the image decoding method of the present invention, when the neighboring sample points are located outside the area of ​​the current block, the sample point value of the integer pixel position closest to the neighboring sample points may be used as the value of the neighboring sample points.

[0014] In the image decoding method of the present invention, the gradient information may be calculated in units of sub-blocks having a predefined size.

[0015] In the image decoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: determining whether the current block is in the BIO mode based on the distance between the first reference picture of the current block and the current picture and the distance between the second reference picture of the current block and the current picture.

[0016] In the image decoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: when the distance between the first reference picture and the current picture is different from the distance between the second reference picture and the current picture, determining that the current block is not in the BIO mode.

[0017] In the image decoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: determining whether the current block is in the BIO mode based on a type of a reference picture of the current block.

[0018] In the image decoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: when at least one of the type of the first reference picture of the current block or the type of the second reference picture of the current block is not a short-term reference picture, determining that the current block is not in the BIO mode.

[0019] In the image decoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: determining whether the current block is in the BIO mode based on the size of the current block.

[0020] A method for encoding an image according to an embodiment of the present invention may include: determining whether a current block is in a bidirectional optical flow (BIO) mode, calculating gradient information of a prediction sample of the current block when the current block is in the BIO mode, and generating a prediction block of the current block using the calculated gradient information. The step of calculating the gradient information of the prediction sample of the current block may include: calculating the gradient information using at least one neighboring sample adjacent to the prediction sample.

[0021] In the image encoding method of the present invention, when the neighboring sample points are located outside the area of ​​the current block, the sample point value of the integer pixel position closest to the neighboring sample points may be used as the value of the neighboring sample points.

[0022] In the image encoding method of the present invention, the gradient information may be calculated in units of sub-blocks having a predefined size.

[0023] In the image encoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: determining whether the current block is in the BIO mode based on the distance between the first reference picture of the current block and the current picture and the distance between the second reference picture of the current block and the current picture.

[0024] In the image encoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: when the distance between the first reference picture and the current picture is different from the distance between the second reference picture and the current picture, determining that the current block is not in the BIO mode.

[0025] In the image encoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: determining whether the current block is in the BIO mode based on a type of a reference picture of the current block.

[0026] In the image encoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: when at least one of the type of the first reference picture of the current block or the type of the second reference picture of the current block is not a short-term reference picture, determining that the current block is not in the BIO mode.

[0027] In the image encoding method of the present invention, the step of determining whether the current block is in the BIO mode may include: determining whether the current block is in the BIO mode based on the size of the current block.

[0028] A computer-readable recording medium according to an embodiment of the present invention may be a non-transitory computer-readable recording medium storing a bit stream generated by a method for encoding an image, wherein the method for encoding an image includes: determining whether a current block is in a bidirectional optical flow (BIO) mode, calculating gradient information of a prediction sample of the current block when the current block is in the BIO mode, and generating a prediction block of the current block using the calculated gradient information, wherein the step of calculating the gradient information of the prediction sample of the current block includes: calculating the gradient information using at least one neighboring sample adjacent to the prediction sample.

[0029] Beneficial Effects

[0030] According to the present invention, an image encoding / decoding method and apparatus with improved encoding / decoding efficiency can be provided.

[0031] According to the present invention, a method and apparatus capable of applying bidirectional optical flow (BIO) by deriving second motion information when an encoding / decoding target block has only one piece of motion information under a condition that bidirectional prediction is possible can be provided.

[0032] According to the present invention, a method / device capable of variably providing a unit size of a subgroup for obtaining a BIO offset to reduce complexity, a method / device for calculating a BIO offset in units of subgroups, and a method / device capable of encoding / decoding by selecting whether to apply BIO in units of blocks can be provided.

[0033] According to the present invention, the memory bandwidth and the amount of calculation in the calculation of the gradient used in the BIO can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a block diagram showing a configuration of an encoding device according to an embodiment to which the present invention is applied.

[0035] Figure 2 is a block diagram showing a configuration of a decoding device according to an embodiment to which the present invention is applied.

[0036] Figure 3 is a diagram schematically showing a partition structure of an image when the image is encoded and decoded.

[0037] Figure 4 is a diagram illustrating an intra prediction process.

[0038] Figure 5 is a diagram illustrating an embodiment of an inter-picture prediction process.

[0039] Figure 6 is a diagram illustrating transform and quantization processing.

[0040] Figure 7 is a diagram showing reference samples that can be used for intra prediction.

[0041] Figure 8 is a diagram illustrating various embodiments of deriving second motion information based on first motion information.

[0042] Fig. 9 is a diagram illustrating an example of calculating gradient values ​​of vertical and horizontal components.

[0043] Fig.10 is a diagram illustrating various embodiments of subsets as units for calculating BIO offsets.

[0044] Fig.11 is a diagram showing the weights that may be applied to each BIO-related parameter value in a subgroup.

[0045] Fig.12 is a diagram showing an embodiment of performing weighted summation on BIO-related parameter values ​​only at specific positions in a subgroup.

[0046] Fig.13 is a diagram showing an embodiment of calculating BIO-related parameter values.

[0047] Fig.14 The calculation of the BIO related parameter S when the size of the subgroup is 4×4 is shown in FIG. group Schematic diagram of an embodiment of the present invention.

[0048] Fig.15 is a diagram illustrating an embodiment of deriving a motion vector of a chrominance component based on a luminance component.

[0049] Fig.16 is an exemplary diagram illustrating motion compensation processing for chroma components.

[0050] Figures 17 to 20 is a diagram illustrating various embodiments of deriving gradient values ​​in a BIO by filling unavailable pixels outside a block boundary with inner boundary pixels of the block in order to calculate gradient values ​​in the BIO.

[0051] Fig.21 is a flowchart illustrating an image decoding method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] Various modifications may be made to the present invention, and there are various embodiments of the present invention, wherein examples of various embodiments of the present invention will now be provided with reference to the accompanying drawings and described in detail. However, the present invention is not limited thereto, although the exemplary embodiments may be interpreted as including all modifications, equivalents or substitutions within the technical concept and technical scope of the present invention. In various aspects, similar figure numerals refer to the same or similar functions. In the accompanying drawings, the shapes and sizes of the elements may be exaggerated for clarity. In the following detailed description of the present invention, reference is made to the accompanying drawings that illustrate specific embodiments of the present invention in a graphical manner. These embodiments are described in sufficient detail to enable those skilled in the art to implement the present disclosure. It should be understood that the various embodiments of the present disclosure, although different, are not necessarily mutually exclusive. For example, without departing from the spirit and scope of the present disclosure, the specific features, structures and characteristics described herein in conjunction with one embodiment may be implemented in other embodiments. In addition, it should be understood that the position or arrangement of each element within each disclosed embodiment may be modified without departing from the spirit and scope of the present disclosure. Therefore, the following detailed description should not be regarded as having a limiting meaning, and the scope of the present disclosure is limited only by the appended claims (when properly interpreted, together with the full range of equivalents claimed by the claims).

[0053] The terms "first", "second", etc. used in the specification may be used to describe various components, but the components should not be interpreted as being limited to these terms. These terms are only used to distinguish one component from other components. For example, without departing from the scope of the present invention, a "first" component may be named a "second" component, and a "second" component may also be similarly named a "first" component. The term "and / or" includes a combination of multiple items or any one of the multiple items.

[0054] It will be understood that in this specification, when an element is simply referred to as being “connected to” or “coupled to” another element rather than being “directly connected to” or “directly coupled to” another element, the element may be “directly connected to” or “directly coupled to” another element, or connected to or coupled to another element with other elements interposed therebetween. Conversely, it will be understood that when an element is referred to as being “directly coupled to” or “directly connected to” another element, there are no intervening elements.

[0055] In addition, the components shown in the embodiments of the present invention are shown independently to represent the characteristic functions that are different from each other. Therefore, this does not mean that each component is composed of a separate hardware or software component unit. In other words, for convenience, each component includes each component in the listed components. Therefore, at least two components of each component can be combined to form a component, or a component can be divided into multiple components to perform each function. If it does not depart from the essence of the present invention, the embodiment in which each component is combined and the embodiment in which a component is divided are also included in the scope of the present invention.

[0056] The terms used in this specification are only used to describe specific embodiments and are not intended to limit the present invention. Unless there is a significantly different meaning in the context, the expression used in the singular includes the expression in the plural form. In this specification, it will be understood that terms such as "including", "having" etc. are intended to indicate the presence of features, numbers, steps, actions, elements, parts or combinations thereof disclosed in the specification, and are not intended to exclude the possibility that one or more other features, numbers, steps, actions, elements, parts or combinations thereof may exist or may be added. In other words, when a particular element is referred to as "included", it does not exclude elements other than the corresponding element, but may include other elements in an embodiment of the present invention or in the scope of the present invention.

[0057] In addition, some components may not be essential components for performing the basic functions of the present invention, but rather selective components that only improve its performance. The present invention may be implemented by including only essential components for implementing the essence of the present invention without including components for improving performance. Structures that include only essential components without including selective components that only improve performance are also included within the scope of the present invention.

[0058] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. When describing the exemplary embodiments of the present invention, well-known functions or constructions will not be described in detail because they may unnecessarily obscure the understanding of the present invention. The same constituent elements in the accompanying drawings are represented by the same reference numerals, and repeated descriptions of the same elements will be omitted.

[0059] Hereinafter, an image may refer to a picture constituting a video, or may refer to the video itself. For example, "encoding or decoding an image or both encoding and decoding" may refer to "encoding or decoding a moving picture or both encoding and decoding", and may refer to "encoding or decoding one of the images of the moving picture or both encoding and decoding".

[0060] Hereinafter, the terms "motion picture" and "video" may be used as the same meaning and may be replaced with each other.

[0061] Hereinafter, a target image may be an encoding target image as an encoding target and / or a decoding target image as a decoding target. In addition, a target image may be an input image input to an encoding device, and an input image input to a decoding device. Here, the target image may have the same meaning as the current image.

[0062] Hereinafter, the terms "image", "picture", "frame" and "screen" may be used as the same meaning and may be replaced with each other.

[0063] Hereinafter, the target block may be an encoding target block as an encoding target and / or a decoding target block as a decoding target. In addition, the target block may be a current block as a target of current encoding and / or decoding. For example, the terms "target block" and "current block" may be used as the same meaning and may be replaced with each other.

[0064] Hereinafter, the terms "block" and "unit" may be used as the same meaning and may be replaced with each other. Alternatively, a "block" may refer to a specific unit.

[0065] In the following, the terms "region" and "segment" are used interchangeably.

[0066] Hereinafter, a specific signal may be a signal representing a specific block. For example, an original signal may be a signal representing a target block. A prediction signal may be a signal representing a prediction block. A residual signal may be a signal representing a residual block.

[0067] In an embodiment, each of the specific information, data, flags, indexes, elements, attributes, etc. may have a value. The value of the information, data, flags, indexes, elements, and attributes equal to "0" may represent a logical false or a first predefined value. In other words, the value "0", false, logical false, and the first predefined value may be replaced with each other. The value of the information, data, flags, indexes, elements, and attributes equal to "1" may represent a logical true or a second predefined value. In other words, the value "1", true, logical true, and the second predefined value may be replaced with each other.

[0068] When the variable i or j is used to represent a column, row, or index, the value of i may be an integer equal to or greater than 0, or an integer equal to or greater than 1. That is, the column, row, index, etc. may be counted from 0, or may be counted from 1.

[0069] Description of terms

[0070] Encoder: This refers to the device that performs encoding. In other words, it refers to the encoding device.

[0071] Decoder: Refers to a device that performs decoding. In other words, it refers to a decoding device.

[0072] Block: is an M×N sample array. Here, M and N may represent positive integers, and a block may represent a sample array in a two-dimensional form. A block may refer to a unit. A current block may represent an encoding target block that becomes a target at the time of encoding, or a decoding target block that becomes a target at the time of decoding. In addition, the current block may be at least one of a coding block, a prediction block, a residual block, and a transform block.

[0073] Sample: It is the basic unit of a block. d ), the sample point can be represented from 0 to 2 Bd In the present invention, a sample point may be used as the meaning of a pixel. That is, a sample point, a pel, and a pixel may have the same meaning as each other.

[0074] Unit: may refer to a coding and decoding unit. When encoding and decoding an image, a unit may be a region generated by partitioning a single image. In addition, when a single image is partitioned into sub-division units during encoding or decoding, a unit may represent a sub-division unit. That is, an image may be partitioned into a plurality of units. When encoding and decoding an image, a predetermined process for each unit may be performed. A single unit may be partitioned into sub-units having a size smaller than that of the unit. According to the function, a unit may represent a block, a macroblock, a coding tree unit, a coding tree block, a coding unit, a coding block, a prediction unit, a prediction block, a residual unit, a residual block, a transform unit, a transform block, and the like. In addition, in order to distinguish a unit from a block, a unit may include a luminance component block, a chrominance component block associated with the luminance component block, and a syntax element for each color component block. A unit may have various sizes and shapes, and specifically, the shape of a unit may be a two-dimensional geometric figure such as a square, a rectangle, a trapezoid, a triangle, a pentagon, and the like. Also, the unit information may include at least one of a unit type indicating a coding unit, a prediction unit, a transformation unit, etc., and a unit size, a unit depth, an order of encoding and decoding of the unit, and the like.

[0075] Coding tree unit: A single coding tree block configured with a luminance component Y and two coding tree blocks associated with chrominance components Cb and Cr. In addition, the coding tree unit may represent a syntax element including a block and each block. Each coding tree unit may be partitioned by using at least one of a quadtree partitioning method, a binary tree partitioning method, and a ternary tree partitioning method to configure a lower-level unit such as a coding unit, a prediction unit, a transform unit, etc. The coding tree unit may be used as a term for specifying a sample block that becomes a processing unit when encoding / decoding an image as an input image. Here, the quadtree may represent a quadtree.

[0076] When the size of the coding block is within a predetermined range, it can be divided using only quadtree partitioning. Here, the predetermined range may be defined as at least one of the maximum size and the minimum size of the coding block that can be divided using only quadtree partitioning. Information indicating the maximum / minimum size of the coding block that allows quadtree partitioning may be signaled through a bitstream, and the information may be signaled in at least one unit of a sequence, a picture parameter, a parallel block group, or a slice (fragment). Optionally, the maximum / minimum size of the coding block may be a fixed size predetermined in the encoder / decoder. For example, when the size of the coding block corresponds to 256×256 to 64×64, it is possible to divide using only quadtree partitioning. Optionally, when the size of the coding block is larger than the size of the maximum conversion block, it is possible to divide using only quadtree partitioning. Here, the block to be divided may be at least one of a coding block and a transform block. In this case, the information indicating the division of the coding block (e.g., split_flag) may be a flag indicating whether quadtree partitioning is performed. When the size of the coding block falls within a predetermined range, it is possible to divide using only binary or ternary tree partitioning. In this case, the above description of the quadtree partition can be applied to the binary tree partition or the ternary tree partition in the same manner.

[0077] Coding tree block: may be used as a term used to designate any one of a Y coding tree block, a Cb coding tree block, and a Cr coding tree block.

[0078] Neighboring block: may refer to a block adjacent to the current block. The block adjacent to the current block may refer to a block that touches the boundary of the current block or a block that is located within a predetermined distance from the current block. The neighboring block may refer to a block adjacent to a vertex of the current block. Here, the block adjacent to a vertex of the current block may refer to a block that is vertically adjacent to a neighboring block that is horizontally adjacent to the current block or a block that is horizontally adjacent to a neighboring block that is vertically adjacent to the current block.

[0079] Reconstructed neighboring block: may represent a neighboring block that is adjacent to the current block and has been encoded or decoded in space / time. Here, the reconstructed neighboring block may represent a reconstructed neighboring unit. The reconstructed spatial neighboring block may be a block that is within the current picture and has been reconstructed by encoding or decoding or both encoding and decoding. The reconstructed temporal neighboring block is a block at a position corresponding to the current block of the current picture within the reference image or a neighboring block of the block.

[0080] Unit depth: can represent the degree of partitioning of a unit. In a tree structure, the highest node (root node) can correspond to the first unit that is not partitioned. In addition, the highest node can have a minimum depth value. In this case, the depth of the highest node can be level 0. A node with a depth of level 1 can represent a unit generated by partitioning the first unit once. A node with a depth of level 2 can represent a unit generated by partitioning the first unit twice. A node with a depth of level n can represent a unit generated by partitioning the first unit n times. A leaf node can be the lowest node and a node that cannot be further partitioned. The depth of a leaf node can be the maximum level. For example, the predefined value of the maximum level can be 3. The depth of the root node can be the lowest, and the depth of the leaf node can be the deepest. In addition, when a unit is represented as a tree structure, the level at which the unit exists can represent the unit depth.

[0081] Bitstream: can represent a stream of bits including coded image information.

[0082] Parameter set: corresponds to header information among the configurations in the bitstream. At least one of a video parameter set, a sequence parameter set, a picture parameter set, and an adaptation parameter set may be included in the parameter set. In addition, the parameter set may include a slice header, a tile group header, and tile header information. The term "tile group" means a group of tiles and has the same meaning as a slice.

[0083] The adaptation parameter set may represent a parameter set that can be shared by being referenced in different pictures, sub-pictures, slices, tile groups, tiles, or partitions. In addition, information in the adaptation parameter set may be used by referring to different adaptation parameter sets for sub-pictures, slices, tile groups, tiles, or partitions within a picture.

[0084] Furthermore, regarding adaptation parameter sets, different adaptation parameter sets may be referenced by using identifiers of different adaptation parameter sets for a sub-picture, a slice, a tile group, a tile, or a partition within a picture.

[0085] Furthermore, regarding adaptation parameter sets, different adaptation parameter sets may be referenced by using identifiers of different adaptation parameter sets for slices, tile groups, tiles, or partitions within a sub-picture.

[0086] Furthermore, regarding adaptation parameter sets, different adaptation parameter sets may be referenced by using identifiers of different adaptation parameter sets for tiles or partitions within a slice.

[0087] Furthermore, with respect to adaptation parameter sets, different adaptation parameter sets may be referenced by using identifiers of different adaptation parameter sets for partitions within a tile.

[0088] Information about the adaptation parameter set identifier may be included in a parameter set or a header of a sub-picture, and an adaptation parameter set corresponding to the adaptation parameter set identifier may be used for the sub-picture.

[0089] Information about the adaptation parameter set identifier may be included in a parameter set or a header of a tile, and an adaptation parameter set corresponding to the adaptation parameter set identifier may be used for the tile.

[0090] Information about the adaptation parameter set identifier may be included in a header of the tile, and an adaptation parameter set corresponding to the adaptation parameter set identifier may be used for the tile.

[0091] A picture may be partitioned into one or more tile rows and one or more tile columns.

[0092] A sub-picture may be partitioned into one or more parallel block rows and one or more parallel block columns within a picture. A sub-picture may be an area having a rectangular / square form within a picture and may include one or more CTUs. In addition, at least one or more parallel blocks / blocks / strips may be included in a sub-picture.

[0093] A tile may be an area having a rectangular / square form within a picture and may include one or more CTUs. In addition, a tile may be partitioned into one or more partitions.

[0094] A partition may represent one or more CTU rows within a tile. A tile may be partitioned into one or more partitions, and each partition may have at least one or more CTU rows. A tile that is not partitioned into two or more may represent a partition.

[0095] A slice may include one or more tiles within a picture, and may include one or more partitions within a tile.

[0096] Parsing: may mean determining the value of a syntax element by performing entropy decoding, or may mean the entropy decoding itself.

[0097] Symbol: can represent at least one of a syntax element, a coding parameter, and a transform coefficient value of a coding / decoding target unit. In addition, the symbol can represent an entropy coding target or an entropy decoding result.

[0098] Prediction mode: may be information indicating a mode for encoding / decoding using intra prediction or a mode for encoding / decoding using inter prediction.

[0099] Prediction unit: may represent a basic unit when performing prediction (such as inter prediction, intra prediction, inter compensation, intra compensation, and motion compensation). A single prediction unit may be partitioned into multiple partitions of smaller size, or may be partitioned into multiple prediction units of lower levels. Multiple partitions may be basic units when performing prediction or compensation. Partitions generated by splitting a prediction unit may also be prediction units.

[0100] Prediction unit partition: may represent a shape obtained by partitioning a prediction unit.

[0101] A reference picture list may refer to a list including one or more reference pictures used for inter prediction or motion compensation. There are several types of available reference picture lists, including LC (List Combination), L0 (List 0), L1 (List 1), L2 (List 2), L3 (List 3).

[0102] The inter prediction indicator may refer to the direction of inter prediction of the current block (unidirectional prediction, bidirectional prediction, etc.). Alternatively, the inter prediction indicator may refer to the number of reference pictures used to generate the prediction block of the current block. Alternatively, the inter prediction indicator may refer to the number of prediction blocks used when performing inter prediction or motion compensation on the current block.

[0103] The prediction list utilization flag indicates whether at least one reference picture in a specific reference picture list is used to generate a prediction block. The prediction list utilization flag may be used to derive the inter prediction indicator, and conversely, the inter prediction indicator may be used to derive the prediction list utilization flag. For example, when the prediction list utilization flag has a first value of zero (0), it indicates that the reference picture in the reference picture list is not used to generate the prediction block. On the other hand, when the prediction list utilization flag has a second value of one (1), it indicates that the reference picture list is used to generate the prediction block.

[0104] The reference picture index may refer to an index indicating a specific reference picture in a reference picture list.

[0105] A reference picture may refer to a reference picture referenced by a specific block for the purpose of inter-frame prediction or motion compensation of the specific block. Alternatively, a reference picture may be a picture including a reference block referenced by a current block for inter-frame prediction or motion compensation. Hereinafter, the terms "reference picture" and "reference image" have the same meaning and are interchangeable.

[0106] A motion vector may be a two-dimensional vector used for inter-frame prediction or motion compensation. A motion vector may represent an offset between a coding / decoding target block and a reference block. For example, (mvX, mvY) may represent a motion vector. Here, mvX may represent a horizontal component, and mvY may represent a vertical component.

[0107] The search range may be a two-dimensional area that is searched during inter prediction to retrieve a motion vector. For example, the size of the search range may be M×N. Here, M and N are both integers.

[0108] The motion vector candidate may refer to a prediction candidate block or a motion vector of the prediction candidate block when predicting a motion vector. In addition, the motion vector candidate may be included in a motion vector candidate list.

[0109] The motion vector candidate list may mean a list consisting of one or more motion vector candidates.

[0110] The motion vector candidate index may represent an indicator indicating a motion vector candidate in the motion vector candidate list. Alternatively, it may be an index of a motion vector predictor.

[0111] The motion information may represent information including at least one of items including a motion vector, a reference picture index, an inter prediction indicator, a prediction list utilization flag, a reference picture list information, a reference picture, a motion vector candidate, a motion vector candidate index, a merge candidate, and a merge index.

[0112] The merge candidate list may mean a list consisting of one or more merge candidates.

[0113] The merge candidate may represent a spatial merge candidate, a temporal merge candidate, a combined merge candidate, a combined bi-prediction merge candidate, or a zero merge candidate. The merge candidate may include motion information such as an inter prediction indicator, a reference picture index for each list, a motion vector, a prediction list utilization flag, and an inter prediction indicator.

[0114] The merge index may represent an indicator indicating a merge candidate in the merge candidate list. Alternatively, the merge index may indicate a block in a reconstructed block that is spatially / temporally adjacent to the current block, from which the merge candidate has been derived. Alternatively, the merge index may indicate at least one piece of motion information of the merge candidate.

[0115] Transform unit: may represent a basic unit when encoding / decoding (such as transform, inverse transform, quantization, inverse quantization, transform coefficient encoding / decoding) is performed on a residual signal. A single transform unit may be partitioned into a plurality of lower-level transform units having a smaller size. Here, the transform / inverse transform may include at least one of a first transform / first inverse transform and a second transform / second inverse transform.

[0116] Scaling: may refer to the process of multiplying the level of quantization by a factor. Transform coefficients may be generated by scaling the level of quantization. Scaling may also be referred to as inverse quantization.

[0117] Quantization parameter: may indicate a value used when a transform coefficient is used to generate a quantized level during quantization. The quantization parameter may also indicate a value used when a transform coefficient is generated by scaling the quantized level during inverse quantization. The quantization parameter may be a value mapped to a quantization step size.

[0118] Delta quantization parameter: may represent the difference between the predicted quantization parameter and the quantization parameter of the encoding / decoding target unit.

[0119] Scan: may refer to a method of ordering coefficients within a cell, block, or matrix. For example, changing a two-dimensional matrix of coefficients to a one-dimensional matrix may be called scanning, and changing a one-dimensional matrix of coefficients to a two-dimensional matrix may be called scanning or inverse scanning.

[0120] Transform coefficient: may refer to a coefficient value generated after performing a transform in an encoder. Transform coefficient may refer to a coefficient value generated after performing at least one of entropy decoding and inverse quantization in a decoder. A quantization level obtained by quantizing a transform coefficient or a residual signal or a quantized transform coefficient level may also fall within the meaning of a transform coefficient.

[0121] Quantization level: may represent a value generated by quantizing a transform coefficient or a residual signal in an encoder. Alternatively, the quantization level may represent a value that is a dequantization target subjected to dequantization in a decoder. Similarly, the quantized transform coefficient level as a result of transformation and quantization may also fall within the meaning of the quantization level.

[0122] Non-zero transform coefficient: may refer to a transform coefficient having a value other than zero, or a transform coefficient level or quantization level having a value other than zero.

[0123] Quantization matrix: may refer to a matrix used in a quantization process or an inverse quantization process performed to improve subjective image quality or objective image quality. The quantization matrix may also be referred to as a scaling list.

[0124] Quantization matrix coefficients: can represent each element in the quantization matrix. Quantization matrix coefficients can also be called matrix coefficients.

[0125] Default matrix: may represent a predetermined quantization matrix predefined in an encoder or a decoder.

[0126] Non-default matrix: may denote a quantization matrix that is not predefined in the encoder or decoder but is signaled by the user.

[0127] Statistical value: The statistical value for at least one of a variable, a coding parameter, a constant value, etc. having a calculable specific value can be one or more of the average value, summed value, weighted average value, weighted sum value, minimum value, maximum value, most frequently occurring value, median value, and interpolation value of the corresponding specific value.

[0128] Figure 1 is a block diagram showing a configuration of an encoding device according to an embodiment to which the present invention is applied.

[0129] The encoding device 100 may be an encoder, a video encoding device, or an image encoding device. A video may include at least one image. The encoding device 100 may sequentially encode at least one image.

[0130] Reference Figure 1 , the encoding device 100 may include a motion prediction unit 111, a motion compensation unit 112, an intra-frame prediction unit 120, a switch 115, a subtractor 125, a transform unit 130, a quantization unit 140, an entropy encoding unit 150, an inverse quantization unit 160, an inverse transform unit 170, an adder 175, a filter unit 180 and a reference picture buffer 190.

[0131] The encoding device 100 may perform encoding of an input image by using an intra mode or an inter mode or both an intra mode and an inter mode. In addition, the encoding device 100 may generate a bit stream including encoding information by encoding the input image, and output the generated bit stream. The generated bit stream may be stored in a computer-readable recording medium, or may be streamed through a wired / wireless transmission medium. When the intra mode is used as a prediction mode, the switch 115 may switch to the intra mode. Alternatively, when the inter mode is used as a prediction mode, the switch 115 may switch to the inter mode. Here, the intra mode may represent an intra prediction mode, and the inter mode may represent an inter prediction mode. The encoding device 100 may generate a prediction block for an input block of the input image. In addition, the encoding device 100 may encode the residual block using the residual of the input block and the prediction block after generating the prediction block. The input image may be referred to as a current image as a current encoding target. The input block may be referred to as a current block as a current encoding target, or may be referred to as an encoding target block.

[0132] When the prediction mode is the intra mode, the intra prediction unit 120 may use samples of a block that has been encoded / decoded and is adjacent to the current block as reference samples. The intra prediction unit 120 may perform spatial prediction on the current block by using the reference samples, or may generate prediction samples of the input block by performing spatial prediction. Here, intra prediction may refer to prediction within a frame.

[0133] When the prediction mode is the inter mode, the motion prediction unit 111 may retrieve the area that best matches the input block from the reference image when performing motion prediction, and derive a motion vector by using the retrieved area. In this case, the search area may be used as the area. The reference image may be stored in the reference picture buffer 190. Here, when encoding / decoding the reference image is performed, the reference image may be stored in the reference picture buffer 190.

[0134] The motion compensation unit 112 may generate a predicted block by performing motion compensation on the current block using a motion vector. Here, inter prediction may refer to prediction or motion compensation between frames.

[0135] When the value of the motion vector is not an integer, the motion prediction unit 111 and the motion compensation unit 112 may generate a prediction block by applying an interpolation filter to a partial area of ​​a reference picture. In order to perform inter-picture prediction or motion compensation on a coding unit, it may be determined which mode among the skip mode, merge mode, advanced motion vector prediction (AMVP) mode, and current picture reference mode is used for motion prediction and motion compensation of a prediction unit included in the corresponding coding unit. Then, inter-picture prediction or motion compensation may be performed differently according to the determined mode.

[0136] The subtractor 125 may generate a residual block by using the difference between the input block and the prediction block. The residual block may be referred to as a residual signal. The residual signal may represent the difference between the original signal and the prediction signal. In addition, the residual signal may be a signal generated by transforming or quantizing the difference between the original signal and the prediction signal, or by transforming and quantizing the difference between the original signal and the prediction signal. The residual block may be a residual signal of a block unit.

[0137] The transform unit 130 may generate a transform coefficient by performing a transform on the residual block and output the generated transform coefficient. Here, the transform coefficient may be a coefficient value generated by performing a transform on the residual block. When the transform skip mode is applied, the transform unit 130 may skip the transform on the residual block.

[0138] The quantized level may be generated by applying quantization to a transform coefficient or to a residual signal. Hereinafter, the quantized level may also be referred to as a transform coefficient in embodiments.

[0139] The quantization unit 140 may generate a quantization level by quantizing the transform coefficient or the residual signal according to the parameter, and output the generated quantization level. Here, the quantization unit 140 may quantize the transform coefficient by using a quantization matrix.

[0140] The entropy encoding unit 150 may generate a bitstream by performing entropy encoding on the value calculated by the quantization unit 140 or the encoding parameter value calculated when encoding is performed according to the probability distribution, and output the generated bitstream. The entropy encoding unit 150 may perform entropy encoding on sample information of the image and information for decoding the image. For example, the information for decoding the image may include a syntax element.

[0141] When entropy coding is applied, symbols are represented so that a smaller number of bits are assigned to symbols with a high probability of generation, and a larger number of bits are assigned to symbols with a low probability of generation, and therefore, the size of the bit stream for the symbol to be encoded can be reduced. The entropy coding unit 150 may use a coding method for entropy coding such as exponential Golomb, context adaptive variable length coding (CAVLC), context adaptive binary arithmetic coding (CABAC), etc. For example, the entropy coding unit 150 may perform entropy coding by using a variable length coding / code (VLC) table. In addition, the entropy coding unit 150 may derive a binarization method of a target symbol and a probability model of a target symbol / binary bit, and perform arithmetic coding by using the derived binarization method and context model.

[0142] In order to encode a transform coefficient level (quantized level), the entropy encoding unit 150 may change a coefficient in a two-dimensional block form into a one-dimensional vector form by using a transform coefficient scanning method.

[0143] The coding parameters may include information such as syntax elements (flags, indexes, etc.) that are encoded in the encoder and sent to the decoder by signaling, as well as information derived when performing encoding or decoding. The coding parameters may represent information required when encoding or decoding an image. For example, at least one value or combination of the following items may be included in the coding parameters: unit / block size, unit / block depth, unit / block partition information, unit / block shape, unit / block partition structure, whether to perform quadtree partitioning, whether to perform binary tree partitioning, binary tree partition direction (horizontal or vertical), binary tree partition form (symmetric partitioning or asymmetric partitioning), whether the current coding unit is partitioned by ternary tree partitioning, the direction of ternary tree partitioning (horizontal or vertical), the type of ternary tree partitioning (symmetric type or asymmetric type), whether the current coding unit is partitioned by multi-type tree partitioning, and the type of multi-type tree partitioning. direction (horizontal or vertical), type of multi-type tree partition (symmetric or asymmetric), tree (binary tree or ternary tree) structure of multi-type tree partition, prediction mode (intra-frame prediction or inter-frame prediction), luminance intra-frame prediction mode / direction, chrominance intra-frame prediction mode / direction, intra-frame partition information, inter-frame partition information, coding block partition flag, prediction block partition flag, transform block partition flag, reference sample filtering method, reference sample filter taps, reference sample filter coefficients, prediction block filtering method, prediction block filter taps, prediction block filter coefficients, prediction block boundary filtering method, prediction block boundary filter taps, prediction block boundary filter coefficients, intra-frame prediction mode , inter prediction mode, motion information, motion vector, motion vector difference, reference picture index, inter prediction angle, inter prediction indicator, prediction list utilization flag, reference picture list, reference picture, motion vector predictor index, motion vector predictor candidate, motion vector candidate list, whether to use merge mode, merge index, merge candidate, merge candidate list, whether to use skip mode, interpolation filter type, interpolation filter tap, interpolation filter coefficient, motion vector size, representation accuracy of motion vector, transform type, transform size, information on whether primary (first) transform is used, information on whether secondary transform is used, primary transform index, secondary transform index , information on whether a residual signal exists, coding block pattern, coding block flag (CBF), quantization parameter, quantization parameter residual, quantization matrix, whether to apply an intra-frame loop filter, intra-frame loop filter coefficients, intra-frame loop filter taps, intra-frame loop filter shape / form, whether to apply a deblocking filter, deblocking filter coefficients, deblocking filter taps, deblocking filter strength, deblocking filter shape / form, whether to apply adaptive sample offset, adaptive sample offset value, adaptive sample offset category, adaptive sample offset type, whether to apply an adaptive loop filter, adaptive loop filter coefficients, adaptive loop filter taps, adaptive loop filter shape / form,Binarization / debinarization method, context model determination method, context model update method, whether to execute normal mode, whether to execute bypass mode, context binary bit, bypass binary bit, valid coefficient flag, last valid coefficient flag, encoding flag for unit of coefficient group, position of last valid coefficient, flag on whether the value of coefficient is greater than 1, flag on whether the value of coefficient is greater than 2, flag on whether the value of coefficient is greater than 3, information on remaining coefficient values, sign information, reconstructed luminance sample, reconstructed chrominance sample, residual luminance sample, residual chrominance sample, luminance transform coefficient, chrominance transform coefficient, quantized luminance level, quantized chrominance level, transform coefficient level scanning method, motion vector search area at decoder side domain size, shape of a motion vector search area at a decoder side, number of motion vector searches at a decoder side, information on a CTU size, information on a minimum block size, information on a maximum block size, information on a maximum block depth, information on a minimum block depth, image display / output order, slice identification information, slice type, slice partition information, tile identification information, tile type, tile partition information, tile group identification information, tile group type, tile group partition information, picture type, bit depth of input samples, bit depth of reconstructed samples, bit depth of residual samples, bit depth of transform coefficients, bit depth of quantization levels, and information on a luminance signal or information on a chrominance signal.

[0144] Here, signaling a flag or an index may mean entropy encoding the corresponding flag or index by an encoder and including it in a bitstream, and may mean entropy decoding the corresponding flag or index from the bitstream by a decoder.

[0145] When the encoding apparatus 100 performs encoding by inter-frame prediction, the encoded current image may be used as a reference image for another image that is subsequently processed. Therefore, the encoding apparatus 100 may reconstruct or decode the encoded current image, or store the reconstructed or decoded image as a reference image in the reference picture buffer 190.

[0146] The quantized level may be dequantized in the dequantization unit 160 or may be inversely transformed in the inverse transform unit 170. The dequantized or inversely transformed coefficient or the dequantized and inversely transformed coefficient may be added to the prediction block by the adder 175. By adding the dequantized or inversely transformed coefficient or the dequantized and inversely transformed coefficient to the prediction block, a reconstructed block may be generated. Here, the dequantized or inversely transformed coefficient or the dequantized and inversely transformed coefficient may mean a coefficient on which at least one of dequantization and inverse transformation is performed, and may mean a reconstructed residual block.

[0147] The reconstructed block may pass through the filter unit 180. The filter unit 180 may apply at least one of a deblocking filter, a sample adaptive offset (SAO), and an adaptive loop filter (ALF) to the reconstructed sample, the reconstructed block, or the reconstructed image. The filter unit 180 may be referred to as an in-loop filter.

[0148] The deblocking filter may remove block distortion generated in the boundary between blocks. In order to determine whether to apply the deblocking filter, it may be determined whether to apply the deblocking filter to the current block based on the samples included in the number of rows or columns included in the block. When the deblocking filter is applied to the block, another filter may be applied according to the required deblocking filter strength.

[0149] In order to compensate for the coding error, a suitable offset value may be added to the sample value by using sample adaptive offset. Sample adaptive offset can correct the offset of the deblocked image from the original image in units of samples. A method of applying the offset in consideration of edge information about each sample may be used, or a method of partitioning the samples of the image into a predetermined number of regions, determining the region to which the offset is applied, and applying the offset to the determined region may be used.

[0150] The adaptive loop filter may perform filtering based on a comparison result of a filtered reconstructed image and an original image. Samples included in an image may be partitioned into predetermined groups, a filter to be applied to each group may be determined, and differential filtering may be performed on each group. Information on whether ALF is applied may be signaled by a coding unit (CU), and the form and coefficient of ALF to be applied to each block may vary.

[0151] The reconstructed block or the reconstructed image that has passed through the filter unit 180 may be stored in the reference picture buffer 190. The reconstructed block processed by the filter unit 180 may be a part of the reference image. That is, the reference image is a reconstructed image composed of the reconstructed blocks processed by the filter unit 180. The stored reference image may be used later in inter-frame prediction or motion compensation.

[0152] Figure 2 is a block diagram showing the configuration of a decoding device according to an embodiment and to which the present invention is applied.

[0153] The decoding device 200 may be a decoder, a video decoding device, or an image decoding device.

[0154] Reference Figure 2 , the decoding device 200 may include an entropy decoding unit 210, a dequantization unit 220, an inverse transform unit 230, an intra-frame prediction unit 240, a motion compensation unit 250, an adder 255, a filter unit 260 and a reference picture buffer 270.

[0155] The decoding device 200 may receive a bit stream output from the encoding device 100. The decoding device 200 may receive a bit stream stored in a computer-readable recording medium, or may receive a bit stream streamed through a wired / wireless transmission medium. The decoding device 200 may decode the bit stream by using an intra mode or an inter mode. In addition, the decoding device 200 may generate a reconstructed image or a decoded image generated by decoding, and output the reconstructed image or the decoded image.

[0156] When the prediction mode used in decoding is the intra mode, the switch may be switched to the intra mode. Alternatively, when the prediction mode used in decoding is the inter mode, the switch may be switched to the inter mode.

[0157] The decoding device 200 can obtain a reconstructed residual block by decoding the input bit stream and generate a prediction block. When the reconstructed residual block and the prediction block are obtained, the decoding device 200 can generate a reconstructed block that becomes a decoding target by adding the reconstructed residual block to the prediction block. The decoding target block may be referred to as a current block.

[0158] The entropy decoding unit 210 may generate symbols by entropy decoding the bit stream according to the probability distribution. The generated symbols may include symbols in the form of quantized levels. Here, the entropy decoding method may be an inverse process of the above entropy encoding method.

[0159] In order to decode the transform coefficient levels (quantized levels), the entropy decoding unit 210 may change the coefficients in the form of a one-way vector into a two-dimensional block form by using a transform coefficient scanning method.

[0160] The quantized level may be dequantized in the dequantization unit 220, or the quantized level may be inversely transformed in the inverse transform unit 230. The quantized level may be a result of dequantization or inverse transformation or both, and may be generated as a reconstructed residual block. Here, the dequantization unit 220 may apply a quantization matrix to the quantized level.

[0161] When the intra mode is used, the intra prediction unit 240 may generate a predicted block by performing spatial prediction on the current block, wherein the spatial prediction uses sample values ​​of a block that is adjacent to the decoding target block and has been decoded.

[0162] When the inter mode is used, the motion compensation unit 250 may generate a prediction block by performing motion compensation on the current block, wherein the motion compensation uses a motion vector and a reference image stored in the reference picture buffer 270 .

[0163] The adder 255 can generate a reconstructed block by adding the reconstructed residual block to the prediction block. The filter unit 260 can apply at least one of a deblocking filter, a sample adaptive offset, and an adaptive loop filter to the reconstructed block or the reconstructed image. The filter unit 260 can output the reconstructed image. The reconstructed block or the reconstructed image can be stored in the reference picture buffer 270 and used when performing inter-frame prediction. The reconstructed block processed by the filter unit 260 can be a part of the reference image. That is, the reference image is a reconstructed image composed of the reconstructed blocks processed by the filter unit 260. The stored reference image can be used later in inter-frame prediction or motion compensation.

[0164] Figure 3 is a diagram schematically showing a partition structure of an image when the image is encoded and decoded. Figure 3 An example of partitioning a single cell into a plurality of lower-level cells is schematically shown.

[0165] In order to effectively partition an image, a coding unit (CU) may be used when encoding and decoding. A coding unit may be used as a basic unit when encoding / decoding an image. In addition, a coding unit may be used as a unit for distinguishing an intra prediction mode from an inter prediction mode when encoding / decoding an image. A coding unit may be a basic unit for prediction, transformation, quantization, inverse transformation, inverse quantization, or encoding / decoding processing of a transform coefficient.

[0166] Reference Figure 3 , the image 300 is partitioned sequentially according to the maximum coding unit (LCU), and the LCU unit is determined as a partition structure. Here, the LCU may be used in the same meaning as the coding tree unit (CTU). Unit partitioning may mean partitioning a block associated with the unit. In the block partition information, information about the unit depth may be included. The depth information may indicate the number or degree of the unit being partitioned or both the number and degree of the unit being partitioned. A single unit may be partitioned into a plurality of lower-level units hierarchically associated with the depth information based on a tree structure. In other words, the unit and the lower-level units generated by partitioning the unit may correspond to a node and a child node of the node, respectively. Each of the partitioned lower-level units may have depth information. The depth information may be information indicating the size of a CU and may be stored in each CU. The unit depth indicates the number and / or degree associated with partitioning the unit. Therefore, the partition information of the lower-level unit may include information about the size of the lower-level unit.

[0167] The partition structure may represent the distribution of coding units (CUs) within the LCU 310. Such distribution may be determined according to whether a single CU is partitioned into multiple (including 2, 4, 8, 16, etc., positive integers equal to or greater than 2) CUs. The horizontal size and vertical size of the CU generated by partitioning may be half of the horizontal size and vertical size of the CU before partitioning, respectively, or may have sizes smaller than the horizontal size and vertical size before partitioning, respectively, according to the number of partitioning. The CU may be recursively partitioned into multiple CUs. By recursive partitioning, at least one of the height and width of the CU after partitioning may be reduced compared to at least one of the height and width of the CU before partitioning. The partitioning of the CU may be recursively performed until a predefined depth or a predefined size. For example, the depth of the LCU may be 0, and the depth of the minimum coding unit (SCU) may be a predefined maximum depth. Here, as described above, the LCU may be a coding unit having a maximum coding unit size, and the SCU may be a coding unit having a minimum coding unit size. Partitioning starts from the LCU 310, and when the horizontal size or the vertical size or both the horizontal size and the vertical size of the CU are reduced by partitioning, the CU depth increases by 1. For example, for each depth, the size of the non-partitioned CU may be 2N×2N. In addition, in the case of a partitioned CU, a CU of size 2N×2N may be partitioned into four CUs of size N×N. As the depth increases by 1, the size of N may be halved.

[0168] In addition, information on whether a CU is partitioned may be indicated by using the partition information of the CU. The partition information may be 1-bit information. All CUs except the SCU may include partition information. For example, when the value of the partition information is a first value, the CU may not be partitioned, and when the value of the partition information is a second value, the CU may be partitioned.

[0169] Reference Figure 3 , an LCU with a depth of 0 may be a block of 64×64. 0 may be the minimum depth. An SCU with a depth of 3 may be a block of 8×8. 3 may be the maximum depth. A CU of a block of 32×32 and a CU of a block of 16×16 may be represented as depth 1 and depth 2, respectively.

[0170] For example, when a single coding unit is partitioned into four coding units, the horizontal size and vertical size of the four coding units partitioned may be half the size of the horizontal size and vertical size of the CU before being partitioned. In one embodiment, when a coding unit of size 32×32 is partitioned into four coding units, each of the four coding units partitioned may have a size of 16×16. When a single coding unit is partitioned into four coding units, it can be said that the coding unit can be partitioned into a quadtree form.

[0171] For example, when one coding unit is partitioned into two sub-coding units, the horizontal size or vertical size (width or height) of each of the two sub-coding units may be half of the horizontal size or vertical size of the original coding unit. For example, when a coding unit of size 32×32 is partitioned vertically into two sub-coding units, each of the two sub-coding units may have a size of 16×32. For example, when a coding unit of size 8×32 is partitioned horizontally into two sub-coding units, each of the two sub-coding units may have a size of 8×16. When one coding unit is partitioned into two sub-coding units, the coding unit may be said to be partitioned into two or partitioned according to a binary tree partition structure.

[0172] For example, when one coding unit is partitioned into three sub-coding units, the horizontal size or the vertical size of the coding unit may be partitioned at a ratio of 1:2:1, thereby generating three sub-coding units having a ratio of 1:2:1 in the horizontal size or the vertical size. For example, when a coding unit having a size of 16×32 is partitioned horizontally into three sub-coding units, the three sub-coding units may have sizes of 16×8, 16×16, and 16×8, respectively, in order from the uppermost sub-coding unit to the lowermost sub-coding unit. For example, when a coding unit having a size of 32×32 is partitioned vertically into three sub-coding units, the three sub-coding units may have sizes of 8×32, 16×32, and 8×32, respectively, in order from the left sub-coding unit to the right sub-coding unit. When one coding unit is partitioned into three sub-coding units, the coding unit may be said to be partitioned into three sub-coding units or partitioned according to a ternary tree partition structure.

[0173] exist Figure 3 , a coding tree unit (CTU) 320 is an example of a CTU to which a quadtree partition structure, a binary tree partition structure, and a ternary tree partition structure are all applied.

[0174] As described above, in order to partition a CTU, at least one of a quadtree partition structure, a binary tree partition structure, and a ternary tree partition structure may be applied. Various tree partition structures may be sequentially applied to a CTU according to a predetermined priority order. For example, a quadtree partition structure may be preferentially applied to a CTU. Coding units that can no longer be partitioned using a quadtree partition structure may correspond to leaf nodes of a quadtree. Coding units corresponding to leaf nodes of a quadtree may be used as root nodes of a binary and / or ternary tree partition structure. That is, coding units corresponding to leaf nodes of a quadtree may be further partitioned according to a binary tree partition structure or a ternary tree partition structure, or may not be further partitioned. Therefore, by preventing the coding blocks obtained from binary tree partitions or ternary tree partitions of coding units corresponding to leaf nodes of a quadtree from undergoing further quadtree partitions, block partitioning operations and / or operations of signaling partition information may be effectively performed.

[0175] The fact that the coding unit corresponding to the node of the quadtree is partitioned may be signaled using the four partition information. The four partition information having a first value (e.g., '1') may indicate that the current coding unit is partitioned according to the quadtree partition structure. The four partition information having a second value (e.g., '0') may indicate that the current coding unit is not partitioned according to the quadtree partition structure. The four partition information may be a flag having a predetermined length (e.g., one bit).

[0176] There may be no priority between binary tree partitioning and ternary tree partitioning. That is, the coding unit corresponding to the leaf node of the quadtree may further undergo any partitioning of the binary tree partitioning and the ternary tree partitioning. In addition, the coding unit generated by the binary tree partitioning or the ternary tree partitioning may undergo further binary tree partitioning or further ternary tree partitioning, or may not be further partitioned.

[0177] A tree structure in which there is no priority between binary tree partitions and ternary tree partitions is called a multi-type tree structure. A coding unit corresponding to a leaf node of a quadtree may be used as a root node of a multi-type tree. At least one of multi-type tree partition indication information, partition direction information, and partition tree information may be used to signal whether to partition a coding unit corresponding to a node of a multi-type tree. In order to partition a coding unit corresponding to a node of a multi-type tree, multi-type tree partition indication information, partition direction information, and partition tree information may be sequentially signaled.

[0178] The multi-type tree partition indication information having a first value (eg, '1') may indicate that the current coding unit will undergo multi-type tree partitioning. The multi-type tree partition indication information having a second value (eg, '0') may indicate that the current coding unit will not undergo multi-type tree partitioning.

[0179] When a coding unit corresponding to a node of a multi-type tree is further partitioned according to a multi-type tree partition structure, the coding unit may include partition direction information. The partition direction information may indicate in which direction the current coding unit will be partitioned for the multi-type tree partition. The partition direction information having a first value (e.g., "1") may indicate that the current coding unit will be partitioned vertically. The partition direction information having a second value (e.g., "0") may indicate that the current coding unit will be partitioned horizontally.

[0180] When the coding unit corresponding to the node of the multi-type tree is further partitioned according to the multi-type tree partition structure, the current coding unit may include partition tree information. The partition tree information may indicate a tree partition structure to be used to partition the node of the multi-type tree. The partition tree information having a first value (e.g., "1") may indicate that the current coding unit will be partitioned according to a binary tree partition structure. The partition tree information having a second value (e.g., "0") may indicate that the current coding unit will be partitioned according to a ternary tree partition structure.

[0181] The partition indication information, the partition tree information and the partition direction information may all be flags having a predetermined length (eg, one bit).

[0182] At least any one of the quadtree partition indication information, the multi-type tree partition indication information, the partition direction information, and the partition tree information may be entropy encoded / decoded. In order to entropy encode / decode those types of information, information about neighboring coding units adjacent to the current coding unit may be used. For example, there is a high probability that the partition type (partitioned or not partitioned, partition tree, and / or partition direction) of the left neighboring coding unit and / or the upper neighboring coding unit of the current coding unit is similar to the partition type of the current coding unit. Therefore, context information for entropy encoding / decoding the information about the current coding unit may be derived from the information about the neighboring coding units. The information about the neighboring coding units may include at least any one of the quadtree partition information, the multi-type tree partition indication information, the partition direction information, and the partition tree information.

[0183] As another example, among binary tree partitioning and ternary tree partitioning, binary tree partitioning may be preferentially performed. That is, the current coding unit may first undergo binary tree partitioning, and then the coding unit corresponding to the leaf node of the binary tree may be set as the root node for the ternary tree partitioning. In this case, for the coding unit corresponding to the node of the ternary tree, neither quadtree partitioning nor binary tree partitioning may be performed.

[0184] A coding unit that cannot be partitioned according to a quadtree partition structure, a binary tree partition structure, and / or a ternary tree partition structure becomes a basic unit for encoding, prediction, and / or transformation. That is, the coding unit cannot be further partitioned for prediction and / or transformation. Therefore, partition structure information and partition information for partitioning a coding unit into a prediction unit and / or a transformation unit may not exist in the bitstream.

[0185] However, when the size of the coding unit (i.e., the basic unit for partitioning) is larger than the size of the maximum transform block, the coding unit may be recursively partitioned until the size of the coding unit is reduced to be equal to or smaller than the size of the maximum transform block. For example, when the size of the coding unit is 64×64 and when the size of the maximum transform block is 32×32, the coding unit may be partitioned into four 32×32 blocks for transforming. For example, when the size of the coding unit is 32×64 and the size of the maximum transform block is 32×32, the coding unit may be partitioned into two 32×32 blocks for transforming. In this case, the partitioning of the coding unit for transforming is not separately signaled, and the partitioning of the coding unit for transforming may be determined by comparison between the horizontal size or vertical size of the coding unit and the horizontal size or vertical size of the maximum transform block. For example, when the horizontal size (width) of the coding unit is larger than the horizontal size (width) of the maximum transform block, the coding unit may be vertically divided into two equal parts. For example, when the vertical size (height) of the coding unit is larger than the vertical size (height) of the maximum transform block, the coding unit may be horizontally divided into two equal parts.

[0186] Information on the maximum and / or minimum size of a coding unit and information on the maximum and / or minimum size of a transform block may be signaled or determined at a higher level of the coding unit. The higher level may be, for example, a sequence level, a picture level, a slice level, a tile group level, a tile block level, etc. For example, the minimum size of a coding unit may be determined as 4×4. For example, the maximum size of a transform block may be determined as 64×64. For example, the minimum size of a transform block may be determined as 4×4.

[0187] Information on the minimum size of the coding unit corresponding to the leaf node of the quadtree (quadtree minimum size) and / or information on the maximum depth from the root node of the multi-type tree to the leaf node (maximum tree depth of the multi-type tree) may be signaled or determined at a higher level of the coding unit. For example, the higher level may be a sequence level, a picture level, a slice level, a tile group level, a tile level, etc. Information on the minimum size of the quadtree and / or information on the maximum depth of the multi-type tree may be signaled or determined for each of the intra-picture slice and the inter-picture slice.

[0188] The difference information between the size of the CTU and the maximum size of the transform block may be signaled or determined at a higher level of the coding unit. For example, the higher level may be a sequence level, a picture level, a slice level, a parallel block group level, a parallel block level, etc. The information of the maximum size of the coding unit corresponding to each node of the binary tree (hereinafter referred to as the maximum size of the binary tree) may be determined based on the size of the coding tree unit and the difference information. The maximum size of the coding unit corresponding to each node of the ternary tree (hereinafter referred to as the maximum size of the ternary tree) may vary according to the type of the slice. For example, for an intra-picture slice, the maximum size of the ternary tree may be 32×32. For example, for an inter-picture slice, the maximum size of the ternary tree may be 128×128. For example, the minimum size of the coding unit corresponding to each node of the binary tree (hereinafter referred to as the minimum size of the binary tree) and / or the minimum size of the coding unit corresponding to each node of the ternary tree (hereinafter referred to as the minimum size of the ternary tree) may be set to the minimum size of the coding block.

[0189] As another example, the maximum size of the binary tree and / or the maximum size of the ternary tree may be signaled or determined at the slice level. Alternatively, the minimum size of the binary tree and / or the minimum size of the ternary tree may be signaled or determined at the slice level.

[0190] According to the sizes and depth information of the above-mentioned various blocks, quad partition information, multi-type tree partition indication information, partition tree information and / or partition direction information may or may not be included in the bitstream.

[0191] For example, when the size of the coding unit is not greater than the minimum size of the quadtree, the coding unit does not include the quad partition information. Therefore, the quad partition information may be inferred from the second value.

[0192] For example, when the size (horizontal size and vertical size) of the coding unit corresponding to the node of the multi-type tree is greater than the maximum size (horizontal size and vertical size) of the binary tree and / or the maximum size (horizontal size and vertical size) of the ternary tree, the coding unit may not be partitioned into two or three partitions. Therefore, the multi-type tree partition indication information may not be signaled, but the multi-type tree partition indication information may be inferred from the second value.

[0193] Optionally, when the size (horizontal size and vertical size) of the coding unit corresponding to the node of the multi-type tree is the same as the maximum size (horizontal size and vertical size) of the binary tree and / or is twice as large as the maximum size (horizontal size and vertical size) of the ternary tree, the coding unit may not be further partitioned into two or three partitions. Therefore, the multi-type tree partition indication information may not be signaled, but may be derived from the second value. This is because when the coding unit is partitioned according to the binary tree partition structure and / or the ternary tree partition structure, a coding unit smaller than the minimum size of the binary tree and / or the minimum size of the ternary tree is generated.

[0194] Optionally, the binary tree partition or ternary tree partition may be limited based on the size of the virtual pipeline data unit (hereinafter, the pipeline buffer size). For example, when the coding unit is divided into sub-coding units that do not fit the pipeline buffer size by binary tree partition or ternary tree partition, the corresponding binary tree partition or ternary tree partition may be limited. The pipeline buffer size may be the size of the maximum transform block (e.g., 64×64). For example, when the pipeline buffer size is 64×64, the following division may be limited.

[0195] - N×M (N and / or M is 128) ternary tree partitions for coding units

[0196] - 128×N (N<=64) binary tree partitions in the horizontal direction for coding units

[0197] - N×128 (N<=64) binary tree partitions in the vertical direction for coding units

[0198] Optionally, when the depth of the coding unit corresponding to the node of the multi-type tree is equal to the maximum depth of the multi-type tree, the coding unit may not be further partitioned into two and / or three partitions. Therefore, the multi-type tree partition indication information may not be sent by a signal, but the multi-type tree partition indication information may be inferred from the second value.

[0199] Optionally, only when at least one of vertical binary tree partitioning, horizontal binary tree partitioning, vertical ternary tree partitioning, and horizontal ternary tree partitioning is possible for a coding unit corresponding to a node of a multi-type tree, a multi-type tree partition indication information may be signaled. Otherwise, the coding unit may not be partitioned into two and / or three partitions. Therefore, the multi-type tree partition indication information may not be signaled, but may be inferred from the second value.

[0200] Optionally, partition direction information may be signaled only when both vertical binary tree partitioning and horizontal binary tree partitioning or both vertical ternary tree partitioning and horizontal ternary tree partitioning are possible for a coding unit corresponding to a node of a multi-type tree. Otherwise, partition direction information may not be signaled, but may be derived from a value indicating a possible partition direction.

[0201] Optionally, partition tree information may be signaled only when both vertical binary tree partitioning and vertical ternary tree partitioning or both horizontal binary tree partitioning and horizontal ternary tree partitioning are possible for a coding tree corresponding to a node of a multi-type tree. Otherwise, partition tree information may not be signaled but may be derived from a value indicating a possible partition tree structure.

[0202] Figure 4 is a diagram illustrating an intra prediction process.

[0203] Figure 4 The arrows from the center to the outside in FIG. 1 represent the prediction direction of the intra prediction mode.

[0204] Intra-frame encoding and / or decoding may be performed by using reference samples of neighboring blocks of the current block. The neighboring blocks may be reconstructed neighboring blocks. For example, intra-frame encoding and / or decoding may be performed by using encoding parameters or values ​​of reference samples included in the reconstructed neighboring blocks.

[0205] The prediction block may represent a block generated by performing intra prediction. The prediction block may correspond to at least one of a CU, a PU, and a TU. The unit of the prediction block may have a size of one of a CU, a PU, and a TU. The prediction block may be a square block of a size of 2×2, 4×4, 16×16, 32×32, or 64×64, etc., or may be a rectangular block of a size of 2×8, 4×8, 2×16, 4×16, and 8×16, etc.

[0206] Intra-prediction may be performed according to an intra-prediction mode for the current block. The number of intra-prediction modes that the current block may have may be a fixed value, and may be a value determined differently according to properties of the prediction block. For example, the properties of the prediction block may include the size of the prediction block, the shape of the prediction block, and the like.

[0207] Regardless of the block size, the number of intra-frame prediction modes can be fixed to N. Alternatively, the number of intra-frame prediction modes can be 3, 5, 9, 17, 34, 35, 36, 65 or 67, etc. Optionally, the number of intra-frame prediction modes can vary according to the block size or the color component type or both the block size and the color component type. For example, the number of intra-frame prediction modes can vary depending on whether the color component is a luminance signal or a chrominance signal. For example, as the block size becomes larger, the number of intra-frame prediction modes can increase. Optionally, the number of intra-frame prediction modes of the luminance component block can be greater than the number of intra-frame prediction modes of the chrominance component block.

[0208] The intra prediction mode may be a non-angle mode or an angle mode. The non-angle mode may be a DC mode or a planar mode, and the angle mode may be a prediction mode having a specific direction or angle. The intra prediction mode may be represented by at least one of a mode number, a mode value, a mode number, a mode angle, and a mode direction. The number of intra prediction modes may be M, which is greater than 1, including non-angle modes and angle modes. In order to perform intra prediction on a current block, a step of determining whether a sample included in a reconstructed neighboring block can be used as a reference sample of the current block may be performed. When there are samples that cannot be used as reference samples of the current block, a value obtained by copying or interpolating at least one sample value of the samples included in the reconstructed neighboring block, or by both copying and interpolating, may be used to replace the unavailable sample value of the sample, so that the replaced sample value is used as the reference sample of the current block.

[0209] Figure 7 is a diagram showing reference samples that can be used for intra prediction.

[0210] like Figure 7 As shown, at least one of the reference sample line 0 to the reference sample line 3 can be used for intra prediction of the current block. Figure 7 In the example, the samples of fragment A and fragment F may be filled with the samples of the closest fragment B and fragment E, respectively, instead of being retrieved from the reconstructed neighboring blocks. Index information indicating the reference sample line to be used for intra prediction of the current block may be signaled. When the upper boundary of the current block is the boundary of the CTU, only the reference sample line 0 may be available. Therefore, in this case, the index information may not be signaled. When reference sample lines other than the reference sample line 0 are used, filtering for the prediction block, which will be described later, may not be performed.

[0211] When intra prediction is performed, a filter may be applied to at least one of a reference sample and a prediction sample based on an intra prediction mode and a current block size.

[0212] In the case of the planar mode, when generating a prediction block of the current block, according to the position of the prediction target sample within the prediction block, the sample value of the prediction target sample may be generated by using the weighted sum of the upper reference sample and the left reference sample of the current sample and the upper right reference sample and the lower left reference sample of the current block. In addition, in the case of the DC mode, when generating the prediction block of the current block, the average value of the upper reference sample and the left reference sample of the current block may be used. In addition, in the case of the angular mode, the prediction block may be generated by using the upper reference sample, the left reference sample, the upper right reference sample and / or the lower left reference sample of the current block. In order to generate the prediction sample value, interpolation of real number units may be performed.

[0213] In the case of intra prediction between color components, a prediction block of a current block of a second color component may be generated based on a corresponding reconstruction block of a first color component. For example, the first color component may be a luminance component, and the second color component may be a chrominance component. For intra prediction between color components, parameters of a linear model between the first color component and the second color component may be derived based on a template. The template may include the upper and / or left neighboring samples of the current block and the upper and / or left neighboring samples of the reconstruction block of the first color component corresponding thereto. For example, the sample value of the first color component having the maximum value among the samples in the template and the sample value of the second color component corresponding thereto, and the sample value of the first color component having the minimum value among the samples in the template and the sample value of the second color component corresponding thereto may be used to derive the parameters of the linear model. When deriving the parameters of the linear model, the corresponding reconstruction block may be applied to the linear model to generate a prediction block of the current block. Depending on the video format, subsampling may be performed on the reconstruction block of the first color component and the neighboring samples of the corresponding reconstruction block. For example, when one sample of the second color component corresponds to four samples of the first color component, the four samples of the first color component may be subsampled to calculate one corresponding sample. In this case, parameter derivation of the linear model and intra prediction between color components may be performed based on the corresponding subsampled samples. Whether to perform intra prediction between color components and / or the range of the template may be signaled as an intra prediction mode.

[0214] The current block may be partitioned into two sub-blocks or four sub-blocks in the horizontal direction or the vertical direction. The partitioned sub-blocks may be reconstructed sequentially. That is, intra prediction may be performed on the sub-block to generate a sub-prediction block. In addition, inverse quantization and / or inverse transformation may be performed on the sub-block to generate a sub-residual block. The reconstructed sub-block may be generated by adding the sub-prediction block to the sub-residual block. The reconstructed sub-block may be used as a reference sample for intra prediction of a subsequent sub-block. The sub-block may be a block including a predetermined number (e.g., 16) or more samples. Thus, for example, when the current block is an 8×4 block or a 4×8 block, the current block may be partitioned into two sub-blocks. In addition, when the current block is a 4×4 block, the current block may not be partitioned into sub-blocks. When the current block has other sizes, the current block may be partitioned into four sub-blocks. Information on whether intra prediction is performed based on sub-blocks and / or partition directions (horizontal or vertical) may be sent by a signal. Intra prediction based on sub-blocks may be performed only when reference sample line 0 is used. When subblock-based intra prediction is performed, filtering for a prediction block, which will be described later, may not be performed.

[0215] The final prediction block may be generated by performing filtering on the prediction block predicted by the intra-frame. The filtering may be performed by applying a predetermined weight to the filtering target sample, the left reference sample, the upper reference sample, and / or the upper left reference sample. The weight and / or reference sample (range, position, etc.) used for filtering may be determined based on at least one of the block size, the intra-frame prediction mode, and the position of the filtering target sample in the prediction block. The filtering may be performed only in the case of a predetermined intra-frame prediction mode (e.g., DC, plane, vertical, horizontal, diagonal, and / or adjacent diagonal mode). The adjacent diagonal mode may be a mode in which k is added to the diagonal mode or subtracted from the diagonal mode. For example, k may be a positive integer of 8 or less.

[0216] The intra-frame prediction mode of the current block may be entropy encoded / decoded by predicting the intra-frame prediction mode of a block existing adjacent to the current block. When the intra-frame prediction mode of the current block is the same as that of the neighboring block, information that the intra-frame prediction mode of the current block is the same as that of the neighboring block may be signaled by using predetermined flag information. In addition, indicator information of the intra-frame prediction mode that is the same as the intra-frame prediction mode of the current block among the intra-frame prediction modes of multiple neighboring blocks may be signaled. When the intra-frame prediction mode of the current block is different from that of the neighboring block, the intra-frame prediction mode information of the current block may be entropy encoded / decoded by performing entropy encoding / decoding based on the intra-frame prediction mode of the neighboring block.

[0217] Figure 5 is a diagram illustrating an embodiment of an inter-picture prediction process.

[0218] exist Figure 5 In , a rectangle can represent a picture. Figure 5In FIG. 1 , the arrow indicates the prediction direction. According to the encoding type of the picture, the picture can be classified into an intra picture (I picture), a predicted picture (P picture) and a bi-predicted picture (B picture).

[0219] An I picture may be encoded by intra prediction without requiring inter-picture prediction. A P picture may be encoded by inter-picture prediction using a reference picture existing in one direction (i.e., forward or backward) relative to the current block. A B picture may be encoded by inter-picture prediction using a reference picture existing in two directions (i.e., forward and backward) relative to the current block. When inter-picture prediction is used, the encoder may perform inter-picture prediction or motion compensation, and the decoder may perform corresponding motion compensation.

[0220] Hereinafter, embodiments of inter-picture prediction will be described in detail.

[0221] Reference pictures and motion information may be used to perform inter-picture prediction or motion compensation.

[0222] The motion information of the current block may be derived during inter-picture prediction by each of the encoding device 100 and the decoding device 200. The motion information of the current block may be derived by using the motion information of a reconstructed neighboring block, the motion information of a co-located block (also referred to as a col block or a co-located block), and / or the motion information of a block adjacent to the co-located block. The co-located block may represent a block in a previously reconstructed co-located picture (also referred to as a col picture or a co-located picture) that is spatially located at the same position as the current block. The co-located picture may be one of one or more reference pictures included in a reference picture list.

[0223] The derivation method of motion information may be different according to the prediction mode of the current block. For example, the prediction modes applied to inter prediction include AMVP mode, merge mode, skip mode, merge mode with motion vector difference, sub-block merge mode, triangle partition mode, inter-intra combined prediction mode, affine mode, etc. Here, the merge mode may be referred to as motion merge mode.

[0224] For example, when AMVP is used as a prediction mode, at least one of a motion vector of a reconstructed neighboring block, a motion vector of a co-located block, a motion vector of a block adjacent to the co-located block, and a (0,0) motion vector may be determined as a motion vector candidate for the current block, and a motion vector candidate list may be generated by using the motion vector candidate. The motion vector candidate for the current block may be derived by using the generated motion vector candidate list. The motion information of the current block may be determined based on the derived motion vector candidate. The motion vector of the co-located block or the motion vector of a block adjacent to the co-located block may be referred to as a temporal motion vector candidate, and the motion vector of the reconstructed neighboring block may be referred to as a spatial motion vector candidate.

[0225] The encoding device 100 may calculate a motion vector difference (MVD) between a motion vector of a current block and a motion vector candidate, and may perform entropy coding on the motion vector difference (MVD). In addition, the encoding device 100 may perform entropy coding on a motion vector candidate index and generate a bitstream. The motion vector candidate index may indicate the best motion vector candidate among the motion vector candidates included in the motion vector candidate list. The decoding device may perform entropy decoding on the motion vector candidate index included in the bitstream, and may select a motion vector candidate of a decoding target block from the motion vector candidates included in the motion vector candidate list by using the entropy-decoded motion vector candidate index. In addition, the decoding device 200 may add the entropy-decoded MVD to the motion vector candidate extracted by entropy decoding, thereby deriving a motion vector of the decoding target block.

[0226] In addition, the encoding apparatus 100 may perform entropy encoding on the resolution information of the calculated MVD. The decoding apparatus 200 may adjust the resolution of the entropy-decoded MVD using the MVD resolution information.

[0227] In addition, the encoding device 100 calculates a motion vector difference (MVD) between a motion vector in the current block and a motion vector candidate based on an affine model, and performs entropy encoding on the MVD. The decoding device 200 derives a motion vector based on each sub-block by deriving an affine controlled motion vector of a decoding target block according to the sum of the entropy-decoded MVD and the affine controlled motion vector candidate.

[0228] The bitstream may include a reference picture index indicating a reference picture. The reference picture index may be entropy encoded by the encoding apparatus 100 and then signaled as a bitstream to the decoding apparatus 200. The decoding apparatus 200 may generate a prediction block of a decoding target block based on the derived motion vector and the reference picture index information.

[0229] Another example of a method of deriving motion information of a current block may be a merge mode. The merge mode may represent a method of merging motions of a plurality of blocks. The merge mode may represent a mode of deriving motion information of a current block from motion information of a neighboring block. When the merge mode is applied, the motion information of the reconstructed neighboring block and / or the motion information of the same-position block may be used to generate a merge candidate list. The motion information may include at least one of a motion vector, a reference picture index, and an inter-picture prediction indicator. The prediction indicator may indicate unidirectional prediction (L0 prediction or L1 prediction) or bidirectional prediction (L0 prediction and L1 prediction).

[0230] The merge candidate list may be a list of stored motion information. The motion information included in the merge candidate list may be at least one of the following: motion information of a neighboring block adjacent to the current block (spatial merge candidate), motion information of a co-located block of the current block in a reference picture (temporal merge candidate), new motion information generated by combining motion information present in the merge candidate list, motion information of a block encoded / decoded before the current block (historical-based merge candidate), and a zero merge candidate.

[0231] The encoding device 100 may generate a bitstream by performing entropy encoding on at least one of a merge flag and a merge index, and may signal the bitstream to the decoding device 200. The merge flag may be information indicating whether a merge mode is performed for each block, and the merge index may be information indicating which neighboring block among neighboring blocks of the current block is a merge target block. For example, the neighboring blocks of the current block may include a left neighboring block located on the left side of the current block, an upper neighboring block arranged above the current block, and a temporal neighboring block temporally adjacent to the current block.

[0232] In addition, the encoding device 100 performs entropy encoding on the correction information for correcting the motion vector in the motion information of the merge candidate, and transmits it to the decoding device 200 by signal. The decoding device 200 may correct the motion vector of the merge candidate selected by the merge index based on the correction information. Here, the correction information may include at least one of information on whether to perform correction, correction direction information, and correction size information. As described above, the prediction mode in which the motion vector of the merge candidate is corrected based on the correction information transmitted by the signal may be referred to as a merge mode with a motion vector difference.

[0233] The skip mode may be a mode in which the motion information of the neighboring blocks is applied to the current block as it is. When the skip mode is applied, the encoding apparatus 100 may perform entropy encoding on information of the fact of which block's motion information is to be used as the motion information of the current block to generate a bitstream, and may signal the bitstream to the decoding apparatus 200. The encoding apparatus 100 may not signal a syntax element regarding at least any one of the motion vector difference information, the coded block flag, and the transform coefficient level to the decoding apparatus 200.

[0234] The sub-block merge mode may represent a mode for deriving motion information in units of sub-blocks of a coding block (CU). When the sub-block merge mode is applied, the sub-block merge candidate list may be generated using motion information of a sub-block co-located with the current sub-block in a reference image (sub-block based temporal merge candidates) and / or affine control point motion vector merge candidates.

[0235] The triangular partition mode may denote a mode of deriving motion information by partitioning the current block into diagonal directions, deriving each prediction sample using each of the derived motion information, and deriving the prediction sample of the current block by weighting each of the derived prediction samples.

[0236] The inter-intra combined prediction mode may mean a mode of deriving a prediction sample of a current block by weighting a prediction sample generated by inter prediction and a prediction sample generated by intra prediction.

[0237] The decoding apparatus 200 may correct the derived motion information by itself. The decoding apparatus 200 may search for a predetermined area based on a reference block indicated by the derived motion information, and derive motion information having a minimum SAD as the corrected motion information.

[0238] The decoding apparatus 200 may compensate for prediction samples derived through inter-frame prediction using optical flow.

[0239] Figure 6 is a diagram illustrating transform and quantization processing.

[0240] like Figure 6 As shown in , a transform process and / or a quantization process is performed on the residual signal to generate a quantized level signal. The residual signal is the difference between the original block and the prediction block (i.e., an intra-frame prediction block or an inter-frame prediction block). The prediction block is a block generated by intra-frame prediction or inter-frame prediction. The transform can be a primary transform, a secondary transform, or both a primary transform and a secondary transform. The primary transform of the residual signal generates a transform coefficient, and the secondary transform of the transform coefficient generates a secondary transform coefficient.

[0241] At least one scheme selected from various predefined transform schemes is used to perform the primary transform. For example, examples of the predefined transform schemes include discrete cosine transform (DCT), discrete sine transform (DST) and Karhunen-Loève transform (KLT). The transform coefficients generated by the primary transform may undergo a secondary transform. The transform scheme for the primary transform and / or the secondary transform may be determined based on the coding parameters of the current block and / or the neighboring blocks of the current block. Optionally, transform information indicating the transform scheme may be sent by a signal. DCT-based transforms may include, for example, DCT-2, DCT-8, etc. DST-based transforms may include, for example, DST-7.

[0242] A quantized level signal (quantized coefficient) may be generated by performing quantization on a residual signal or on a result of performing a primary transform and / or a secondary transform. Depending on the intra prediction mode of the block or the block size / shape, the quantized level signal may be scanned according to at least one of a diagonal upper right scan, a vertical scan, and a horizontal scan. For example, when the coefficients are scanned according to a diagonal upper right scan, the coefficients in block form are changed to a one-dimensional vector form. In addition to the diagonal upper right scan, a horizontal scan that scans the coefficients in a two-dimensional block form horizontally or a vertical scan that scans the coefficients in a two-dimensional block form vertically may be used according to the intra prediction mode and / or the size of the transform block. The scanned quantized level coefficients may be entropy encoded for insertion into a bitstream.

[0243] The decoder performs entropy decoding on the bit stream to obtain quantized level coefficients. The quantized level coefficients can be arranged in a two-dimensional block form by inverse scanning. For the inverse scanning, at least one of diagonal upper right scanning, vertical scanning, and horizontal scanning can be used.

[0244] The quantized level coefficients may then be dequantized, then inversely transformed secondary if necessary, and finally inversely transformed primary if necessary to generate a reconstructed residual signal.

[0245] Inverse mapping in the dynamic range can be performed for the luminance component reconstructed by intra prediction or inter prediction before in-loop filtering. The dynamic range can be divided into 16 equal segments, and the mapping function of each segment can be sent by signal. The mapping function can be sent by signal at the slice level or parallel block group level. The inverse mapping function for performing inverse mapping can be derived based on the mapping function. In-loop filtering, reference picture storage and motion compensation are performed in the inverse mapping area, and the prediction block generated by inter prediction is converted to the mapping area via mapping using the mapping function, and then used to generate the reconstructed block. However, since intra prediction is performed in the mapping area, the prediction block generated by intra prediction can be used to generate the reconstructed block without mapping / inverse mapping.

[0246] When the current block is a residual block of a chroma component, the residual block can be converted to an inverse mapping area by performing scaling on the chroma component of the mapping area. The availability of scaling can be signaled at the slice level or the parallel block group level. Scaling can be applied only when the mapping of the luminance component is available and the division of the luminance component and the division of the chroma component follow the same tree structure. Scaling can be performed based on the average value of the sample value of the luminance prediction block corresponding to the chroma block. In this case, when the current block uses inter-frame prediction, the luminance prediction block can represent the mapped luminance prediction block. The value required for scaling can be derived by using the index reference lookup table of the fragment to which the average value of the sample value of the luminance prediction block belongs. Finally, the residual block can be converted to an inverse mapping area by scaling the residual block using the derived value. Then, chroma component block recovery, intra-frame prediction, inter-frame prediction, in-loop filtering, and reference picture storage can be performed in the inverse mapping area.

[0247] Information indicating whether mapping / inverse mapping of luma components and chroma components is available may be signaled through a sequence parameter set.

[0248] The prediction block of the current block may be generated based on a block vector indicating the displacement between the current block and the reference block in the current picture. In this way, the prediction mode for generating the prediction block with reference to the current picture is called an intra-block copy (IBC) mode. The IBC mode may be applied to M×N (M<=64, N<=64) coding units. The IBC mode may include a skip mode, a merge mode, an AMVP mode, and the like. In the case of a skip mode or a merge mode, a merge candidate list is constructed, and a merge index is signaled so that a merge candidate can be specified. The block vector of the specified merge candidate may be used as the block vector of the current block. The merge candidate list may include at least one of a spatial candidate, a history-based candidate, a candidate based on the average of two candidates, and a zero merge candidate. In the case of the AMVP mode, a difference block vector may be signaled. In addition, a prediction block vector may be derived from the left neighboring block and the upper neighboring block of the current block. The index of the neighboring block to be used may be signaled. The prediction block in the IBC mode is included in the current CTU or the left CTU and is limited to blocks in the reconstructed area. For example, the value of the block vector may be limited so that the prediction block of the current block is located in the region of three 64×64 blocks preceding the 64×64 block to which the current block belongs in the encoding / decoding order. By limiting the value of the block vector in this way, the memory consumption and device complexity of the implementation scheme according to the IBC mode may be reduced.

[0249] In the following, reference will be made to Figures 8 to 21 Embodiments of the present invention are described in detail.

[0250] Bidirectional optical flow (BIO) may refer to a motion correction technique in pixel or sub-pixel units performed according to block-based motion compensation. That is, BIO may refer to a technique for correcting a bidirectional prediction signal in pixel or sub-pixel units.

[0251] For example, given the pixel value I at time t t When , equation 1 can be obtained through the first-order Taylor expansion.

[0252] Equation 1

[0253]

[0254] In the assumption I t0 Located in I t The following equation 2 is established on the motion trajectory and the optical flow is valid along the motion trajectory.

[0255] Equation 2

[0256]

[0257] Based on Equation 2, the following Equation 3 can be derived from Equation 1.

[0258] Equation 3

[0259]

[0260] and When it is considered as the movement speed, V x0 and V y0 Therefore, the following equation 4 can be derived from equation 3.

[0261] Equation 4

[0262] I t =I t0 -G x0 ·V x0 (t-t0)-G y0 ·V y0 ·(t-t0)

[0263] When a forward reference picture at time t0 and a backward reference picture at time t1 exist, and (t-t0)=(t-t1)=Δt=1, a pixel value at time t may be calculated based on Equation 5 below.

[0264] Equation 5

[0265] I t =I t0 -G x0 ·V x0 (t-t0)-G y0 ·Vy0 (t-t0) = I t0 +G x0 ·V x0 +G y0 ·V y0

[0266] I t =I t1 -G x1 ·V x1 (t-t1)-G y1 ·V y1 (t-t1) = I t1 -G x1 ·V x1 -G y1 ·V y1

[0267]

[0268] Furthermore, since the motion follows a trajectory, it is assumed that V x0 =V x1 =V x , V y0 =V y1 =V y Therefore, the following equation 6 can be derived from the above equation 5.

[0269] Equation 6

[0270]

[0271] ΔG x =G x0 -G x1 , ΔG y =G y0 -G y1

[0272] In the above equation, ΔG can be obtained from the reconstructed reference picture x , ΔG y .

[0273] In Equation 6 above, Can correspond to general bidirectional prediction. In addition, Can represent BIO offset.

[0274] The motion correction vector V can be obtained using the following equation 7 in the encoder and decoder: x and V y .

[0275] Equation 7

[0276]

[0277] In the above Equation 7, r and m may have a value of 0, and the threshold value 'limit' may be determined according to the bit depth of the luma component.

[0278] In Equation 7 above, the pixel values ​​I at time t0 and t1 can be used (0) and I (1) and G x0 , G y0 , G x1 , G x1 To calculate s1, s2, s3, s5 and s6, as shown in Equation 8 below.

[0279] Equation 8

[0280]

[0281] In the above Equation 8, s1, s2, s3, s5, and s6 may represent BIO-related parameters of a pixel position (i, j) in a target block region for motion correction based on a pixel unit.

[0282] In Equation 8 above, Represents the pixel gradient value G of the horizontal component at the position (i, j) of the L1 reference image prediction block x1 . Represents the pixel gradient value G of the vertical component at the position (i, j) of the L1 reference image prediction block y1 . Represents the pixel gradient value G of the horizontal component at the position (i, j) of the L0 reference image prediction block x0 . Represents the pixel gradient value G of the vertical component at the position (i, j) of the L0 reference image prediction block y0 I (1) (i,j) and I (0) (i, j) may represent a predicted pixel value at a position (i, j) of an L1 reference image prediction block and a predicted pixel value at a position (i, j) of an L0 reference image prediction block. Here, the gradient value may represent a gradient value.

[0283] When BIO-based motion correction is performed in units of sub-blocks, the above Equation 8 may be derived in units of sub-blocks as shown in the following Equation 9.

[0284] Equation 9

[0285]

[0286] θ(i,j)=(I (1) (i,j)>>n b )-(I (0)(i,j)>>n b )

[0287]

[0288] In the above equation 9, ψ can be expressed by the BIO parameter in the present invention. x (i,j),ψ y (i, j), θ(i, j), and can be represented by the BIO related parameters in the present invention S1, S2, S3, S5, S6.

[0289] In Equation 9 above, the range of values ​​i and j may be determined by the size of the sub-block, the position of the sub-block in the block, and the window size applied to the sub-block.

[0290] For example, when there are four 4×4 sub-blocks for an 8×8 block and the window size applied to the sub-blocks is 6×6, (i, j) for the first 4×4 sub-block on the upper left side may have a value of -1≤i≤4, -1≤j≤4, (i, j) for the second 4×4 sub-block on the upper right side may have a value of 3≤i≤8, -1≤j≤4, (i, j) for the third 4×4 sub-block on the lower left side may have a value of -1≤i≤4, 3≤j≤8, and (i, j) for the fourth 4×4 sub-block on the lower right side may have a value of 3≤i≤8, 3≤j≤8.

[0291] In Equation 9 above, n a and n b May be a parameter used to adjust the bit depth and may have a positive integer value.

[0292] For example, a =3,n b =6.

[0293] When calculating BIO-related parameters in units of sub-blocks, pixel values ​​and gradient values ​​at block area positions close to the boundaries can be used to fill in unavailable pixel values ​​and gradient values ​​outside the boundary of the decoding target block area. That is, unavailable pixel values ​​and gradient values ​​outside the boundary of the decoding target block area are replaced with pixel values ​​and gradient values ​​at block area positions close to the boundaries that can be used to calculate BIO-related parameters in units of sub-blocks.

[0294] As shown in the following equation 10, V can be calculated in units of sub-blocks using the BIO-related parameters s1, s2, s3, s5, and s6 obtained in units of sub-blocks through the above equation 9. x and V y .

[0295] Equation 10

[0296]

[0297] In Equation 10 above, S 2,m =S2>>12,S 2,s =S2&(2 12 -1) and limit = 2 13-Bitdepth .

[0298] In the present invention, s1, s2, s3, s5 and s6 may be represented by s.

[0299] Based on the use of V calculated in pixels or sub-blocks x and V y As well as equation 6 of the pixel value and gradient value at each pixel position, the BIO offset value at each pixel position can be obtained, and then the BIO offset value at each pixel position is added to the prediction signal subjected to bidirectional weighted summation to calculate the final prediction signal of the block.

[0300] As shown in Equation 11 below, when two different reference pictures are temporally located before or after a current picture, a prediction signal at time t may be calculated considering a temporal distance between the current picture and the reference picture.

[0301] Equation 11

[0302] I t =I t0 -G x0 ·V x0 (t-t0)-G y0 ·V y0 (t-t0) = I t0 +G x0 ·V x0 TD0+G y0 ·V y0 TD0

[0303] I t =I t1 -G x1 ·V x1 (t-t1)-G y1 ·V y1 (t-t1) = I t1 +G x1 ·V x1 TD1+G y1 ·V y1 TD1

[0304]

[0305] G′ x0 =G x0 TD0, G′ y0 =G y0TD0, G′ x1 =G x1 TD1, G′ y1 =G y1 TD1

[0306] According to one embodiment of the present invention, when a picture and / or a slice belonging to a current block can be encoded / decoded by inter-frame prediction using a reference picture present in a bidirectional reference picture list, encoding / decoding can be performed by applying BIO even if the current block has only first motion information. In the present invention, the first motion information may represent motion information in the L0 direction or motion information in the L1 direction of the current block.

[0307] According to the present invention, when the current block has only the first motion information, the second motion information may be derived, and then encoding / decoding may be performed by applying the BIO.

[0308] When deriving the second motion information, whether to derive the second motion information may be determined based on the first motion vector of the current block.

[0309] For example, whether to derive the second motion information may be determined based on a result of comparing the first motion vector value of the current block with a predetermined threshold. For example, as shown in the following equation 12, the first X-direction motion vector MV of the current block may be calculated based on the first X-direction motion vector MV of the current block. x0 and Y direction motion vector MV y0 According to one embodiment, when MV x0 and MV y0 When both are lower than the threshold, it can be determined to derive the second motion information.

[0310] Equation 12

[0311] |MV x0 |<=Th&&|MV y0 |<=Th

[0312] After determining to derive the second motion information, for example, when the above conditions are met, the second motion information may be derived based on the first motion information of the current block. In addition, the BIO may be applied to the current block using the first motion information and the derived second motion information.

[0313] The threshold Th for determining whether to derive the second motion information may be a predefined value, or may be included in the bitstream and transmitted in the bitstream. The threshold may be adaptively determined based on encoding parameters of the current block, such as the size and / or shape of the current block.

[0314] For example, the threshold may be sent via syntax data at a sequence parameter, a picture parameter, a slice header, or a block level.

[0315] The second motion information may be derived based on a temporal distance between the current picture and the reference picture.

[0316] Figure 8 is a diagram illustrating various embodiments of deriving second motion information based on first motion information.

[0317] For example, Figure 8 As shown in (a) of FIG. 1 , only when there is a first reference picture Ref0 having a temporal distance TD between the current picture CurPic and the first motion information MV0 of the current block in the second reference picture list c When a picture at the same temporal distance and having a POC different from that of the first reference picture Ref0 , the corresponding picture may be used as the second reference picture Ref1 , thereby deriving second motion information MV1 .

[0318] As shown in Equation 13 below, when a picture having the same temporal distance is used as a second reference picture, the second motion vector may be derived from the first motion vector.

[0319] Equation 13

[0320] MV x1 =-MV x0 ,MV y1 =-MV y0

[0321] For example, Figure 8 As shown in (b), when there is no picture having the same temporal distance TD0 as the temporal distance TD0 between the current picture CurPic and the first reference picture Ref0 indicated by the first motion information MV0 of the current block in the second reference picture list, a picture having the shortest temporal distance to the current picture CurPic and having a POC different from that of the first reference picture Ref0 may be used as the second reference picture Ref1, thereby deriving the second motion information MV1.

[0322] In the above example, as shown in Equation 14 below, the second motion vector MV1 may be derived using the first motion vector MV0, the temporal distance TD0 between the current picture CurPic and the first reference picture Ref0, and the temporal distance TD1 between the current picture CurPic and the second reference picture Ref1.

[0323] Equation 14

[0324]

[0325] For example, regardless of the temporal distance TD0 between the current picture CurPic and the first reference picture Ref0 indicated by the first motion information MV0 of the current block, the picture in the second reference picture list that always has the shortest temporal distance to the current picture CurPic and has a POC different from the POC of the first reference picture Ref0 can be used as the second reference picture Ref1, thereby deriving the second motion information MV1.

[0326] In the above example, as shown in Equation 15 below, the second motion vector MV1 may be derived using the first motion vector MV0, the temporal distance TD0 between the current picture CurPic and the first reference picture Ref0, and the temporal distance TD1 between the current picture CurPic and the second reference picture Ref1.

[0327] Equation 15

[0328]

[0329] When deriving the second motion information MV1 , the second motion information MV1 may be derived through motion prediction using the current picture CurPic and reference pictures in a reference picture list in a direction different from the direction of the first motion information MV0 , Ref0 of the current block.

[0330] For example, Figure 8 As shown in (c) of FIG. 1 , a block having a minimum distortion value with respect to the prediction block P0 within a predetermined search range for reference pictures in the reference picture list in different directions may be found based on the prediction block P0 generated from the first motion information of the current block. A (0,0) motion vector indicating the same position corresponding to the current block may be used as an initial motion vector for motion search, and as shown in FIG. Figure 8 As shown in (a) or (b) of , a motion vector derived based on the first motion vector MV0 may be used to find a block having a minimum distortion value within a predetermined search range.

[0331] As shown in the following Equation 16, the block P having the minimum distortion value indicating the prediction block P0 and the prediction block P0 is used. min The distance between the offset MV search And MV0 indicating the distance offset between the current block and the prediction block P0, the second motion information of the current block can be derived. In addition, the reference picture index of the reference picture Ref1 including the prediction block P0 can be used as the reference picture index of the second motion information.

[0332] Equation 16

[0333] MV x1 =MV search,x +MV x0

[0334] MV y1 =MV search,y +MV y0

[0335] When the second motion information obtained in the above embodiment is available, a final prediction signal may be generated by applying a pixel-unit or sub-block-unit BIO to the current block using the first reference picture Ref0 and the first motion information of the current block and the second reference picture Ref1 and the second motion information.

[0336] When the BIO is applied to the current block using the first motion information and the second motion information, when the first reference picture Ref1 indicated by the first motion information and the second reference picture Ref1 indicated by the second motion information exist on different time axes based on the current picture at time t, and the temporal distance TD0 between the current picture and the first reference picture and the temporal distance TD1 between the current picture and the second reference picture are different from each other, the BIO offset can be calculated considering the temporal distance between the current picture and the reference picture.

[0337] For example, when the following conditions are met, the BIO offset may be calculated considering the temporal distance between the current picture and the reference picture as shown in the following Equation 17, thereby obtaining a final prediction signal of the current block.

[0338] Equation 17

[0339] TD0=(t0-t),

[0340] TD1=(t-t1),

[0341] TD0×TD1>0, TD0≠TD1

[0342] I t =I t0 -G x0 ·V x0 (t-t0)-G y0 ·V y0 (t-t0) = I t0 +G x0 ·V x0 TD0+G y0 ·V y0 TD0

[0343] I t =I t1 -G x1 ·V x1 (t-t1)-G y1 ·V y1 (t-t1) = I t1 -G x1 ·V x1TD1-G y1 ·V y1 TD1

[0344]

[0345] G′ x0 =G x0 TD0, G′ y0 =G y0 TD0, G′ x1 =G x1 TD1, G′ y1 =G y1 TD1

[0346] When BIO is applied to the current block using the first motion information of the current block and the second motion information of the current block, only when the first reference picture Ref0 indicated by the first motion information and the second reference picture Ref1 indicated by the second motion information exist on different time axes based on the current picture at time t, and the time distance TD0 between the current picture and the first reference picture and the time distance TD1 between the current picture and the second reference picture are equal to each other, BIO can be applied, and the BIO offset can be added to the prediction signal of the current block to obtain the final prediction signal. The first motion information may represent motion information in a first prediction direction, and the second motion information may represent motion information in a second prediction direction. In addition, two reference pictures existing on different axes may represent that the two reference pictures are located in different directions relative to the current picture. For example, when the relationship between the POC POC_ref0 of the first reference picture, the POC POC_cur of the current picture, and the POCPOC_ref1 of the second reference picture satisfies the following conditions, this may represent that the two reference pictures exist on different time axes based on the current picture, and the time distance TD_0 between the current picture and the first reference picture and the time distance TD_1 between the current picture and the second reference picture are equal to each other.

[0347] Condition: (POC_ref0–POC_cur)==(POC_cur–POC_ref1)

[0348] When applying BIO to the current block using the first motion information and the second motion information, when the first reference picture Ref0 indicated by the first motion information and the second reference picture Ref1 indicated by the second motion information exist on different time axes based on the current picture at time t, the motion vector of the first motion information is (0,0) and the motion vector of the second motion information is (0,0), BIO may not be applied to the current block.

[0349] When BIO is applied to the current block using the first motion information and the second motion information, when the first reference picture Ref0 indicated by the first motion information and the second reference picture Ref1 indicated by the second motion information exist on different time axes based on the current picture at time t, and the temporal distance TD0 between the current picture and the first reference picture and the temporal distance TD1 between the current picture and the second reference picture are equal to each other, the motion vector of the first motion information is (0,0), and the motion vector of the second motion information is (0,0), BIO may not be applied to the current block.

[0350] Whether the two reference pictures exist on different axes based on the current picture can be determined by the difference between the POC of the current picture and the POC of the reference picture. For example, when the relationship between the POC POC_ref0 of the first reference picture, the POC POC_cur of the current picture, and the POC POC_ref1 of the second reference picture satisfies the following conditions, this may indicate that the two reference pictures exist on different time axes relative to the current picture.

[0351] Condition: (POC_ref0–POC_cur)×(POC_cur–POC_ref1)>0

[0352] The first motion information and the second motion information may be used to calculate the gradient G used in the process of calculating the BIO offset. x0 ,G y0 ,G x1 ,G y1 .

[0353] When the motion vector indicates a sub-pixel position in a reference picture, the gradient values ​​of the vertical and horizontal components at the corresponding sub-pixel position may be calculated by applying a filter using values ​​of neighboring integer pixel positions. Tables 1 and 2 below show filter coefficients of the interpolation filter.

[0354] Table 1

[0355]

[0356] Table 2

[0357]

[0358] When the motion vector indicates a sub-pixel position, the motion vector may be rounded to an integer pixel position closest to the sub-pixel position, and then the gradient values ​​of the vertical component and the horizontal component may be calculated using the neighboring integer pixel values. That is, when the motion vector indicates a sub-pixel position, the motion vector may be rounded, and then the pixel value of the integer pixel position indicated by the rounded motion vector may be used for gradient calculation. In this case, the gradient values ​​of the vertical component and the horizontal component may be calculated using only the filter coefficient at the pixel position 0 of Table 1.

[0359] For example, in the case of 1 / 16 motion vector accuracy and horizontal and vertical motion vector magnitudes of (15, 15), rounding may be performed as shown in the following Equation 18 to have a motion vector (16, 16) value, and then the gradient value of the horizontal component may be calculated using the pixel values ​​of integer pixel positions and the filter coefficients (8, -39, -3, 46, -17, 5). In the case of 1 / 16 motion vector accuracy, the shift value may be 4, and in the case of 1 / 8 motion vector accuracy, the shift value may be 3.

[0360] Equation 18

[0361] roundMV(x,y)=((MV x +(1<<shift-1))>>shift)<<shift, ((MV y +(1<<shift-1))>>shift)<<shift

[0362] When the motion vector indicates a sub-pixel position, a pixel value interpolated at the sub-pixel position may be generated, and then the interpolated pixel value may be used to calculate the gradient values ​​of the vertical component and the horizontal component. The gradient may be calculated using a [-1, 0, 1] filter for the interpolated pixel value. The gradient values ​​of the vertical component and the horizontal component at the corresponding position in the prediction block may be calculated for the motion-compensated prediction signal.

[0363] For example, as shown in the following Equation 19, when a [-1, 0, 1] filter is applied, the gradient values ​​of the vertical component and the horizontal component may be calculated for the motion compensated prediction signal of the L0 reference picture. (0) (i, j) may represent the motion compensated prediction signal value at the position (i, j).

[0364] Equation 19

[0365]

[0366] Fig. 9is a diagram illustrating an example of generating a motion-compensated predicted pixel value at a sub-pixel position and then calculating gradient values ​​of a vertical component and a horizontal component using the corresponding pixel value.

[0367] Fig. 9 (a) shows an embodiment of obtaining gradients at all pixel positions for a 4×4 block, ie, pixel values ​​of an area filled with a pattern are required.

[0368] For example, if a [-1,0,1] filter is applied to calculate the gradient value at each position in a 4×4 block, then the pixel value at the (-1,0) position outside the 4×4 block area is required to calculate the horizontal gradient (G) at the upper left (0,0) position. 0,0 In addition, the pixel value at (4,0) outside the 4×4 block area is needed to calculate the horizontal gradient (G 3,0 ). When an 8-tap interpolation filter is applied to generate interpolated pixel values ​​for a block of size W (horizontal) × H (vertical), (W+7) × (H+7) pixel values ​​from a reference picture are required, but in the above method, a total of (W+7+2) × (H+7+2) pixel values ​​obtained by adding 2 pixels in the horizontal and vertical directions are required to additionally calculate the gradient value.

[0369] In order to reduce the memory bandwidth of the reference picture, a bilinear interpolation filter may be used for pixel values ​​outside the target block area that are otherwise required to calculate the gradient at the block boundary position.

[0370] In order to reduce the memory bandwidth of the reference image, the pixel value at the integer pixel position closest to the left or right of the sub-pixel position indicated by the motion vector in the reference picture can be used as the pixel value outside the target block area without performing a separate interpolation process.

[0371] In order to reduce the memory bandwidth of the reference picture, the pixel value at the integer pixel position closest to the sub-pixel position indicated by the motion vector in the reference picture may be used as the pixel value outside the target block area without performing a separate interpolation process. Fig. 9 In (a), the pixel value at the integer pixel position in the reference picture that is closest to the sub-pixel position indicated by the motion vector can be used as the pixel value of the area filled with the pattern (i.e., the pixel value outside the target block area) without performing separate interpolation processing.

[0372] That is, when a motion vector indicates a sub-pixel position outside a target block area in a reference image, the motion vector may be rounded to the nearest integer position without performing separate interpolation processing, and gradient calculation may be performed using the pixel value of the rounded integer pixel position.

[0373] For example, the pixel value outside the target block area may be replaced by the pixel value of the position indicated by (xIntL+(xFracL>>3)-1), (yIntL+(yFracL>>3)–1) in the reference picture, thereby performing gradient calculation. Here, (xIntL, yIntL) may represent the integer pixel unit position of the motion vector, and (xFracL, yFracL) may represent the sub-pixel unit position of the motion vector.

[0374] The above method can be used in the gradient value calculation process at each position in the 4×4 sub-block required for prediction signal correction in the affine prediction mode, wherein the affine prediction mode is used to derive a motion vector in units of sub-blocks by deriving an affine controlled motion vector of a target block based on an affine model.

[0375] That is, if a motion vector derived in sub-block units by deriving an affine-controlled motion vector of a target block based on an affine model indicates a sub-pixel position outside a target block (4×4 sub-block) in a reference picture, the motion vector can be rounded to the nearest integer position without performing separate interpolation processing, and then gradient calculation can be performed using the pixel value of the rounded integer pixel position.

[0376] For example, pixel values ​​outside the target block (4×4 sub-block) may be replaced by pixel values ​​at positions indicated by (xIntL+(xFracL>>3)-1), (yIntL+(yFracL>>3)-1) in the reference picture, thereby performing gradient calculation. Here, (xIntL, yIntL) may represent an integer-pixel-unit position of a motion vector, and (xFracL, yFracL) may represent a sub-pixel-unit position of a motion vector.

[0377] In order to generate pixel values ​​outside the target block area without increasing the memory bandwidth of the reference picture, the pixel values ​​outside the block area can be generated using an existing 8-tap interpolation filter after filling with the pixel values ​​of the integer pixel positions in the reference picture closest to the encoding position indicated by the motion vector.

[0378] In order to reduce the memory bandwidth of the reference picture, such as Fig. 9 As shown in (b), the gradient value can be calculated only at the internal position of the 4×4 block area, and the gradient value is used to calculate the BIO offset.

[0379] For example, in the case of a 4×4 block, gradients may be calculated only at (1,1), (2,1), (1,2), and (2,2) which are internal positions of the block, and the gradients may be used to calculate the BIO offset. Fig. 9 (c) shows an embodiment in which gradients are obtained only at internal positions of an 8×8 block.

[0380] Additionally, for locations where gradients are not computed, the gradients computed inside the block can be copied and used. Fig. 9 (d) shows an embodiment of copying and using the gradient calculated inside the block. Fig. 9 As shown in (d), for a position where the gradient is not calculated, the gradient value of the adjacent position can be used as the gradient value of the corresponding position.

[0381] As another example for reducing the memory bandwidth of the reference picture, as shown in Equation 20, the gradient at each pixel position in the block may be calculated using only the available pixel values ​​in the block area.

[0382] Equation 20

[0383]

[0384] In the above equation 20, (b) and (c) are equations for obtaining the gradient when the pixel value outside the block boundary is not available at the block boundary position, and (a) is an equation for obtaining the gradient when the left / right or upper / lower neighboring pixels in the block area are available. That is, different gradient calculation equations can be applied based on the gradient calculation position in the block.

[0385] exist Fig. 9 In the embodiment shown in (a) of FIG. 1 , when gradients are obtained at all pixel positions in units of 4×4 sub-blocks for a 4×4 block, for example, if the pixel value at the position (-1,0) is unavailable, only the pixel values ​​at the positions (0,0) and (1,0) may be used to calculate the horizontal gradient (G ) at the upper left position (0,0). 0,0 In addition, if the pixel value at the (4,0) position is not available, the horizontal gradient (G) at the upper right position (3,0) can be calculated using only the pixel values ​​at the (3,0) and (2,0) positions. 3,0 In addition, if the pixel value at the position (0, -1) is not available, the vertical gradient (G) at the upper left position (0, 0) can be calculated using only the pixel values ​​at the positions (0, 0) and (0, 1). 0,0 In addition, if the pixel value at the position (3, -1) is not available, the vertical gradient (G) at the upper right position (3, 0) can be calculated using only the pixel values ​​at the positions (3, 0) and (3, 1). 3,0 ).

[0386] In the above equation 20, (b) and (c) have the same effect as calculating the gradient values ​​at all positions in the block using equation 20(a) after padding the pixel values ​​at positions outside the 4×4 block boundary with the nearby inner boundary pixel values ​​of the block. For example, if the pixel value at the (-1,0) position is not available, then after padding the pixel value at the (0,0) position to the (-1,0) position, the horizontal gradient (G) at the upper left (0,0) position can be calculated by equation 20(a) using the pixel values ​​at the (-1,0) and (1,0) positions. 0,0 ). In addition, if the pixel value at the (4,0) position is not available, after padding the pixel value at the (3,0) position to the (4,0) position, the horizontal gradient (G) at the upper right (3,0) position can be calculated using Equation 20(a) using the pixel values ​​at the (4,0) and (2,0) positions. 3,0 ). In addition, if the pixel value at the position (0, -1) is not available, after padding the pixel value at the position (0, 0) to the position (0, -1), the vertical gradient (G) at the upper left position (0, 0) can be calculated using the pixel values ​​at the positions (0, -1) and (0, 1) by using Equation 20(a). 0,0 ). In addition, if the pixel value at the position (3, -1) is not available, after padding the pixel value at the position (3, 0) to the position (3, -1), the vertical gradient (G) at the upper right position (3, 0) can be calculated using the pixel values ​​at the positions (3, -1) and (3, 1) by using Equation 20(a). 3,0 ). For example, the gradient values ​​in the BIO can be derived by filling the unavailable pixels outside the block boundary with the inner boundary pixels of the block.

[0387] As another example for reducing the memory bandwidth of a reference picture, the gradient at each pixel position in the block may be calculated by Equation 21 using only the available pixel values ​​in the block.

[0388] Equation 21

[0389]

[0390] In the above equation 21, (b) and (c) are equations for obtaining the gradient when the pixel value outside the block boundary is not available at the block boundary position, and (a) is an equation for obtaining the gradient when the left / right or upper / lower neighboring pixels in the block are available. That is, different gradient calculation equations may be applied based on the gradient calculation position in the block. For example, Fig. 9 As shown in , the gradient values ​​in the BIO can be derived by filling the unavailable pixels outside the block boundary with the inner boundary pixels of the block.

[0391] exist Fig. 9In the embodiment shown in (a) of FIG. 1 , when gradients are obtained at all pixel positions in units of 4×4 sub-blocks for a 4×4 block, for example, if the pixel value at the position (-1,0) is unavailable, only the pixel values ​​at the positions (0,0) and (1,0) may be used to calculate the horizontal gradient (G ) at the upper left position (0,0). 0,0 In addition, if the pixel value at the (4,0) position is not available, the horizontal gradient (G) at the upper right position (3,0) can be calculated using only the pixel values ​​at the (3,0) and (2,0) positions. 3,0 In addition, if the pixel value at the position (0, -1) is not available, the vertical gradient (G) at the upper left position (0, 0) can be calculated using only the pixel values ​​at the positions (0, 0) and (0, 1). 0,0 In addition, if the pixel value at the position (3, -1) is not available, the vertical gradient (G) at the upper left position (3, 0) can be calculated using only the pixel values ​​at the positions (3, 0) and (3, 1). 3,0 ). In the above equation 21, (b) and (c) have the same effect as calculating the gradient values ​​at all positions in the block using equation 21(a) after filling the unavailable pixel values ​​at positions outside the 4×4 block boundary with values ​​calculated from the inner boundary pixel values ​​of the block as shown in equation 22. For example, if the pixel value at the (-1,0) position is unavailable, the horizontal gradient (G) at the upper left (0,0) position can be calculated using the pixel values ​​at the (-1,0) and (1,0) positions after filling the pixel value at the (-1,0) position with a value calculated from equation 22(a) using the pixel value at the (0,0) position and the pixel value at the (1,0) position. 0,0 ). In addition, if the pixel value at the (4,0) position is not available, the horizontal gradient (G) at the upper right (3,0) position can be calculated using the pixel values ​​at the (4,0) and (2,0) positions after filling the pixel value at the (4,0) position with the value calculated from equation 22(b) using the pixel value at the (3,0) position and the pixel value at the (2,0) position. 3,0 ). In addition, if the pixel value at the (0, -1) position is not available, the vertical gradient (G) at the upper left (0, 0) position can be calculated using the pixel values ​​at the (0, -1) and (0, 1) positions after filling the pixel value at the (0, -1) position with the value calculated from equation 22(c) using the pixel value at the (0, 0) position and the pixel value at the (0, 1) position. 0,0 ). In addition, if the pixel value at the position (3, -1) is not available, the vertical gradient (G) at the upper right position (3, 0) can be calculated using the pixel values ​​at the positions (3, -1) and (3, 1) after filling the pixel value at the position (3, -1) with the value calculated from equation 22(d) using the pixel value at the position (3, 0) and the pixel value at the position (3, 1). 3,0). For example, the gradient values ​​in the BIO can be derived by filling the unavailable pixels outside the block boundary with the inner boundary pixels of the block.

[0392] Equation 22

[0393] p[x-1]=(p[x]<<1-p[x+1]), x=0, (a)

[0394] p[x+1]=(p[x]<<1-p[x-1]), x=W-1, (b)

[0395] p[y-1]=(p[y]<<1-p[y+1]), y=0, (c)

[0396] p[y+1]=(p[y]<<1-p[y-1]), y=H-1, (d)

[0397] The motion correction vector V used to calculate the BIO offset of the current block may be calculated in units of pixels or one or more subgroups. x and V y .

[0398] In the calculation in subgroup units, the size of the subgroup may be determined by the ratio of the horizontal size to the vertical size of the current target block, or information about the size of the subgroup may be entropy encoded / decoded. In addition, a subgroup unit having a predefined fixed size may be used according to the size and / or shape of the current block.

[0399] Fig.10 is a diagram illustrating various embodiments of subsets as units for calculating BIO offsets.

[0400] For example, Fig.10 As shown in (a), if the size of the current target block is 16×16, V can be calculated in units of 4×4 subgroups. x and V y .

[0401] For example, Fig.10 As shown in (b), if the size of the current target block is 8×16, V can be calculated in units of 2×4 subgroups. x and V y .

[0402] For example, Fig.10 As shown in (c), if the size of the current target block is 16×8, V can be calculated in units of 4×2 subgroups. x and V y .

[0403] For example, Fig.10As shown in (d) of FIG. 1 , if the size of the current target block is 8×16, V can be calculated in units of one 8×8 subgroup, two 4×4 subgroups, and eight 2×2 subgroups. x and V y .

[0404] Optionally, the horizontal size and vertical size of the current target block can be used to derive V x and V y The size of the subgroup unit may be defined by at least one of a minimum depth information value of a predetermined depth information value or a predefined minimum subgroup size. The minimum depth information value may be entropy encoded and transmitted.

[0405] For example, when the size of the current target block is 64×64, the minimum depth information value is 3, and the predefined minimum sub-group size is 4, the size of the sub-group unit may be determined to be 8×8 by the following Equation 23.

[0406] Equation 23

[0407] max(average length (horizontal, vertical) >> minimum depth information value, predefined minimum subgroup size)

[0408] For example, when the size of the current target block is 128×64, the minimum depth information value is 3, and the predefined minimum sub-group size is 4, the size of the sub-group unit may be determined to be 8×8 by the following Equation 24.

[0409] Equation 24

[0410] max(min(horizontal, vertical) >> minimum depth information value, predefined minimum subgroup size)

[0411] When a BIO in a sub-group unit is applied to the current target block, after determining whether to apply deblocking filtering in a sub-group unit, deblocking filtering may be performed on the current target block.

[0412] For example, Fig.10 As shown in (b), when the size of the sub-group unit is 2×4, deblocking filtering can be performed after determining whether to apply the deblocking filter to the boundaries between sub-groups whose horizontal length (width) is greater than 4 and the boundaries between sub-groups whose vertical length (height) is greater than 4 in the current target block.

[0413] When a BIO in sub-group units is applied to the current target block, transformation and inverse transformation may be performed in sub-group units.

[0414] For example, Fig.10 As shown in (a) of FIG. 5 , when the size of the sub-group unit is 4×4, the transform and the inverse transform can be performed in units of 4×4.

[0415] S can be calculated from the subgroup group The motion correction vector V of the subgroup unit is calculated x and V y .

[0416] Without extending the current block, the pixel value and gradient value (G) at each pixel position in the block area can be used according to Equation 8. x ,G y ) calculated at each pixel position, and S representing the BIO-related parameters of the subgroup unit was calculated. group .

[0417] For example, as shown in the following Equation 25, when the size of the subgroup of the current block is 4×4, the motion correction vector of the subgroup unit may be calculated by summing the BIO-related parameters at the respective pixel positions calculated from Equation 8 using only the pixel value and the gradient value at each pixel position without expanding the block area. As shown in Equation 8, S represents a pixel-unit BIO-related parameter calculated from the pixel value and the gradient value at each pixel position.

[0418] Equation 25

[0419]

[0420] Fig.11 It is shown that in order to obtain the subgroup S group And a diagram of the weights that can be applied to the S values ​​in a sub-group.

[0421] In obtaining the subgroup unit S group In the processing of values, such as Fig.11 As shown in (a) of FIG. 1 , a value obtained by summing the respective S values ​​in the subgroup to which the same weight is applied may be used as the S of the subgroup. group value.

[0422] like Fig. 9 As shown in (b) of FIG. 1 , when the gradient is calculated only at the internal position of the sub-group, the S value calculated by weighted summing the S values ​​calculated using the corresponding gradients can be used as the S of the sub-group. group value.

[0423] like Fig. 9 As shown in (d) of FIG. 1 , when the gradient is calculated only at the inner position of the sub-group and the calculated gradient is copied and used as the gradient at the outer position, the S value calculated by weighted summing the S values ​​calculated using only the gradient values ​​at the inner position can be used as the S of the sub-group. group value.

[0424] like Fig. 9As shown in (d) of FIG. 1 , when the gradient is calculated only at the inner position of the subgroup and the calculated gradient is copied and used as the gradient at the outer position, the S value obtained by weighted summing the S values ​​calculated from the gradient values ​​at all positions of the subgroup can be used as the S of the subgroup. group value.

[0425] In obtaining the subgroup unit S group In the processing of values, such as Fig.11 As shown in (b) of FIG. 1 , a value obtained by summing the respective S values ​​in the subgroup to which different weights are applied can be used as the S of the subgroup. group value.

[0426] Optionally, the S value at a specific position in a subgroup may be used as the subgroup S group For example, Fig.11 As shown in (c), when the size of the subgroup of the current block is 4×4, S 10 The S value at the position can be used as the S of the subgroup group Value. Information about a specific position may be predetermined in an encoder / decoder, or may be signaled via a bitstream, or may be derived based on encoding parameters (size, shape, etc.) of the current block.

[0427] Fig.12 It is shown that in order to obtain the subgroup S group A diagram of an embodiment in which weighted summation is performed only on S values ​​at specific positions in a subgroup.

[0428] like Fig.12 As shown in (a) to (d) of FIG. 1 , a value obtained by weighted summing only the S values ​​at specific positions in the subgroup can be used as S group .

[0429] As shown in Equation 7, the S of the subgroup unit obtained by the above method can be used group The motion correction vector V of the subgroup unit is obtained by x and V y The gradient value at each pixel position in the subgroup and the motion correction vector V can be used x and V y To calculate the value of each pixel position in the subgroup and the value in Equation 6 The corresponding BIO offset value. In the calculation of the BIO offset value, for the pixel position where the gradient is not calculated, such as Fig. 9 As shown in (d), the gradient values ​​calculated within the block can be copied and used for calculation.

[0430] The BIO offset may be calculated in subgroup units using the motion correction vector derived in subgroup units, the representative value of the gradient value at each pixel position in the subgroup, and the representative value of the pixel value in the subgroup, and the same BIO offset may be applied to each pixel position in the subgroup. The representative value of the gradient value in the subgroup may represent at least one of a minimum value, a maximum value, an average value, a weighted average value, a most frequent value, an interpolation value, or a median value of the gradient value.

[0431] The representative value of the pixel values ​​in the subgroup may represent at least one of a minimum value, a maximum value, an average value, a weighted average value, a most frequent value, an interpolated value, or a median value of the pixel values.

[0432] A representative value of the BIO offset values ​​obtained at each pixel position in the subgroup may be obtained, and the same BIO offset value may be applied to each pixel position in the subgroup. The representative value may represent at least one of a minimum value, a maximum value, an average value, a weighted average value, a most frequent value, an interpolated value, or a median value of the BIO offset value.

[0433] The motion correction vector V calculated in units of pixels in the subgroup can be used x and V y To calculate the motion correction vector V of the subgroup unit x and V y .

[0434] For example, when the size of the subgroup is 2×2, V of the subgroup may be calculated as shown in the following Equation 26: x and V y .

[0435] Equation 26

[0436] V x =average(V x0 ,V x1 ,V x2 ,V x3 ),V y =average(V y0 ,V y1 ,V y2 ,V y3 )

[0437] V x =min(V x0 ,V x1 ,V x2 ,V x3 ),V y =min(V y0 ,V y1 ,V y2 ,V y3 )

[0438] Vy =max(V x0 ,V x1 ,V x2 ,V x3 ),V y =max(V y0 ,V y1 ,V y2 ,V y3 )

[0439] The V of the derived sub-group unit can be determined based on the size of the current block x and V y The sub-group unit may be determined based on a comparison between the size of the current block and a predetermined threshold. The predetermined threshold may represent a value for determining the derivation V x and V y The predetermined threshold may be a fixed value predetermined in the encoder / decoder, may be variably derived based on coding parameters of the current block (e.g., motion vector size, etc.), or may be signaled via a bitstream (e.g., sequence, picture, slice, block level, etc.).

[0440] For example, V may be calculated in units of sub-blocks for blocks whose horizontal length and vertical length have a product equal to or greater than 256. x and V y , and V can be calculated in pixels for other blocks x and V y .

[0441] For example, V may be calculated in units of sub-blocks for blocks whose minimum length among horizontal length and vertical length is equal to or greater than 8. x and V y , and V can be calculated in pixels for other blocks x and V y .

[0442] V of subgroup unit x and V y Can be used to deduce V only by comparing the gradient value in the horizontal direction with the gradient value in the vertical direction x or V y To calculate the BIO offset.

[0443] For example, when the sum of the absolute values ​​of the horizontal gradient value for the L0 reference prediction block and the horizontal gradient value for the L1 reference prediction block is greater than the sum of the absolute values ​​of the vertical gradient value for the L0 reference prediction block and the vertical gradient value for the L1 reference prediction block, only V x can be calculated and used to calculate the BIO offset. In this case, V y The value can represent 0.

[0444] For example, when the sum of the absolute values ​​of the vertical gradient value for the L0 reference prediction block and the vertical gradient value for the L1 reference prediction block is greater than the sum of the absolute values ​​of the horizontal gradient value for the L0 reference prediction block and the horizontal gradient value for the L1 reference prediction block, only V y can be calculated and used to calculate the BIO offset. In this case, V x The value can represent 0.

[0445] The gradient values ​​of neighboring pixel positions for the current block may be considered together, and the sub-group unit S may be calculated from the S value calculated in pixel units. group value.

[0446] Fig.13 is a diagram showing an example of calculating an S value.

[0447] The S value at the upper left position (0,0) in the current block can be calculated by applying a 5×5 window to the corresponding position and considering the gradient values ​​of the neighboring pixel positions together with the gradient value at the current position. Fig.13 As shown in (a) of FIG. 1 , the gradient value in the current block may be used to calculate the gradient value at the position corresponding to the outside of the current block, or the gradient value at the position corresponding to the outside of the current block may be directly calculated and used. The S value at another position in the current block may be calculated equivalently.

[0448] The subgroup unit S in the current block can be calculated by applying different weights according to the position group At this time, without expanding the block, only the gradient value in the current block can be used.

[0449] For example, Fig.13 As shown in (b), when the size of the subgroup is 2×2 and a 5×5 window is applied to each pixel position, the S of the subgroup can be calculated by applying a 6×6 weight table to the S value calculated by considering the gradient values ​​of the neighboring pixel positions. group value.

[0450] For example, Fig.13 As shown in (c), when the size of the subgroup is 4×4 and a 5×5 window is applied to each pixel position, the S of the subgroup can be calculated by applying an 8×8 weight table to the S value calculated by considering the gradient values ​​of the neighboring pixel positions. group value.

[0451] For example, Fig.13 As shown in (d) of FIG. 1 , when the size of the subgroup is 8×8 and a 5×5 window is applied to each pixel position, the S of the subgroup can be calculated by applying a 12×12 weight table to the S value calculated by considering the gradient values ​​of the neighboring pixel positions. group value.

[0452] Fig.14 It shows that when the size of the subgroup is 4×4, S is calculated. group Schematic diagram of an embodiment of the present invention. Fig.14 (a) shows the gradients at a pixel position and neighboring pixel positions in a 4×4 block. Fig.14 (b) shows the S values ​​at the pixel position and the neighboring pixel positions in the 4×4 block.

[0453] For example, when the size of the subgroup is 4×4, such as Fig.14 As shown in , S can be calculated by summing all S values ​​at positions obtained from gradients obtained from neighboring pixel positions and pixel positions of the current block area. group For example, the following equation 27 may be used. At this time, the weights at the respective positions may be equal to a predetermined value (e.g., 1), or different weights may be applied. When the S value at the adjacent pixel position is not available, S may be calculated by adding the available adjacent S value to the S value in the current subgroup. group When a neighboring pixel position is outside the boundary of a block region and is unavailable, the neighboring pixel position may be used after being filled with S values ​​near the block region boundary. When a neighboring pixel position is outside the boundary of a block region and the gradient and pixel values ​​are unavailable, the S value at the position may be calculated by filling in the gradient values ​​and pixel values ​​near the block region boundary.

[0454] Equation 27

[0455]

[0456] In the above embodiment, the size of the weight table and the weight may vary according to the size of the M×N window applied to each pixel position. M and N are natural numbers greater than 0, and M and N may be equal to or different from each other.

[0457] By calculating V in subgroup units x and V y Applied to the first motion information and the second motion information, the motion information of the current block may be updated and stored in units of sub-groups, and then the motion information of the current block may be used for the next target block. When the motion vector is updated in units of sub-groups, only the motion correction vector V of the predefined sub-group position may be used. x and V y To update the motion information of the current block. Fig.10 As shown in (a), when the size of the target block is 16×16 and the size of the sub-group is 4×4, the motion vector obtained by applying only the motion correction vector of the first sub-group on the upper left side to the first motion vector and the second motion vector of the current block can be stored as the motion vector of the current block.

[0458] In the following, the derivation of the motion vector for the chrominance component will be described.

[0459] According to one embodiment, a motion correction vector (V) calculated from the luma component in units of subgroups may be used in a motion compensation process for the chroma component. x ,V y ) is applied to the motion vector of the chrominance component.

[0460] Alternatively, a motion vector obtained by applying the motion correction vectors of the subset of predefined relative positions to the first motion vector and the second motion vector of the current block may be used as the motion vector of the chroma component.

[0461] Fig.15 is a diagram illustrating an embodiment of deriving a motion vector of a chrominance component based on a luminance component.

[0462] like Fig.15 As shown in , when the size of the current target block is 8×8 and the size of the sub-group is 4×4, the motion vector of the chroma block can be obtained by converting the motion correction vector V of the sub-block ④ into x and V y The first motion vector MV applied to the current target block 0x , MV 0y and the second motion vector MV 1x , MV 1y The obtained motion vector, that is, the motion vector of the chrominance component can be derived as shown in the following equation 28.

[0463] Equation 28

[0464] First motion vector = (MV 0x +V x ,MV 0y +V y ),

[0465] The second motion vector = (MV 1x +V x ,MV 1y +V y )

[0466] As a motion correction vector for calculating the motion vector of the chroma block, the motion correction vector of sub-block ④ may be replaced by the motion correction vector of another sub-block. Alternatively, at least one of the maximum value, minimum value, median value, average value, weighted average value, or most frequent value of the motion correction vectors of two or more sub-blocks among sub-blocks ① to ④ may be used.

[0467] The motion correction vector V calculated from the luma component in subgroup units can be used in the motion compensation process for the chroma component. x and Vy The motion vector obtained by applying it to the chrominance components Cb and Cr.

[0468] Fig.16 is a diagram showing an example of motion compensation processing for chroma components.

[0469] like Fig.16 As shown in , when the size of the sub-block unit of the luminance block is 4×4, each of the corresponding chrominance blocks Cb and Cr has a sub-group of size 2×2, and the motion correction vector of each sub-group in the chrominance block can use the motion correction vector of the sub-group of the corresponding luminance block.

[0470] For example, the motion correction vectors V for the first subset of chrominance blocks Cb and Cr are cx1 and V cy1 The motion correction vectors V corresponding to the first subset of luminance blocks can be used x1 and V y1 .

[0471] Using the motion vector value obtained in units of subgroups of the chrominance components Cb and Cr and the pixel values ​​of the reconstructed chrominance components Cb and Cr, similar to the luma component, the BIO offset can be calculated in units of subgroups of the chrominance components Cb and Cr. At this time, the following equation 29 can be used.

[0472] Equation 29

[0473]

[0474] ΔG cx =G cx0 -G cx1 , ΔG cy =G cy0 -G cy1

[0475] In the above equation, ΔG can be obtained from the reconstructed pixels of the reference pixels of the chrominance components Cb and Cr. cx ,ΔG cy .

[0476] For example, it can be obtained as follows Fig.16 The pixel value positions G at the second subgroup of chrominance components shown in cx ,G cy .

[0477] The G of the x component at the P2 position can be calculated by the difference between the pixel value at the P1 position and the pixel value at the P3 position. c .

[0478] The G of the x component at the P3 position can be calculated by the difference between the pixel value at the P2 position and the pixel value at the P3 position. c .

[0479] The G of the y component at the P2 position can be calculated by the difference between the pixel value at the P6 position and the pixel value at the P2 position. c .

[0480] The pixel value at position P2 can be compared with P 10 The difference between the pixel values ​​at the position is used to calculate the G of the y component at the position P6 c .

[0481] The encoder may determine whether to perform a BIO for the current block, and then encode (e.g., entropy encoding) information indicating whether the BIO is performed. Whether the BIO is performed may be determined by comparing a distortion value between a prediction value before applying the BIO and a prediction signal after applying the BIO. The decoder may decode (e.g., entropy decoding) information indicating whether the BIO is performed from the bitstream, and perform the BIO based on the received information. In addition, information indicating whether the BIO is activated may be signaled at a higher level (sequence, picture, slice, CTU, etc.). For example, information indicating whether the BIO is performed may be determined only when the information indicating whether the BIO is activated indicates that the BIO is activated.

[0482] The information indicating whether the BIO is performed may be entropy encoded / decoded based on the encoding parameters of the current block.

[0483] Optionally, encoding / decoding of information indicating whether BIO is performed may be omitted based on encoding parameters of the current block. That is, the information indicating whether BIO is performed may be determined based on encoding parameters of the current block. Here, the encoding parameters may include at least one of a prediction mode in a decoder, motion compensation accuracy, size and shape of the current block, a partition form (quadtree partition, binary tree partition, or ternary tree partition), a picture type, a global motion compensation mode, or a motion correction mode.

[0484] For example, the encoder / decoder may determine the accuracy of motion compensation using a prediction signal generated by performing motion compensation based on the first motion information of the current block and a prediction signal generated by performing motion compensation based on the second motion information. For example, the encoder / decoder may determine the accuracy of motion compensation based on a difference signal between the two prediction signals and based on a comparison between the difference signal and a predetermined threshold. The difference signal may represent a SAD value between the two prediction signals.

[0485] The predetermined threshold may represent a reference value for determining the accuracy of the difference signal to determine whether BIO is performed. This may be expressed in the form of at least one of a minimum value or a maximum value. The predetermined threshold may be a fixed value predetermined in the encoder / decoder, may be determined by encoding parameters such as the size, shape, and bit depth of the current block, and may be signaled at the SPS, PPS, slice header, tile, CTU, or CU level.

[0486] In addition, the encoder / decoder may determine whether BIO is performed in units of sub-blocks of the current block. The encoder / decoder may determine whether BIO is performed in units of sub-blocks based on the comparison between the predetermined threshold and the difference signal between the two prediction signals corresponding to the sub-blocks in each sub-block unit. The predetermined threshold used in the sub-block unit may be equal to or different from the threshold used in the block unit. This can be expressed in the form of at least one of a minimum value or a maximum value. The predetermined threshold may be a fixed value predetermined in the encoder / decoder, may be determined by encoding parameters such as the size, shape, and bit depth of the current block, and may be signaled at the SPS, PPS, slice header, parallel block, CTU, or CU level. For example, when the SAD of a sub-block of the current block is less than a threshold determined based on the size of the current block (e.g., 2×the horizontal length (width) of the sub-block×the vertical length (height) of the sub-block), BIO may not be applied to the current sub-block.

[0487] For example, when the current block is in the merge mode, the encoder / decoder may always apply the BIO without entropy encoding / decoding information indicating whether the BIO is performed.

[0488] For example, when the current block is in the AMVP mode, the encoder / decoder may entropy encode / decode information indicating whether the BIO is performed and perform the BIO according to the information.

[0489] For example, when the current block is in the AMVP mode, the encoder / decoder may always apply the BIO without entropy encoding / decoding information indicating whether the BIO is performed.

[0490] For example, when the current block is in the AMVP mode, the encoder / decoder may not always apply the BIO without entropy encoding / decoding information indicating whether the BIO is performed.

[0491] For example, when the current block is in the merge mode, the encoder / decoder may entropy encode / decode information indicating whether BIO is performed, and perform BIO according to the corresponding information.

[0492] For example, when the current block is in the AMVP mode and 1 / 4 pixel (one quarter) unit motion compensation is performed, the encoder / decoder may always apply BIO without entropy encoding / decoding the information indicating whether BIO is performed. In addition, when the current block is in the AMVP mode and integer pixel (one pixel or four pixels) unit motion compensation is performed, the encoder / decoder may entropy encode / decode the information indicating whether BIO is performed, and perform BIO according to the corresponding information.

[0493] For example, when the current block is in the AMVP mode and 1 / 4 pixel (one quarter) unit motion compensation is performed, the encoder / decoder may not always apply BIO without entropy encoding / decoding the information indicating whether BIO is performed. In addition, when the current block is in the AMVP mode and integer pixel (one pixel or four pixels) unit motion compensation is performed, the encoder / decoder may not always apply BIO without entropy encoding / decoding the information indicating whether BIO is performed.

[0494] For example, when the current block is in the AMVP mode and integer pixel (one pixel or four pixels) unit motion compensation is performed, the encoder / decoder may always apply BIO without entropy encoding / decoding the information indicating whether BIO is performed. In addition, when 1 / 4 pixel (one quarter) unit motion compensation is performed, the encoder / decoder may entropy encode / decode the information indicating whether BIO is performed, and perform BIO according to the corresponding information.

[0495] For example, when the current block is in AMVP mode and has a size less than or equal to 256 luma pixels, the encoder / decoder may entropy encode / decode information indicating whether BIO is performed and perform BIO according to the corresponding information. When the condition is not met, BIO may always be performed.

[0496] For example, when the current block is in the inter-intra combined prediction mode, the encoder / decoder may not always perform BIO.

[0497] For example, when the current block is in the motion prediction mode based on the affine motion model, the encoder / decoder may not always perform BIO. The motion prediction mode based on the affine motion model may correspond to the case where the encoding parameter MotionModelIdc value is a non-zero value.

[0498] For example, when the current block is in the symmetric motion vector difference mode, the encoder / decoder may not always perform BIO. The symmetric motion vector difference mode may mean a mode in which, without entropy encoding / decoding the motion vector difference in the L1 direction, values ​​-MVD0x and -MVD0y obtained by mirroring the horizontal component value MVD0x and the vertical component value MVD0y of the motion vector difference in the L0 direction in the L1 direction are used as the motion vector difference in the L1 direction.

[0499] For example, when the size of the current block is equal to or smaller than a predefined size, the encoder / decoder may not perform BIO.

[0500] For example, when the vertical length (height) of the current block is 4, the encoder / decoder may not perform BIO.

[0501] For example, when the horizontal length (width) of the current block is 4 and the vertical length (height) of the current block is 8, the encoder / decoder may not perform BIO.

[0502] For example, when the vertical length (height) of the current block is less than 8, the encoder / decoder may not perform BIO.

[0503] For example, when the horizontal length (width) of the current block is less than 8, the encoder / decoder may not perform BIO.

[0504] For example, when the area of ​​the current block is less than 128, the encoder / decoder may not perform BIO.

[0505] For example, when the size of the current block is less than or equal to the predefined size and binary tree partitioning is performed, the encoder / decoder may not perform BIO.

[0506] For example, when the size of the current block is less than or equal to the predefined size and ternary tree partitioning is performed, the encoder / decoder may not perform BIO.

[0507] For example, the encoder / decoder may not perform BIO when the current block is in illumination compensation mode, affine mode, sub-block merging mode, a mode for correcting motion information in the decoder (e.g., PMMVD (pattern matching motion vector derivation), DMVR (decoder-side motion vector revision)), or a current picture reference (CPR) mode for performing inter-frame prediction by referring to a current image including the current block or reconstructed pixels in a CTU including the current block.

[0508] The encoder / decoder may determine whether BIO is performed based on a reference picture of the current block.

[0509] For example, when all reference pictures of the current block are short-term reference pictures, the encoder / decoder may perform BIO. Conversely, when at least one of the reference pictures of the current block is not a short-term reference picture, the encoder / decoder may not perform BIO.

[0510] Whether to apply the BIO to the current target block may be determined according to flag information entropy decoded in at least one of a CTU unit or a CTU sub-unit. At this time, the sub-unit may include at least one of a CTU sub-unit, a CU unit, or a PU unit.

[0511] For example, when the CTU block size is 128×128 and information about BIO is entropy decoded in units of 32×32 blocks as CTU sub-units, the encoder / decoder may perform BIO based on the information about BIO entropy decoded in units of 32×32 blocks for blocks belonging to the 32×32 blocks and having a size smaller than the 32×32 block unit.

[0512] For example, when the block depth of a CTU is 0 and information about a BIO is entropy decoded in sub-units of a CTU having a block depth of 1, the encoder / decoder may perform BIO based on the information about the BIO entropy decoded in sub-units of a CTU having a block depth of 1 for a block included in a sub-unit of the CTU and having a block depth of 1 or more.

[0513] The predicted sample signal P obtained via BIO can be used optical flow and the prediction signal P obtained through the existing bidirectional prediction conventional bi-prediction The final prediction sample signal of the current block is generated by the weighted sum of . At this time, the following equation 30 can be used.

[0514] Equation 30

[0515] P=(1-σ)P conventional bi-prediction +σP optical flow

[0516] In the above equation, the weights σ or 1-σ applied to the block may be equal to each other and may be determined differently according to the encoding parameters of the current block. The encoding parameters may include a prediction mode, motion compensation accuracy, the size and shape of the current block, a partition form (quadtree partition, binary tree partition, or ternary tree partition), a global motion compensation mode, a motion correction mode in a decoder, or at least one of a layer of a current picture to which the current block belongs.

[0517] For example, the weight σ may vary depending on whether the current block is in merge mode or AMVP mode.

[0518] For example, when the current block is in the AMVP mode, the weight σ may vary according to whether there is a ¼ pixel (quarter) unit motion vector difference MVD or an integer unit motion vector difference MVD.

[0519] For example, when the current block is in the merge mode, the weight σ may vary according to an affine mode, an illumination compensation mode, a mode for correcting motion information in a decoder (eg, PMMVD or DMVR).

[0520] For example, the weight σ may vary according to the size and / or shape of the current block.

[0521] For example, the weight σ may vary according to the temporal layer of the current picture to which the current block belongs.

[0522] For example, when BIO is applied to the current block in units of subgroups, the weight σ may vary in units of subgroups.

[0523] Fig.21 is a flowchart illustrating an image decoding method according to an embodiment of the present invention.

[0524] Reference Fig.21 , the decoder may determine whether the current block is in the bidirectional optical flow mode based on a distance between a first reference picture of the current block and the current picture and a distance between a second reference picture of the current block and the current picture ( S2110 ).

[0525] For example, when the distance between the first reference picture and the current picture and the distance between the second reference picture and the current picture are not the same, the decoder may determine that the current block is not in the bidirectional optical flow mode.

[0526] In addition, the decoder may determine whether the current block is in the bidirectional optical flow mode based on the type of the reference picture of the current block.

[0527] For example, when at least one of the type of the first reference picture of the current block or the type of the second reference picture of the current block is not a short-term reference picture, the decoder may determine that the current block is not in the bidirectional optical flow mode.

[0528] Additionally, the decoder may determine whether the current block is in the bidirectional optical flow mode based on the size of the current block.

[0529] In addition, when the current block is in the bidirectional optical flow mode (S2110-Yes), the decoder may calculate the gradient information of the prediction sample of the current block. Specifically, the decoder may calculate the gradient information using at least one neighboring sample adjacent to the prediction sample. At this time, when the neighboring sample is located outside the area of ​​the current block, the sample value of the integer pixel position closest to the neighboring sample may be used as the value of the neighboring sample.

[0530] In addition, the gradient information may be calculated in units of sub-blocks having a predefined size.

[0531] Furthermore, the decoder may generate a prediction block of the current block using the calculated gradient information ( S2130 ).

[0532] To derive the same prediction result in the encoder as in the decoder, the same Fig.21 The image decoding method is equivalent to the image encoding method.

[0533] A bit stream generated by the image encoding method of the present invention may be temporarily stored in a non-transitory computer-readable recording medium, and may be a bit stream encoded by the above-mentioned image encoding method.

[0534] Specifically, a non-transitory computer-readable recording medium storing a bitstream generated by a method for encoding an image. The method for encoding an image may include: determining whether a current block is in a bidirectional optical flow (BIO) mode, calculating gradient information of a prediction sample of the current block when the current block is in the BIO mode, and using the calculated gradient information to generate a prediction block of the current block. The step of calculating the gradient information of the prediction sample of the current block includes: calculating the gradient information using at least one neighboring sample adjacent to the prediction sample.

[0535] The above-described embodiments may be performed in the same manner in an encoder and a decoder.

[0536] At least one embodiment or a combination of the above embodiments may be used to encode / decode a video.

[0537] The order in which the above-described embodiments are applied may be different between an encoder and a decoder, or the order in which the above-described embodiments are applied may be the same in an encoder and a decoder.

[0538] The above-described embodiments may be performed on each luminance signal and each chrominance signal, or may be performed identically on the luminance signal and the chrominance signal.

[0539] The block form to which the above-described embodiment of the present invention is applied may have a square form or a non-square form.

[0540] The above-mentioned embodiments of the present invention may be applied according to the size of at least one of a coding block, a prediction block, a transform block, a block, a current block, a coding unit, a prediction unit, a transform unit, a unit, and a current unit. Here, the size may be defined as a minimum size or a maximum size or both a minimum size and a maximum size so that the above-mentioned embodiments are applied, or the size may be defined as a fixed size to which the above-mentioned embodiments are applied. In addition, in the above-mentioned embodiments, the first embodiment may be applied to a first size, and the second embodiment may be applied to a second size. In other words, the above-mentioned embodiments may be applied in combination according to the size. In addition, the above-mentioned embodiments may be applied when the size is equal to or greater than the minimum size and equal to or less than the maximum size. In other words, the above-mentioned embodiments may be applied when the block size is included in a specific range.

[0541] For example, when the size of the current block is 8×8 or larger, the above embodiment may be applied. For example, when the size of the current block is only 4×4, the above embodiment may be applied. For example, when the size of the current block is 16×16 or smaller, the above embodiment may be applied. For example, when the size of the current block is equal to or larger than 16×16 and equal to or smaller than 64×64, the above embodiment may be applied.

[0542] The above-described embodiments of the present invention may be applied according to time layers. In order to identify the time layers to which the above-described embodiments may be applied, a corresponding identifier may be signaled, and the above-described embodiments may be applied to the specified time layers identified by the corresponding identifier. Here, the identifier may be defined as the lowest layer or the highest layer or both the lowest layer and the highest layer to which the above-described embodiments may be applied, or may be defined as a specific layer to which the embodiments may be applied. In addition, a fixed time layer to which the embodiments may be applied may be defined.

[0543] For example, when the temporal layer of the current image is the lowest layer, the above embodiment may be applied. For example, when the temporal layer identifier of the current image is 1, the above embodiment may be applied. For example, when the temporal layer of the current image is the highest layer, the above embodiment may be applied.

[0544] A slice type or a tile group type to which the above-described embodiments of the present invention are applied may be defined, and the above-described embodiments may be applied according to the corresponding slice type or tile group type.

[0545] In the above embodiments, the method is described based on a flowchart having a series of steps or units, but the present invention is not limited to the order of the steps, but some steps may be performed simultaneously with other steps or in a different order. In addition, it should be understood by those skilled in the art that the steps in the flowchart are not mutually exclusive, and other steps may be added to the flowchart, or some steps may be deleted from the flowchart without affecting the scope of the present invention.

[0546] The embodiments include various aspects of the examples. All possible combinations for various aspects may not be described, but those skilled in the art will be able to recognize different combinations. Therefore, the present invention may include all substitutions, modifications and changes within the scope of the claims.

[0547] Embodiments of the present invention may be implemented in the form of program instructions that can be executed by various computer components and recorded in a computer-readable recording medium. A computer-readable recording medium may include a separate program instruction, a data file, a data structure, etc., or a combination of a program instruction, a data file, a data structure, etc. The program instruction recorded in the computer-readable recording medium may be specifically designed and constructed for the present invention, or may be known to a person of ordinary skill in the field of computer software technology. Examples of computer-readable recording media include: magnetic recording media (such as hard disks, floppy disks, and magnetic tapes) that are specifically constructed to store and implement program instructions, optical data storage media (such as CD-ROMs or DVD-ROMs), magneto-optical media (such as floppy disks), and hardware devices (such as read-only memory (ROM), random access memory (RAM), flash memory, etc.). Examples of program instructions include not only machine language codes formatted by a compiler, but also high-level language codes that can be implemented by a computer using an interpreter. A hardware device may be configured to be operated by one or more software modules to perform processing according to the present invention, or vice versa.

[0548] Although the present invention has been described according to specific items such as detailed elements and limited embodiments and drawings, they are provided only to help a greater understanding of the present invention, and the present invention is not limited to the above embodiments. It will be appreciated by those skilled in the art that various modifications and changes can be made from the above description.

[0549] Therefore, the spirit of the present invention should not be limited to the above-described embodiments, and the entire scope of the claims and their equivalents will fall within the scope and spirit of the present invention.

[0550] Industrial Applicability

[0551] The present invention can be used to encode or decode images.

Claims

1. A method for decoding an image, the method comprising: Build a merge candidate list for the current block; Obtaining a motion vector of a current block from the merge candidates included in the merge candidate list; Determine whether the bidirectional optical flow BIO mode is applied to the current block; When the BIO mode is applied to the current block, a prediction sample of an extended sub-block is obtained based on a motion vector of the current block, wherein the extended sub-block is composed of a sub-block included in the current block and an extended area around the sub-block; Obtaining gradient information of a first prediction sample point in the extended sub-block; deriving a motion refinement vector of the sub-block based on the gradient information of the first prediction sample in the extended sub-block; and obtaining a final prediction sample in the sub-block based on a prediction offset derived from the motion refinement vector, The gradient information of the first prediction sample point is derived by using at least one neighboring sample point adjacent to the first prediction sample point. wherein the BIO mode is applied to the current block only when a first distance representing a picture order count difference between a first reference picture of the current block and the current picture and a second distance representing a picture order count difference between the current picture and a second reference picture of the current block are the same, and Whether the BIO mode is applied to a subblock included in the current block is determined by comparing a difference signal between an L0 prediction subblock of the subblock and an L1 prediction subblock of the subblock with a threshold.

2. The method of claim 1, wherein: The second prediction sample in the extended area is derived from a sample value of an integer pixel position closest to a reference position, wherein the reference position is specified by a motion vector of the current block, and the integer pixel position is determined by rounding the reference position.

3. The method of claim 1, wherein: The threshold is adaptively determined based on the size of the sub-block.

4. A method for encoding an image, the method comprising: Build a merge candidate list for the current block; Obtaining a motion vector of a current block from the merge candidates included in the merge candidate list; Determine whether the bidirectional optical flow BIO mode is applied to the current block; When the BIO mode is applied to the current block, a prediction sample of an extended sub-block is obtained based on a motion vector of the current block, wherein the extended sub-block is composed of a sub-block included in the current block and an extended area around the sub-block; Obtaining gradient information of a first prediction sample point in the extended sub-block; deriving a motion refinement vector of the sub-block based on the gradient information of the first prediction sample in the extended sub-block; and obtaining a final prediction sample in the sub-block based on a prediction offset derived from the motion refinement vector, The gradient information of the first prediction sample point is derived by using at least one neighboring sample point adjacent to the first prediction sample point. wherein the BIO mode is applied to the current block only when a first distance representing a picture order count difference between a first reference picture of the current block and the current picture and a second distance representing a picture order count difference between the current picture and a second reference picture of the current block are the same, and Whether the BIO mode is applied to a subblock included in the current block is determined by comparing a difference signal between an L0 prediction subblock of the subblock and an L1 prediction subblock of the subblock with a threshold.

5. The method of claim 4, wherein: The second prediction sample in the extended area is derived from a sample value of an integer pixel position closest to a reference position, wherein the reference position is specified by a motion vector of the current block, and the integer pixel position is determined by rounding the reference position.

6. The method of claim 4, wherein: The threshold is adaptively determined based on the size of the sub-block.

7. A non-transitory computer-readable recording medium storing a bit stream generated by a method of encoding an image by an encoder, the method comprising: Build a merge candidate list for the current block; Obtaining a motion vector of a current block from the merge candidates included in the merge candidate list; Determine whether the bidirectional optical flow BIO mode is applied to the current block; When the BIO mode is applied to the current block, a prediction sample of an extended sub-block is obtained based on a motion vector of the current block, wherein the extended sub-block is composed of a sub-block included in the current block and an extended area around the sub-block; Obtaining gradient information of a first prediction sample point in the extended sub-block; deriving a motion refinement vector of the sub-block based on the gradient information of the first prediction sample in the extended sub-block; and obtaining a final prediction sample in the sub-block based on a prediction offset derived from the motion refinement vector, The gradient information of the first prediction sample point is derived by using at least one neighboring sample point adjacent to the first prediction sample point. wherein the BIO mode is applied to the current block only when a first distance representing a picture order count difference between a first reference picture of the current block and the current picture and a second distance representing a picture order count difference between the current picture and a second reference picture of the current block are the same, and Whether the BIO mode is applied to a subblock included in the current block is determined by comparing a difference signal between an L0 prediction subblock of the subblock and an L1 prediction subblock of the subblock with a threshold.

Images