Method, apparatus, and recording medium for image encoding / decoding

Abstract

Disclosed herein are a method, apparatus, and storage medium for image encoding / decoding. In a typical image encoding / decoding method, a decoder-side motion information derivation method can be used finitely. Therefore, an increase in coding efficiency attributable to a decoder-side motion information derivation method may also be limited. In an embodiment, a motion information search method for use in an inter prediction mode, an intra block copy mode, and an intra template matching prediction mode is disclosed. By using various motion search methods, coding efficiency can be improved.

Description

Technical Field

[0001] The present disclosure generally relates to methods, devices, and storage media for image encoding / decoding.

[0002] This application claims the benefit of Korean Patent Application No. 10-2022-0044186, filed on April 8, 2022, and No. 10-2023-0046211, filed on April 7, 2023, which are hereby incorporated by reference in their entirety into this application. Background Art

[0003] With the continuous development of the information and communication industry, broadcast services supporting high-definition (HD) resolution have become popular around the world. With this popularity, a large number of users have become accustomed to high-resolution and high-definition images and / or videos.

[0004] To meet user demand for higher definition, numerous organizations have accelerated the development of next-generation imaging devices. In addition to high-definition television (HDTV) and full-high-definition (FHD) television, interest in ultra-high-definition television (UHD) has also grown. UHD television offers a resolution more than four times that of full-high-definition (FHD). This growing interest has led to a demand for image encoding and decoding technologies that achieve higher resolution and definition.

[0005] As image compression technology, there are various technologies such as inter-frame prediction technology, intra-frame prediction technology, transform, quantization technology, and entropy coding technology.

[0006] Inter-frame prediction technology is a technology for predicting the values ​​of pixels included in the current picture using pictures before and / or after the current picture. Intra-frame prediction technology is a technology for predicting the values ​​of pixels included in the current picture using information about the pixels in the current picture. Transformation and quantization technology can be a technology for compressing the energy of the residual signal. Entropy coding technology is a technology for assigning short codewords to frequently occurring values ​​and long codewords to less frequently occurring values.

[0007] By utilizing these image compression technologies, data regarding images can be efficiently compressed, transmitted, and stored. Summary of the Invention

[0008] Technical issues

[0009] The embodiments are directed to providing an apparatus, method, and storage medium for performing encoding / decoding on a target block in an inter prediction mode, an intra block copy mode, and an intra template matching prediction mode.

[0010] Technical Solution

[0011] According to one aspect, there is provided an image encoding method including: determining prediction information to be used for encoding of a target block; generating encoded encoding information by performing encoding on the encoding information; and generating a bitstream including the encoded encoding information.

[0012] The prediction information may include a motion information candidate list and final motion information.

[0013] The image encoding method may further include performing prediction on the target block using inter prediction, intra block copying, or intra template matching.

[0014] A motion vector search method for prediction may be used.

[0015] The motion vector search method may be template matching.

[0016] Inter prediction can be bidirectional prediction.

[0017] The motion vector may be fixed for one of the two directions in bi-directional prediction.

[0018] A motion search may be performed for the remaining one of the two directions.

[0019] A motion vector search method for inter prediction may be used.

[0020] The motion vector search method may be bilateral matching.

[0021] According to another aspect, there is provided an image decoding method including: acquiring a bitstream including encoded encoding information; generating encoding information by performing decoding on the encoded encoding information; and determining prediction information to be used for decoding of a target block.

[0022] The prediction information may include a motion information candidate list and final motion information.

[0023] The image decoding method may further include performing prediction on the target block based on the prediction information using inter prediction, intra block copying, or intra template matching.

[0024] A motion vector search method for prediction may be used.

[0025] The motion vector search method may be template matching.

[0026] Inter prediction can be bidirectional prediction.

[0027] The motion vector may be fixed for one of the two directions of bi-directional prediction.

[0028] A motion search may be performed for the remaining one of the two directions.

[0029] A motion vector search method for inter prediction may be used.

[0030] The motion vector search method may be bilateral matching.

[0031] According to another aspect, a computer-readable storage medium for storing a bitstream for image decoding is provided, wherein the bitstream includes encoded encoding information, encoding information is generated by performing decoding on the encoded encoding information, and prediction information to be used for decoding of a target block is determined.

[0032] The prediction information may include a motion information candidate list and final motion information.

[0033] Prediction of the target block may be performed based on the prediction information using inter prediction, intra block copying, or intra template matching.

[0034] A motion vector search method for prediction may be used.

[0035] The motion vector search method may be template matching.

[0036] Inter prediction can be bidirectional prediction.

[0037] The motion vector may be fixed for one of the two directions of bi-directional prediction.

[0038] A motion search may be performed for the remaining one of the two directions.

[0039] Beneficial effects

[0040] Provided are an apparatus, method, and storage medium for performing encoding / decoding on a target block in an inter prediction mode, an intra block copy mode, and an intra template matching mode. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a block diagram showing a configuration of an embodiment of an encoding device to which the present disclosure is applied;

[0042] Figure 2 is a block diagram showing a configuration of an embodiment of a decoding device to which the present disclosure is applied;

[0043] Figure 3 is a diagram schematically illustrating a partition structure of an image when the image is encoded and decoded;

[0044] Figure 4 is a diagram illustrating a form of prediction unit (PU) that a coding unit (CU) can include;

[0045] Figure 5 is a diagram illustrating a form of a transform unit (TU) that can be included in a CU;

[0046] Figure 6 shows the division of blocks according to an example;

[0047] Figure 7 is a diagram for explaining an embodiment of an intra prediction process;

[0048] Figure 8 is a diagram showing reference samples used in the intra prediction process;

[0049] Figure 9 is a diagram for explaining an embodiment of an inter-frame prediction process;

[0050] Figure 10 shows spatial candidates according to an embodiment;

[0051] Figure 11 shows the order in which motion information of spatial candidates is added to a merge list according to an embodiment;

[0052] Figure 12 shows a transform and quantization process according to an example;

[0053] Figure 13 shows a diagonal scan according to an example;

[0054] Figure 14 shows a horizontal scan according to an example;

[0055] Figure 15 shows vertical scanning according to an example;

[0056] Figure 16 is a configuration diagram of an encoding device according to an embodiment;

[0057] Figure 17 is a configuration diagram of a decoding device according to an embodiment;

[0058] Figure 18 is a flowchart illustrating a target block prediction method and a bitstream generation method according to an embodiment;

[0059] Figure 19 is a flowchart illustrating a target block prediction method using a bitstream according to an embodiment;

[0060] Figure 20 shows motion vector differences in a symmetric motion vector difference pattern according to an example;

[0061] Figure 21 shows syntax elements signaled / encoded / decoded in symmetric motion vector difference mode according to an example;

[0062] Figure 22 shows a two-sided matching according to an example;

[0063] Figure 23 illustrates template matching according to an embodiment;

[0064] Figure 24 A first relationship between a search pattern and a resolution according to an example is shown;

[0065] Figure 25 shows a second relationship between search pattern and resolution according to an example;

[0066] Figure 26 shows a third relationship between search pattern and resolution according to an example;

[0067] Figure 27 shows a fourth relationship between search pattern and resolution according to an example;

[0068] Figure 28 shows a fifth relationship between search pattern and resolution according to an example;

[0069] Figure 29 shows a sixth relationship between search pattern and resolution according to an example;

[0070] Figure 30 A first template configuration method in an affine mode according to an example is shown;

[0071] Figure 31 A second template configuration method in an affine mode according to an example is shown;

[0072] Figure 32 is a first flowchart illustrating a method for selecting a scheme for configuring a motion information candidate list according to an example;

[0073] Figure 33 is a second flowchart illustrating a method for selecting a scheme for configuring a motion information candidate list according to an example;

[0074] Figure 34 is a third flowchart illustrating a method for selecting a scheme for configuring a motion information candidate list according to an example;

[0075] Figure 35 is a fourth flowchart illustrating a method for selecting a scheme for configuring a motion information candidate list according to an example;

[0076] Figure 36 shows reconfiguration of a motion information candidate list according to an embodiment;

[0077] Figure 37 shows the setting of a flag depending on the value of a motion information index according to an example;

[0078] Figure 38shows a first setting of an index depending on the value of a motion information index according to an example;

[0079] Figure 39 shows a second setting of the index depending on the value of the motion information index according to an example;

[0080] Figure 40 shows a first setting of an index depending on the value of a reference image index according to an example;

[0081] Figure 41 shows a second setting of the index depending on the value of the reference image index according to an example;

[0082] Figure 42 shows the setting of a flag depending on the value of a reference picture index according to an example;

[0083] Figure 43 A method for determining the sign of each component of a motion vector difference according to an embodiment is shown;

[0084] Figure 44 A first code for signaling / encoding / decoding of motion vector differences according to an example is shown;

[0085] Figure 45 shows a second code for signaling / encoding / decoding of motion vector differences according to an example;

[0086] Figure 46 shows a third code for signaling / encoding / decoding of motion vector differences according to an example;

[0087] Figure 47 shows a fourth code for signaling / encoding / decoding of motion vector differences according to an example;

[0088] Figure 48 shows a fifth code for signaling / encoding / decoding of motion vector differences according to an example;

[0089] Figure 49 A method for signaling information for a motion vector search mode according to an example is shown;

[0090] Figure 50 A first code indicating encoding information signaled in association with a block partition structure is shown according to an example; and

[0091] Figure 51 A second code indicating encoding information signaled in association with a block partition structure is shown according to an example.

[0092] Figure 52shows the filling of candidates for replacing zero vectors in an IBC list according to an example;

[0093] Figure 53 shows an IBC reference area depending on a first position of a target block according to an example;

[0094] Figure 54 shows an IBC reference area depending on a second position of a target block according to an example;

[0095] Figure 55 shows an IBC reference area depending on a third position of a target block according to an example;

[0096] Figure 56 shows an IBC reference area depending on a fourth position of a target block according to an example;

[0097] Figure 57 shows intra-frame template matching according to an example; and

[0098] Figure 58 The use of intraTMP block vectors for an IBC block according to an example is shown. DETAILED DESCRIPTION

[0099] The present invention can be variously modified and can have various embodiments, and specific embodiments will be described in detail below with reference to the accompanying drawings. However, it should be understood that these embodiments are not intended to limit the present invention to the specific disclosed forms, and they include all changes, equivalent forms or modified forms included in the spirit and scope of the present invention.

[0100] The following exemplary embodiments will be described in detail with reference to the accompanying drawings showing specific embodiments. These embodiments are described so that those of ordinary skill in the art to which the present disclosure pertains can easily implement these embodiments. It should be noted that the various embodiments are different from one another, but do not need to be mutually exclusive. For example, the specific shapes, structures, and characteristics described herein may be implemented as the other embodiments described without departing from the spirit and scope of the other embodiments associated with an embodiment. In addition, it should be understood that the position or arrangement of the various components in each disclosed embodiment can be changed without departing from the spirit and scope of the embodiments. Therefore, the attached detailed description is not intended to limit the scope of the present disclosure, and the scope of the exemplary embodiments is limited only by the attached claims and their equivalents (as long as they are appropriately described).

[0101] In the accompanying drawings, like reference numerals are used to designate the same or similar functions in various aspects. The shapes, sizes, etc. of components in the accompanying drawings may be exaggerated to make the description clear.

[0102] Terms such as "first" and "second" may be used to describe various components, but the components are not limited by the terms. The terms are only used to distinguish one component from another. For example, a first component may be referred to as a second component without departing from the scope of this specification. Similarly, a second component may be referred to as a first component. The term "and / or" may include a combination of multiple related description items or any one of the multiple related description items.

[0103] It will be understood that when a component is referred to as being “connected” or “coupled” to another component, the two components may be directly connected or coupled to each other, or intervening components may be present between the two components. On the other hand, it will be understood that when a component is referred to as being “directly connected or coupled,” there are no intervening components between the two components.

[0104] The components described in the embodiments are shown separately to indicate different feature functions, but this does not mean that each component is formed by a separate piece of hardware or software. That is, for the convenience of description, multiple components are arranged and included separately. For example, at least two components of the multiple components can be integrated into a single component. Conversely, a component can be divided into multiple components. As long as it does not depart from the essence of this specification, embodiments in which multiple components are integrated or embodiments in which some components are separated are included in the scope of this specification.

[0105] The terms used in the embodiments are only used to describe specific embodiments and are not intended to limit the present invention. Singular expressions include plural expressions, unless the contrary description is specifically indicated in the context. In the embodiments, it should be understood that terms such as "including" or "having" are only intended to indicate the presence of features, numbers, steps, operations, components, parts or a combination thereof, and are not intended to exclude the possibility of one or more other features, numbers, steps, operations, components, parts or a combination thereof being present or added. That is, in the embodiments, the expression describing a component "including" a specific component means that other components may be included in the practice of the present invention or the scope of the technical spirit of the present invention, but does not exclude the presence of components other than the specific components.

[0106] In an embodiment, the term "at least one" may mean one of one or more quantities (such as 1, 2, 3, and 4). In an embodiment, the term "plurality" may mean one of two or more quantities (such as 2, 3, and 4).

[0107] Some components of the embodiment are not essential components for performing necessary functions, but may be optional components that are only used to improve performance. The embodiment can be implemented using only the essential components for realizing the essence of the embodiment. For example, a structure that only includes essential components (excluding optional components that are only used to improve performance) is also included in the scope of the embodiment.

[0108] The embodiments will be described in detail below with reference to the accompanying drawings so that those skilled in the art can easily implement the embodiments. In the following description of the embodiments, detailed descriptions of well-known functions or configurations that are deemed to obscure the main points of this specification will be omitted. In addition, the same reference numerals are used to designate the same components throughout the drawings, and repeated descriptions of the same components will be omitted.

[0109] Hereinafter, "image" may refer to a single frame constituting a video, or may refer to the video itself. For example, "encoding and / or decoding an image" may refer to "encoding and / or decoding a video" or "encoding and / or decoding any one of a plurality of images constituting a video."

[0110] Hereinafter, the terms "video" and "moving picture" may be used to have the same meaning and may be used interchangeably with each other.

[0111] Hereinafter, a target image may be an encoding target image to be encoded and / or a decoding target image to be decoded. Furthermore, a target image may be an input image input to an encoding device or an input image input to a decoding device. Furthermore, a target image may be a current image, that is, a target image currently to be encoded and / or decoded. For example, the terms "target image" and "current image" may be used interchangeably to have the same meaning.

[0112] Hereinafter, the terms "image," "picture," "frame," and "screen" may be used to have the same meaning and may be used interchangeably with each other.

[0113] Hereinafter, a target block may be an encoding target block (i.e., a target to be encoded) and / or a decoding target block (i.e., a target to be decoded). In addition, a target block may be a current block, i.e., a target to be currently encoded and / or decoded. Herein, the terms "target block" and "current block" may be used to have the same meaning and may be used interchangeably with each other. The current block may refer to an encoding target block that is an encoding target during encoding and / or a decoding target block that is a decoding target during 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.

[0114] Hereinafter, the terms "block" and "unit" may be used to have the same meaning and may be used interchangeably with each other. Alternatively, a "block" may refer to a specific unit.

[0115] Hereinafter, the terms "region" and "segment" are used interchangeably with each other.

[0116] In the following embodiments, specific information, data, flags, indexes, elements, and attributes may have their own values. The value "0" corresponding to each of the information, data, flags, indexes, elements, and attributes may indicate false, logically false, or a first predefined value. In other words, the values ​​"0", false, logically false, and the first predefined value may be used interchangeably. The value "1" corresponding to each of the information, data, flags, indexes, elements, and attributes may indicate true, logically true, or a second predefined value. In other words, the values ​​"1", true, logically true, and the second predefined value may be used interchangeably.

[0117] When a variable such as i or j is used to indicate a row, column, or index, the value i may be an integer 0 or greater than 0, or may be an integer 1 or greater than 1. In other words, in an embodiment, each of the rows, columns, and indexes may be counted starting from 0, or may be counted starting from 1.

[0118] In an embodiment, the term “one or more” or the term “at least one” may mean the term “plurality.” The term “one or more” or the term “at least one” may be used interchangeably with “plurality.”

[0119] Hereinafter, terms to be used in the embodiments will be described.

[0120] Encoder: An encoder refers to a device for performing encoding. In other words, an encoder can refer to an encoding device.

[0121] Decoder: A decoder refers to a device for performing decoding. In other words, a decoder can refer to a decoding device.

[0122] Unit: A unit may refer to a unit of image encoding and decoding. The terms "unit" and "block" may be used to have the same meaning and may be used interchangeably with each other.

[0123] – A cell may be an M×N array of samples. Each of M and N may be a positive integer. A cell may generally represent a two-dimensional array of samples.

[0124] During image encoding and decoding, a "unit" may be a region generated by partitioning an image. In other words, a "unit" may be a specified region within an image. A single image may be partitioned into multiple units. Alternatively, an image may be partitioned into sub-parts, and a unit may represent each sub-part when encoding or decoding is performed on the partitioned sub-parts.

[0125] – During the encoding and decoding of an image, predefined processing can be performed on each unit according to the type of the unit.

[0126] – According to functions, unit types may be classified into macro units, coding units (CUs), prediction units (PUs), residual units, transform units (TUs), etc. Alternatively, according to functions, a unit may refer to a block, a macro block, 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, etc. For example, a target unit as a target of encoding and / or decoding may be at least one of a CU, a PU, a residual unit, and a TU.

[0127] The term "unit" may mean information including a luma component block, a chroma component block corresponding to the luma component block, and syntax elements for each block, such that the unit is designated to be distinguished from the block.

[0128] - The size and shape of the unit can be implemented differently. In addition, the unit can have any of a variety of sizes and shapes. Specifically, the shape of the unit can include not only a square but also a geometric shape that can be represented in two dimensions (2D), such as a rectangle, a trapezoid, a triangle, and a pentagon.

[0129] In addition, the unit information may include one or more of a unit type, a unit size, a unit depth, a unit encoding order, a unit decoding order, etc. For example, the unit type may indicate one of CU, PU, ​​residual unit, and TU.

[0130] - A cell can be partitioned into sub-cells, each sub-cell having a size smaller than that of the associated cell.

[0131] Depth: Depth may indicate the degree to which a cell is partitioned. In addition, the depth of a cell may indicate the level at which the corresponding cell exists when the cell is represented by a tree structure.

[0132] - The cell partition information may include a depth indicating the depth of the cell. The depth may indicate the number of times the cell is partitioned and / or the extent to which the cell is partitioned.

[0133] In a tree structure, the root node can be considered to have the smallest depth and the leaf node can be considered to have the largest depth. The root node can be the highest (top) node. The leaf node can be the lowest node.

[0134] A single cell can be hierarchically partitioned into multiple sub-cells, with the sub-cell having depth information based on a tree structure. In other words, a cell and the sub-cells generated by partitioning the cell may correspond to a node and its child nodes, respectively. Each partitioned sub-cell may have a cell depth. Since the depth indicates the number of times a cell has been partitioned and / or the extent to which a cell has been partitioned, the sub-cell partition information may include information about the size of the sub-cell.

[0135] In the tree structure, the top node may correspond to the initial node before partitioning. The top node may be referred to as the "root node." Furthermore, the root node may have the smallest depth value. Here, the depth of the top node may be level "0."

[0136] - A node at a depth level of "1" may represent a cell generated when the original cell is partitioned once. A node at a depth level of "2" may represent a cell generated when the original cell is partitioned twice.

[0137] - Leaf nodes at depth level "n" may represent cells generated when the initial cell is partitioned n times.

[0138] A leaf node may be a bottom node that cannot be partitioned further. The depth of a leaf node may be a maximum level. For example, a predefined value for the maximum level may be 3.

[0139] – QT depth may indicate the depth for a four-partition zone. BT depth may indicate the depth for a two-partition zone. TT depth may indicate the depth for a three-part zone.

[0140] - Sample: Sample can be the basic unit of building blocks. Available from 0 to 2 according to the bit depth (Bd) Bd A value of -1 is used to represent a sample point.

[0141] – A sample can be a pixel or a pixel value.

[0142] - Hereinafter, the terms "pixel" and "sample" may be used to have the same meaning and may be used interchangeably with each other.

[0143] Coding Tree Unit (CTU): A CTU may consist of a single luma (Y) coding tree block and two chroma (Cb, Cr) coding tree blocks associated with the luma coding tree block. Furthermore, a CTU may represent information including the above blocks and syntax elements for each block.

[0144] – Each coding tree unit (CTU) may be partitioned using one or more partitioning methods, such as quadtree (QT), binary tree (BT), and ternary tree (TT), to configure subunits such as coding units, prediction units, and transform units. Quadtree may represent a quadtree. In addition, each coding tree unit may be partitioned using a multi-type tree (MTT) using one or more partitioning methods.

[0145] – “CTU” may be used as a term to designate a pixel block as a processing unit in image decoding and encoding processes (such as in the case of partitioning an input image).

[0146] Coding Tree Block (CTB): “CTB” may be used as a term designating any one of a Y coding tree block, a Cb coding tree block, and a Cr coding tree block.

[0147] Neighboring block: A neighboring block (or neighboring block) may refer to a block adjacent to a target block. A neighboring block may refer to a reconstructed neighboring block.

[0148] Hereinafter, the terms “neighboring block” and “adjacent block” may be used to have the same meaning and may be used interchangeably with each other.

[0149] The neighboring block may represent a reconstructed neighboring block.

[0150] Spatially neighboring blocks: Spatially neighboring blocks may be blocks that are spatially adjacent to the target block. Neighboring blocks may include spatially neighboring blocks.

[0151] – A target block and spatially neighboring blocks may be included in a target picture.

[0152] - The spatially neighboring block may mean a block whose boundary contacts the target block or a block located within a predetermined distance from the target block.

[0153] - A spatially adjacent block may refer to a block adjacent to a vertex of the target block. Here, a block adjacent to a vertex of the target block may refer to a block vertically adjacent to a neighboring block horizontally adjacent to the target block or a block horizontally adjacent to a neighboring block vertically adjacent to the target block.

[0154] Temporally neighboring blocks: Temporally neighboring blocks may be blocks that are temporally adjacent to the target block. Neighboring blocks may include temporally neighboring blocks.

[0155] - Temporally neighboring blocks may include co-located blocks (col blocks).

[0156] The col block may be a block in a previously reconstructed co-located picture (col picture). The position of the col block in the col picture may correspond to the position of the target block in the target picture. Alternatively, the position of the col block in the col picture may be equal to the position of the target block in the target picture. The col picture may be a picture included in the reference picture list.

[0157] - The temporally neighboring block may be a block temporally adjacent to the spatially neighboring block of the target block.

[0158] Prediction mode: The prediction mode may be information indicating a mode for intra prediction or a mode for inter prediction.

[0159] Prediction unit: A prediction unit may be a basic unit for prediction such as inter prediction, intra prediction, inter compensation, intra compensation, and motion compensation.

[0160] A single prediction unit can be divided into multiple partitions or sub-prediction units of smaller size. The multiple partitions can also be the basic units when performing prediction or compensation. Partitions generated by dividing the prediction unit can also be prediction units.

[0161] Prediction unit partition: A prediction unit partition may be a shape into which a prediction unit is divided.

[0162] Reconstructed neighboring unit: The reconstructed neighboring unit may be a unit that is adjacent to the target unit and has been decoded and reconstructed.

[0163] The reconstructed neighboring cell may be a cell that is spatially adjacent to the target cell or temporally adjacent to the target cell.

[0164] - The reconstructed spatially neighboring unit may be a unit included in the target picture that has been reconstructed through encoding and / or decoding.

[0165] –The reconstructed temporally adjacent unit may be a unit included in the reference image and already reconstructed by encoding and / or decoding. The position of the reconstructed temporally adjacent unit in the reference image may be the same as the position of the target unit in the target picture, or may correspond to the position of the target unit in the target picture. In addition, the reconstructed temporally adjacent unit may be a block adjacent to the corresponding block in the reference image. Here, the position of the corresponding block in the reference image may correspond to the position of the target block in the target image. Here, the fact that the positions of the blocks correspond to each other may mean that the positions of the blocks are the same as each other, may mean that one block is included in another block, or may mean that one block occupies a specific position in another block.

[0166] Sub-picture: A picture can be divided into one or more sub-pictures. A sub-picture can be composed of one or more tile rows and one or more tile columns.

[0167] - A sub-picture may be an area in a picture having a square shape or a rectangular (ie, non-square rectangular) shape. In addition, a sub-picture may include one or more CTUs.

[0168] - A sub-picture can be a rectangular area of ​​one or more slices in a picture.

[0169] - A sub-picture may include one or more tiles, one or more bricks and / or one or more slices.

[0170] Tile: A tile may be an area in a picture having a square shape or a rectangular (ie, non-square rectangular) shape.

[0171] – A tile may include one or more CTUs.

[0172] – A parallel block can be partitioned into one or more partitions.

[0173] Block: A block may represent one or more CTU rows in a tile.

[0174] A tile can be partitioned into one or more partitions, each of which can include one or more CTU rows.

[0175] – Parallel blocks that are not partitioned into two parts can also represent partitions.

[0176] Slice: A slice may include one or more tiles in a picture. Alternatively, a slice may include one or more partitions in a tile.

[0177] - A sub-picture may comprise one or more slices that together cover a rectangular area of ​​the picture. Thus, every sub-picture boundary is also always a slice boundary, and every vertical sub-picture boundary is also always a vertical tile boundary.

[0178] Parameter set: The parameter set may correspond to header information in the internal structure of the bitstream.

[0179] The parameter set may include at least one of a video parameter set (VPS), a sequence parameter set (SPS), a picture parameter set (PPS), an adaptation parameter set (APS), a decoding parameter set (DPS), etc.

[0180] – Information signaled by each parameter set can be applied to pictures that reference the corresponding parameter set. For example, information in a VPS can be applied to pictures that reference the VPS. Information in an SPS can be applied to pictures that reference the SPS. Information in a PPS can be applied to pictures that reference the PPS.

[0181] – Each parameter set can refer to a higher parameter set. For example, PPS can refer to SPS. SPS can refer to VPS.

[0182] – In addition, the parameter set may include a parallel block group, slice header information, and parallel block header information. A parallel block group may be a group including multiple parallel blocks. In addition, the meaning of "parallel block group" may be the same as the meaning of "slice".

[0183] Rate-distortion optimization: The encoding device may use rate-distortion optimization to provide high encoding efficiency by utilizing a combination of the following: the size of the coding unit (CU), the prediction mode, the size of the prediction unit (PU), motion information, and the size of the transform unit (TU).

[0184] – The rate-distortion optimization scheme calculates the rate-distortion cost of each combination to select the optimal combination from these combinations. The rate-distortion cost can be calculated using the equation "D + λ * R". Generally, the combination that minimizes the rate-distortion cost can be selected as the optimal combination under the rate-distortion optimization scheme.

[0185] −D may represent distortion. D may be the average of the squares of the differences between the original transform coefficients and the reconstructed transform coefficients in the transform unit (ie, the mean square error).

[0186] -R may represent the rate, which may use relevant context information to represent the bit rate.

[0187] –λ represents the Lagrange multiplier. R may include not only encoding parameter information such as prediction mode, motion information, and coding block flag, but also bits generated due to encoding of transform coefficients.

[0188] – The encoding device may perform processes such as inter-frame prediction and / or intra-frame prediction, transformation, quantization, entropy coding, inverse quantization (dequantization), and / or inverse transformation in order to calculate accurate D and R. These processes significantly increase the complexity of the encoding device.

[0189] – Bitstream: A bitstream may refer to a stream of bits including encoded image information.

[0190] Parsing: Parsing can be the decision of the value of the syntax element by performing entropy decoding on the bitstream. Alternatively, the term "parsing" can refer to such entropy decoding itself.

[0191] Symbol: A symbol may be at least one of a syntax element, a coding parameter, and a transform coefficient of a coding target unit and / or a decoding target unit. In addition, a symbol may be a target of entropy coding or a result of entropy decoding.

[0192] Reference picture: A reference picture may be an image referenced by a unit in order to perform inter-frame prediction or motion compensation. Alternatively, a reference picture may be an image including a reference unit referenced by a target unit in order to perform inter-frame prediction or motion compensation.

[0193] Hereinafter, the terms “reference picture” and “reference image” may be used to have the same meaning and may be used interchangeably with each other.

[0194] Reference picture list: A reference picture list may be a list including one or more reference images used for inter prediction or motion compensation.

[0195] – The types of reference picture lists may include combined list (LC), list 0 (L0), list 1 (L1), list 2 (L2), list 3 (L3), etc.

[0196] – For inter prediction, one or more reference picture lists may be used.

[0197] Inter-frame prediction indicator: The inter-frame prediction indicator may indicate the inter-frame prediction direction for the target unit. Inter-frame prediction can be one of unidirectional prediction and bidirectional prediction. Optionally, the inter-frame prediction indicator may indicate the number of reference pictures used to generate the prediction unit for the target unit. Optionally, the inter-frame prediction indicator may indicate the number of prediction blocks used for inter-frame prediction or motion compensation for the target unit.

[0198] Prediction list utilization flag: The prediction list utilization flag may indicate whether to use at least one reference picture in a specific reference picture list to generate a prediction unit.

[0199] – The prediction list utilization flag can be used to derive the inter-frame prediction indicator. Conversely, the prediction list utilization flag can be used to derive the inter-frame prediction indicator. For example, if the prediction list utilization flag indicates "0" (as a first value), it can indicate that the reference picture in the reference picture list is not used to generate the prediction block for the target unit. If the prediction list utilization flag indicates "1" (as a second value), it can indicate that the reference picture list is used to generate the prediction unit for the target unit.

[0200] Reference picture index: A reference picture index may be an index indicating a specific reference picture in a reference picture list.

[0201] Picture Order Count (POC): The POC value of a picture may indicate the order in which the corresponding picture is displayed.

[0202] Motion Vector (MV): A motion vector is a 2D vector used for inter-frame prediction or motion compensation. A motion vector represents the offset between a target image and a reference image.

[0203] – For example, you can use a command such as (mv x , mv y ) to express MV. mv x Can indicate the horizontal component, mv y May indicate the vertical component.

[0204] -Search range: The search range may be a 2D area where a search for an MV is performed during inter prediction. For example, the size of the search range may be M×N. M and N may be positive integers, respectively.

[0205] Motion vector candidate: A motion vector candidate may be a block that is a prediction candidate when a motion vector is predicted or a motion vector of a block that is a prediction candidate.

[0206] – A motion vector candidate may be included in a motion vector candidate list.

[0207] Motion vector candidate list: A motion vector candidate list may be a list configured using one or more motion vector candidates.

[0208] Motion vector candidate index: The motion vector candidate index may be an indicator for indicating a motion vector candidate in the motion vector candidate list. Alternatively, the motion vector candidate index may be an index of a motion vector predictor.

[0209] Motion information: The motion information may be information including at least one of a reference picture list, a reference image, a motion vector candidate, a motion vector candidate index, a merge candidate, and a merge index, as well as a motion vector, a reference picture index, and an inter prediction indicator.

[0210] Merge candidate list: A merge candidate list may be a list configured using one or more merge candidates.

[0211] Merge candidate: A merge candidate may be a spatial merge candidate, a temporal merge candidate, a combined merge candidate, a combined bi-predictive merge candidate, a history-based candidate, an average merge candidate based on the average of two candidates, a zero merge candidate, etc. The merge candidate may include an inter prediction indicator and may include motion information such as prediction type information, reference picture index for each list, motion vector, prediction list utilization flag, and an inter prediction indicator.

[0212] Merge index: The merge index may be an indicator for indicating a merge candidate in a merge candidate list.

[0213] – The merge index may indicate a reconstructed unit for deriving a merge candidate among reconstructed units spatially adjacent to the target unit and reconstructed units temporally adjacent to the target unit.

[0214] – The merge index may indicate at least one of a plurality of pieces of motion information of the merge candidate.

[0215] Transform unit: A transform unit may be a basic unit for residual signal encoding and / or residual signal decoding (such as transform, inverse transform, quantization, inverse quantization, transform coefficient encoding, and transform coefficient decoding). A single transform unit may be partitioned into multiple sub-transform units of smaller size. Here, the transform may include one or more of a primary transform and a secondary transform, and the inverse transform may include one or more of a primary inverse transform and a secondary inverse transform.

[0216] Scaling: Scaling can be referred to as the process of multiplying the levels of the transform coefficients by a factor.

[0217] – As a result of scaling the transform coefficient levels, transform coefficients may be generated. Scaling may also be referred to as “inverse quantization”.

[0218] Quantization Parameter (QP): A quantization parameter may be a value used to generate a transform coefficient level for a transform coefficient during quantization. Alternatively, a quantization parameter may also be a value used to generate a transform coefficient by scaling the transform coefficient level during inverse quantization. Alternatively, a quantization parameter may be a value mapped to a quantization step size.

[0219] Delta quantization parameter: A delta quantization parameter represents the difference between the target unit's quantization parameter and the predicted quantization parameter.

[0220] Scanning: Scanning can refer to a method of arranging the order of coefficients in a cell, block, or matrix. For example, a method for arranging a 2D array in the form of a one-dimensional (1D) array can be referred to as "scanning." Alternatively, a method for arranging a 1D array in the form of a 2D array can also be referred to as "scanning" or "inverse scanning."

[0221] Transform coefficient: The transform coefficient may be a coefficient value generated when the encoding device performs transformation. Alternatively, the transform coefficient may be a coefficient value generated when the decoding device performs at least one of entropy decoding and inverse quantization.

[0222] - A quantized level or a quantized transform coefficient level generated by applying quantization to a transform coefficient or a residual signal may also be included in the meaning of the term "transform coefficient".

[0223] Quantization level: The quantization level may be a value generated when the encoding device performs quantization on the transform coefficient or residual signal. Alternatively, the quantization level may be a value that is a target of inverse quantization when the decoding device performs inverse quantization.

[0224] - A quantized transform coefficient level as a result of transformation and quantization may also be included in the meaning of the quantization level.

[0225] Non-zero transform coefficient: A non-zero transform coefficient may be a transform coefficient having a value other than 0, or may be a transform coefficient level having a value other than 0. Alternatively, a non-zero transform coefficient may be a transform coefficient having a value with a magnitude other than 0, or may be a transform coefficient level having a value with a magnitude other than 0.

[0226] Quantization Matrix: A quantization matrix may be a matrix used in a quantization process or an inverse quantization process to improve the subjective or objective image quality of an image. A quantization matrix may also be referred to as a "scaling list."

[0227] Quantization matrix coefficient: A quantization matrix coefficient can be each element in the quantization matrix. A quantization matrix coefficient can also be called a "matrix coefficient."

[0228] Default matrix: The default matrix may be a quantization matrix predefined by the encoding device and the decoding device.

[0229] Non-default matrix: A non-default matrix may be a quantization matrix that is not pre-defined by the encoding device and the decoding device. The non-default matrix may refer to a quantization matrix that is signaled by a user from the encoding device to the decoding device.

[0230] Most Probable Mode (MPM): The MPM may represent an intra prediction mode that is highly likely to be used for intra prediction for a target block.

[0231] The encoding apparatus and the decoding apparatus may determine one or more MPMs based on encoding parameters related to the target block and properties of an entity related to the target block.

[0232] The encoding device and the decoding device may determine one or more MPMs based on the intra-frame prediction mode of the reference block. The reference block may include multiple reference blocks. The multiple reference blocks may include a spatially neighboring block adjacent to the left of the target block and a spatially neighboring block adjacent to the top of the target block. In other words, depending on which intra-frame prediction mode has been used for the reference block, one or more different MPMs may be determined.

[0233] - One or more MPMs may be determined in the same manner in both the encoding device and the decoding device. That is, the encoding device and the decoding device may share the same MPM list including one or more MPMs.

[0234] MPM list: The MPM list may be a list including one or more MPMs. The number of the one or more MPMs in the MPM list may be predefined.

[0235] MPM indicator: The MPM indicator may indicate an MPM to be used for intra prediction for a target block among one or more MPMs in an MPM list. For example, the MPM indicator may be an index for the MPM list.

[0236] Since the MPM list is determined in the same manner in both the encoding device and the decoding device, there may be no need to transmit the MPM list itself from the encoding device to the decoding device.

[0237] – The MPM indicator may be signaled from the encoding apparatus to the decoding apparatus. Since the MPM indicator is signaled, the decoding apparatus may determine an MPM to be used for intra prediction for a target block among the MPMs in the MPM list.

[0238] MPM usage indicator: The MPM usage indicator may indicate whether an MPM usage mode is to be used for prediction for a target block. The MPM usage mode may be a mode of determining an MPM to be used for intra prediction for a target block using an MPM list.

[0239] – An MPM usage indicator may be signaled from an encoding device to a decoding device.

[0240] Signaling: "Signaling" may mean that information is sent from an encoding device to a decoding device. Alternatively, "signaling" may mean that information is included in a bitstream or recording medium by the encoding device. Information signaled by the encoding device can be used by the decoding device.

[0241] The encoding device may generate coded information by encoding information to be signaled. The coded information may be transmitted from the encoding device to the decoding device. The decoding device may obtain the information by decoding the transmitted coded information. Here, the encoding may be entropy encoding, and the decoding may be entropy decoding.

[0242] Selective signaling: Information can be selectively signaled. Selective signaling of information can mean that an encoding device selectively includes information in a bitstream or recording medium (depending on specific conditions). Selective signaling of information can mean that a decoding device selectively extracts information from a bitstream (depending on specific conditions).

[0243] Omission of signaling: Signaling for information may be omitted. Regarding information, omission of signaling for information may mean that the encoding device (under certain conditions) does not include the information in the bitstream or recording medium. Omission of signaling for information may mean that the decoding device (under certain conditions) does not extract the information from the bitstream.

[0244] Statistic: A variable, encoding parameter, constant, etc. may have a computable value. A statistic may be a value generated by performing a calculation (operation) on the value of a specified target. For example, a statistic may indicate one or more of the following: the average, weighted average, weighted sum, minimum, maximum, mode, median, and interpolated values ​​of the values ​​of a particular variable, encoding parameter, constant, etc.

[0245] Figure 1 is a block diagram showing a configuration of an embodiment of an encoding device to which the present disclosure is applied.

[0246] The encoding device 100 may be an encoder, a video encoding device, or an image encoding device. A video may include one or more images (pictures). The encoding device 100 may sequentially encode one or more images of a video.

[0247] Reference Figure 1, the encoding device 100 includes an inter-frame prediction unit 110, 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 (dequantization) unit 160, an inverse transform unit 170, an adder 175, a filter unit 180 and a reference picture buffer 190.

[0248] The encoding apparatus 100 may perform encoding on a target image using an intra mode and / or an inter mode. In other words, the prediction mode of the target block may be one of the intra mode and the inter mode.

[0249] Hereinafter, the terms “intra mode”, “intra prediction mode”, “intra-screen mode” and “intra prediction mode” may be used to have the same meaning and may be used interchangeably with each other.

[0250] Hereinafter, the terms “inter mode”, “inter prediction mode”, “inter-picture mode” and “inter-picture prediction mode” may be used to have the same meaning and may be used interchangeably with each other.

[0251] Hereinafter, the term "image" may refer to only a portion of an image, or may refer to a block. Furthermore, processing of an "image" may refer to sequential processing of a plurality of blocks.

[0252] In addition, the encoding device 100 can generate a bitstream including the encoded information by encoding the target image, and can output and store the generated bitstream. The generated bitstream can be stored in a computer-readable storage medium and can be streamed via a wired and / or wireless transmission medium.

[0253] When the intra mode is used as the prediction mode, the switch 115 may switch to the intra mode. When the inter mode is used as the prediction mode, the switch 115 may switch to the inter mode.

[0254] The encoding apparatus 100 may generate a prediction block of the target block. In addition, after having generated the prediction block, the encoding apparatus 100 may encode a residual block for the target block using a residual between the target block and the prediction block.

[0255] When the prediction mode is intra mode, the intra prediction unit 120 may use pixels of a previously encoded / decoded neighboring block adjacent to the target block as reference samples. The intra prediction unit 120 may use the reference samples to perform spatial prediction on the target block and may generate prediction samples for the target block via spatial prediction. The prediction samples may represent samples in the prediction block.

[0256] The inter prediction unit 110 may include a motion prediction unit and a motion compensation unit.

[0257] When the prediction mode is inter mode, the motion prediction unit may search for an area in the reference image that best matches the target block during the motion prediction process, and may derive a motion vector for the target block and the area based on the area found. Here, the motion prediction unit may use the search range as the target area for the search.

[0258] The reference image may be stored in the reference picture buffer 190. More specifically, when encoding and / or decoding of a reference image has been processed, the encoded and / or decoded reference image may be stored in the reference picture buffer 190.

[0259] Since decoded pictures are stored, the reference picture buffer 190 may be a decoded picture buffer (DPB).

[0260] The motion compensation unit can generate a prediction block for the target block by performing motion compensation using a motion vector. Here, the motion vector can be a two-dimensional (2D) vector used for inter-frame prediction. In addition, the motion vector can indicate the offset between the target image and the reference image.

[0261] When the motion vector has a value other than an integer, the motion prediction unit and the motion compensation unit may generate a prediction block by applying an interpolation filter to a partial area of ​​the reference image. In order to perform inter-frame prediction or motion compensation, it may be determined which mode among the skip mode, merge mode, advanced motion vector prediction (AMVP) mode, and current picture reference mode corresponds to a method for predicting and compensating the motion of the PU included in the CU based on the CU, and inter-frame prediction or motion compensation may be performed according to the mode.

[0262] The subtractor 125 may generate a residual block, which is the difference between the target block and the prediction block. The residual block may also be referred to as a "residual signal."

[0263] The residual signal may be the difference between the original signal and the predicted signal. Alternatively, the residual signal may be a signal generated by transforming or quantizing the difference between the original signal and the predicted signal, or a signal generated by transforming and quantizing the difference. The residual block may be a residual signal for a block unit.

[0264] The transform unit 130 may generate a transform coefficient by transforming the residual block, and may output the generated transform coefficient. Here, the transform coefficient may be a coefficient value generated by transforming the residual block.

[0265] The transform unit 130 may use one of a plurality of predefined transform methods when performing the transform.

[0266] The plurality of predefined transform methods may include discrete cosine transform (DCT), discrete sine transform (DST), Karhunen-Loeve transform (KLT), and the like.

[0267] The transform method used to transform the residual block may be determined based on at least one of the encoding parameters for the target block and / or the neighboring blocks. For example, the transform method may be determined based on at least one of the inter-prediction mode for the PU, the intra-prediction mode for the PU, the size of the TU, and the shape of the TU. Alternatively, transform information indicating the transform method may be signaled from the encoding device 100 to the decoding device 200.

[0268] When the transform skip mode is used, the transform unit 130 may omit the operation of transforming the residual block.

[0269] By performing quantization on the transform coefficients, quantized transform coefficient levels or quantized levels may be generated. Hereinafter, in an embodiment, each of the quantized transform coefficient levels and the quantized levels may also be referred to as a 'transform coefficient'.

[0270] The quantization unit 140 may generate quantized transform coefficient levels (i.e., quantized levels or quantized coefficients) by quantizing the transform coefficients according to the quantization parameters. The quantization unit 140 may output the generated quantized transform coefficient levels. In this case, the quantization unit 140 may quantize the transform coefficients using a quantization matrix.

[0271] The entropy encoding unit 150 may generate a bitstream by performing entropy encoding based on a probability distribution based on a value calculated by the quantization unit 140 and / or an encoding parameter value calculated during encoding. The entropy encoding unit 150 may output the generated bitstream.

[0272] The entropy encoding unit 150 may perform entropy encoding on information about pixels of an image and information required for decoding the image. For example, the information required for decoding the image may include syntax elements and the like.

[0273] When entropy coding is applied, fewer bits are allocated to symbols that appear more frequently, and more bits are allocated to symbols that appear less frequently. Since symbols are represented by this allocation, the size of the bit string used to encode the target symbol can be reduced. Therefore, entropy coding can improve the compression performance of video encoding.

[0274] In addition, to perform entropy coding, the entropy coding unit 150 may use a coding method such as Exponential Golomb, Context Adaptive Variable Length Coding (CAVLC), or Context Adaptive Binary Arithmetic Coding (CABAC). For example, the entropy coding unit 150 may use a variable length coding / code (VLC) table to perform entropy coding. For example, the entropy coding unit 150 may derive a binarization method for the target symbol. In addition, the entropy coding unit 150 may derive a probability model for the target symbol / bin. The entropy coding unit 150 may use the derived binarization method, probability model, and context model to perform arithmetic coding.

[0275] The entropy encoding unit 150 may transform coefficients in a 2D block form into a 1D vector form through a transform coefficient scanning method in order to encode quantized transform coefficient levels.

[0276] Coding parameters may be information required for encoding and / or decoding. Coding parameters may include information encoded by encoding device 100 and transmitted from encoding device 100 to a decoding device, and may also include information that can be derived during the encoding or decoding process. For example, the information transmitted to the decoding device may include syntax elements.

[0277] The coding parameters may include not only information (or flags or indices) such as syntax elements that are encoded by the encoding device and sent by the encoding device to the decoding device using a signal, but also information derived in the encoding or decoding process. In addition, the coding parameters may include information required for encoding or decoding an image. For example, the coding parameters may include at least one value of the following items, a combination of the following items, or statistics: the size of the unit / block, the shape / form of the unit / block, the depth of the unit / block, the partition information of the unit / block, the partition structure of the unit / block, information indicating whether the unit / block is partitioned in a quadtree structure, information indicating whether the unit / block is partitioned in a binary tree structure, the partition direction of the binary tree structure (horizontal or vertical), the partition form of the binary tree structure (symmetric partitioning or asymmetric partitioning), information indicating whether the unit / block is partitioned in a ternary tree structure, the partition direction of the ternary tree structure (horizontal or vertical), the partition form of the ternary tree structure (symmetric partitioning or asymmetric partitioning), symmetrical partitioning, etc.), information indicating whether the unit / block is partitioned in a multi-type tree structure, a combination and direction of partitions of the multi-type tree structure (horizontal direction or vertical direction, etc.), a partition form of the multi-type tree structure (symmetrical partitioning or asymmetrical partitioning, etc.), a partition tree in the multi-type tree form (binary tree or ternary tree), a prediction type (intra-frame prediction or inter-frame prediction), an intra-frame prediction mode / direction, an intra-frame luminance prediction mode / direction, an intra-frame chrominance prediction mode / direction, an intra-frame partition information, an inter-frame partition information, a coding block partition flag, a prediction block partition flag, a transform block partition flag, a reference sample filtering method, a reference sample filter tap, a reference sample filter coefficient, a 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, inter-frame prediction mode, motion information, motion vector, motion vector difference, reference picture index, inter-frame prediction direction, inter-frame prediction indicator, prediction list utilization flag, reference picture list, reference image, POC, motion vector predictor, motion vector prediction index, motion vector prediction candidate, motion vector candidate list, information indicating whether merge mode is used, merge index, merge candidate, merge candidate list, information indicating whether skip mode is used, type of interpolation filter, taps of interpolation filter, interpolation filter filter coefficients of a loop filter, size of a motion vector, accuracy of motion vector representation, transform type, transform size, information indicating whether a first transform is used, information indicating whether an additional (second) transform is used, first transform selection information (or first transform index), second transform selection information (or second transform index), information indicating the presence or absence of a residual signal, a coding block pattern, a coding block flag, a quantization parameter, a residual quantization parameter, a quantization matrix, information about a loop filter, information indicating whether a loop filter is applied, coefficients of a loop filter, taps of a loop filter, a shape / form of a loop filter, information indicating whether a deblocking filter is applied,Deblocking filter coefficients, deblocking filter taps, deblocking filter strength, deblocking filter shape / form, information indicating whether adaptive sample offset is applied, adaptive sample offset value, adaptive sample offset category, adaptive sample offset type, information indicating whether adaptive loop filter is applied, adaptive loop filter coefficients, adaptive loop filter taps, adaptive loop filter shape / form, binarization / debinarization method, context model, context model determination method, context model update method, information indicating whether normal mode is performed, information indicating whether bypass mode is performed, significant coefficient flag, last significant coefficient flag, coefficient group encoding flag, position of last significant coefficient, information indicating whether the coefficient value is greater than 1, information indicating whether the coefficient value is greater than 2, information indicating whether the coefficient value is greater than 3, remaining coefficient value information, sign information, reconstructed luminance sample, reconstructed chrominance sample, context binary bit , bypass binary bits, residual luminance samples, residual chrominance samples, transform coefficients, luminance transform coefficients, chrominance transform coefficients, quantization levels, luminance quantization levels, chrominance quantization levels, transform coefficient levels, transform coefficient level scanning methods, size of the motion vector search area on the decoding device side, shape / form of the motion vector search area on the decoding device side, number of motion vector searches on the decoding device side, size of CTU, minimum block size, maximum block size, maximum block depth, minimum block depth, image display / output order, stripe identification information, stripe type, stripe partition information, parallel block group identification information, parallel block group type, parallel block group partition information, parallel block identification information, parallel block type, parallel block partition information, picture type, bit depth, input sample bit depth, reconstructed sample bit depth, residual sample bit depth, transform coefficient bit depth, quantization level bit depth, information about the luminance signal, information about the chrominance signal, color space of the target block, and color space of the residual block. In addition, the above-mentioned coding parameter related information may also be included in the coding parameters. Information used to calculate and / or derive the above-mentioned coding parameters may also be included in the coding parameters. Information calculated or derived using the above encoding parameters may also be included in the encoding parameters.

[0278] In an embodiment, the coding parameters may refer to coding parameters of a target. A target may refer to a target of a particular process, such as a target block and a target picture. The coding parameters of a target block may refer to coding parameters of neighboring blocks of the target block, coding parameters of a reference block for the target block, or coding parameters of a reference picture for the target block. The coding parameters of a target picture may refer to coding parameters of a reference picture for the target picture or coding parameters of a block included in the target picture.

[0279] The first transform selection information may indicate a first transform applied to the target block.

[0280] The second transform selection information may indicate a second transform applied to the target block.

[0281] The residual signal may represent the difference between the original signal and the predicted signal. Alternatively, the residual signal may be a signal generated by transforming the difference between the original signal and the predicted signal. Alternatively, the residual signal may be a signal generated by transforming and quantizing the difference between the original signal and the predicted signal. The residual block may be a residual signal for a block.

[0282] Here, signaling information may mean that the encoding device 100 includes entropy-coded information generated by performing entropy encoding on a flag or index in the bitstream, and may mean that the decoding device 200 obtains information by performing entropy decoding on the entropy-coded information extracted from the bitstream. Here, the information may include a flag, an index, etc.

[0283] A signal may refer to information to be transmitted using a signal. Hereinafter, information for an image and a block may be referred to as a "signal." Furthermore, hereinafter, the terms "information" and "signal" may be used interchangeably to have the same meaning. For example, a specific signal may be a signal representing a specific block. An original signal may be a signal representing a target block. A predicted signal may be a signal representing a predicted block. A residual signal may be a signal representing a residual block.

[0284] The bitstream may include information based on a specific syntax. The encoding apparatus 100 may generate a bitstream including information according to the specific syntax. The decoding apparatus 200 may acquire information from the bitstream according to the specific syntax.

[0285] Since the encoding apparatus 100 performs encoding via inter-frame prediction, the encoded target image can be used as a reference image for another image to be subsequently processed. Therefore, the encoding apparatus 100 can reconstruct or decode the encoded target image and store the reconstructed or decoded image as a reference image in the reference picture buffer 190. For decoding, the encoded target image can be dequantized and inversely transformed.

[0286] The quantized levels may be dequantized by the dequantization unit 160 and inversely transformed by the inverse transform unit 170. The dequantization unit 160 may generate dequantized coefficients by performing inverse transform on the quantized levels. The inverse transform unit 170 may generate dequantized and inversely transformed coefficients by performing inverse transform on the dequantized coefficients.

[0287] The dequantized and inverse-transformed coefficients may be added to the prediction block by the adder 175. The dequantized and inverse-transformed coefficients are added to the prediction block, and then a reconstructed block may be generated. Here, the dequantized and / or inverse-transformed coefficients may refer to coefficients on which one or more of dequantization and inverse transformation have been performed, and may also refer to a reconstructed residual block. Here, the reconstructed block may refer to a restored block or a decoded block.

[0288] The reconstructed block may be filtered by the filter unit 180. The filter unit 180 may apply one or more filters selected from the group consisting of a deblocking filter, a sample adaptive offset (SAO) filter, an adaptive loop filter (ALF), and a non-local filter (NLF) to the reconstructed samples, the reconstructed block, or the reconstructed picture. The filter unit 180 may also be referred to as a "loop filter."

[0289] The deblocking filter can eliminate block distortion that occurs at boundaries between blocks in a reconstructed picture. To determine whether to apply the deblocking filter, the number of columns or rows of pixels included in the block and including the basis for determining whether to apply the deblocking filter to the target block can be determined.

[0290] When a deblocking filter is applied to a target block, the applied filter may be different depending on the required strength of the deblocking filter. In other words, among different filters, a filter determined in consideration of the strength of the deblocking filter may be applied to the target block. When a deblocking filter is applied to a target block, one or more filters selected from a long tap filter, a strong filter, a weak filter, and a Gaussian filter may be applied to the target block depending on the required strength of the deblocking filter.

[0291] Also, when vertical filtering and horizontal filtering are performed on a target block, the horizontal filtering and the vertical filtering may be performed in parallel.

[0292] SAO can add an appropriate offset to pixel values ​​to compensate for coding errors. SAO can perform pixel-by-pixel correction on an image to which deblocking has been applied, wherein the correction uses an offset that is the difference between the original image and the image to which deblocking has been applied. To perform offset correction on an image, a method can be used for dividing the pixels included in the image into a specific number of regions, determining the regions to which the offset is applied within the divided regions, and applying the offset to the determined regions. A method can also be used for applying the offset while taking into account edge information for each pixel.

[0293] The ALF can perform filtering based on values ​​obtained by comparing a reconstructed image with an original image. After the pixels included in the image have been divided into a predetermined number of groups, the filter to be applied to each group can be determined, and filtering can be performed differently for each group. Information related to whether an adaptive loop filter is applied can be signaled for each CU. This information can be signaled for the luma signal. The shape and filter coefficients of the ALF to be applied to each block can be different for each block. Alternatively, an ALF having a fixed form can be applied to the block regardless of its characteristics.

[0294] A non-local filter can perform filtering based on a reconstructed block similar to a target block. A region similar to the target block can be selected from the reconstructed image, and the target block can be filtered using statistical properties of the selected similar region. Information on whether a non-local filter is applied can be signaled for each coding unit (CU). Furthermore, the shape and filter coefficients of the non-local filter applied to a block can vary depending on the block.

[0295] The reconstructed block or reconstructed image filtered by the filter unit 180 may be stored as a reference picture in the reference picture buffer 190. The reconstructed block filtered by the filter unit 180 may be part of a reference picture. In other words, the reference picture may be a reconstructed picture composed of the reconstructed blocks filtered by the filter unit 180. The stored reference picture may then be used for inter-frame prediction or motion compensation.

[0296] Figure 2 is a block diagram showing a configuration of an embodiment of a decoding device to which the present disclosure is applied.

[0297] The decoding device 200 may be a decoder, a video decoding device, or an image decoding device.

[0298] Reference Figure 2 , the decoding device 200 may include an entropy decoding unit 210, an inverse quantization (dequantization) unit 220, an inverse transform unit 230, an intra-frame prediction unit 240, an inter-frame prediction unit 250, a switch 245, an adder 255, a filter unit 260 and a reference picture buffer 270.

[0299] The decoding apparatus 200 may receive a bitstream output from the encoding apparatus 100. The decoding apparatus 200 may receive a bitstream stored in a computer-readable storage medium, and may receive a bitstream streamed through a wired / wireless transmission medium.

[0300] The decoding apparatus 200 may perform decoding on a bitstream in an intra mode and / or an inter mode. In addition, the decoding apparatus 200 may generate a reconstructed image or a decoded image through decoding, and may output the reconstructed image or the decoded image.

[0301] For example, an operation of switching to intra mode or inter mode based on the prediction mode used for decoding may be performed by the switch 245. When the prediction mode used for decoding is intra mode, the switch 245 may be operated to switch to intra mode. When the prediction mode used for decoding is inter mode, the switch 245 may be operated to switch to inter mode.

[0302] The decoding device 200 can obtain a reconstructed residual block by decoding the input bit stream and can generate a prediction block. When the reconstructed residual block and the prediction block are obtained, the decoding device 200 can generate a reconstructed block as a target to be decoded by adding the reconstructed residual block to the prediction block.

[0303] The entropy decoding unit 210 may generate symbols by performing entropy decoding on the bitstream based on the probability distribution of the bitstream. The generated symbols may include symbols in the form of quantized transform coefficient levels (i.e., quantized levels or quantized coefficients). Here, the entropy decoding method may be similar to the entropy encoding method described above. That is, the entropy decoding method may be an inverse process of the entropy encoding method described above.

[0304] The entropy decoding unit 210 may change a coefficient having a one-dimensional (1D) vector form into a 2D block shape through a transform coefficient scanning method in order to decode quantized transform coefficient levels.

[0305] For example, the coefficients of the block can be changed to a 2D block shape by scanning the block coefficients using an upper right diagonal scan. Optionally, which of the upper right diagonal scan, vertical scan, and horizontal scan to use can be determined based on the size of the corresponding block and / or the intra prediction mode.

[0306] The quantized coefficients may be dequantized by the dequantization unit 220. The dequantization unit 220 may generate dequantized coefficients by performing dequantization on the quantized coefficients. Furthermore, the dequantized coefficients may be inversely transformed by the inverse transform unit 230. The inverse transform unit 230 may generate a reconstructed residual block by performing inverse transform on the dequantized coefficients. As a result of performing dequantization and inverse transform on the quantized coefficients, a reconstructed residual block may be generated. Here, when generating the reconstructed residual block, the dequantization unit 220 may apply a quantization matrix to the quantized coefficients.

[0307] When the intra mode is used, the intra prediction unit 240 may generate a prediction block by performing spatial prediction on a target block using pixel values ​​of a previously decoded neighboring block adjacent to the target block.

[0308] The inter-frame prediction unit 250 may include a motion compensation unit. Alternatively, the inter-frame prediction unit 250 may be designated as a "motion compensation unit."

[0309] When the inter mode is used, the motion compensation unit may generate a prediction block by performing motion compensation on the target block using a motion vector and a reference image stored in the reference picture buffer 270 .

[0310] The motion compensation unit may apply an interpolation filter to a partial area of ​​a reference image when a motion vector has a value other than an integer, and may generate a prediction block using the reference image to which the interpolation filter is applied. In order to perform motion compensation, the motion compensation unit may determine which mode of skip mode, merge mode, advanced motion vector prediction (AMVP) mode, and current picture reference mode corresponds to a motion compensation method for a PU included in the CU based on the CU, and may perform motion compensation according to the determined mode.

[0311] The reconstructed residual block and the prediction block may be added to each other by the adder 255. The adder 255 may generate a reconstructed block by adding the reconstructed residual block and the prediction block.

[0312] The reconstructed block may be filtered by the filter unit 260. The filter unit 260 may apply at least one of a deblocking filter, an SAO filter, an ALF, and an NLF to the reconstructed block or the reconstructed image. The reconstructed image may be a picture including the reconstructed block.

[0313] The filter unit may output a reconstructed image.

[0314] The reconstructed image and / or reconstructed block filtered by the filter unit 260 may be stored as a reference picture in the reference picture buffer 270. The reconstructed block filtered by the filter unit 260 may be part of a reference picture. In other words, the reference picture may be an image composed of the reconstructed blocks filtered by the filter unit 260. The stored reference picture may then be used for inter-frame prediction or motion compensation.

[0315] Figure 3 is a diagram schematically showing a partition structure of an image when the image is encoded and decoded.

[0316] Figure 3 An example in which a single unit is partitioned into a plurality of subunits may be schematically shown.

[0317] To efficiently partition an image, coding units (CUs) may be used in encoding and decoding. The term "unit" may be used to collectively designate 1) a block containing image samples and 2) a syntax element. For example, "partition of a unit" may refer to "partition of blocks corresponding to the unit."

[0318] A CU may be used as a basic unit for image encoding / decoding. A CU may be used as a unit to which one mode selected from intra mode and inter mode is applied in image encoding / decoding. In other words, in image encoding / decoding, it may be determined which mode of intra mode and inter mode is to be applied to each CU.

[0319] Also, a CU may be a basic unit for predicting, transforming, quantizing, inversely transforming, dequantizing, and encoding / decoding a transform coefficient.

[0320] Reference Figure 3 , the image 300 may be sequentially partitioned into units corresponding to the largest coding unit (LCU), and a partition structure may be determined for each LCU. Here, LCU may be used to have the same meaning as a coding tree unit (CTU).

[0321] Partitioning a cell may mean partitioning a block corresponding to the cell. Block partition information may include depth information regarding the depth of the cell. The depth information may indicate the number of times the cell is partitioned and / or the extent to which the cell is partitioned. A single cell may be hierarchically partitioned into multiple sub-cells, with the single cell having depth information based on a tree structure.

[0322] Each partitioned sub-unit may have depth information. The depth information may be information indicating the size of the CU. The depth information may be stored for each CU.

[0323] Each CU may have depth information. When a CU is partitioned, the depth of the CU generated from the partition may increase by 1 from the depth of the partitioned CU.

[0324] The partition structure may indicate the distribution of coding units (CUs) for efficiently encoding an image in the LCU 310. This distribution may be determined based on whether a single CU is partitioned into multiple CUs. The number of CUs generated by partitioning may be a positive integer of 2 or greater, including 2, 3, 4, 8, 16, and the like.

[0325] Depending on the number of CUs generated by partitioning, the horizontal size and vertical size of each CU generated by partitioning may be smaller than the horizontal size and vertical size of the CU before partitioning. For example, the horizontal size and vertical size of each CU generated by partitioning may be half the horizontal size and vertical size of the CU before partitioning.

[0326] Each partitioned CU may be recursively partitioned into four CUs in the same manner. Compared to at least one of the horizontal size and the vertical size of the CU before partitioning, at least one of the horizontal size and the vertical size of each partitioned CU may be reduced through recursive partitioning.

[0327] Partitioning of the CU may be performed recursively up to a predefined depth or a predefined size.

[0328] For example, the depth of the CU may have a value ranging from 0 to 3. The size of the CU may range from a size of 64×64 to a size of 8×8 according to the depth of the CU.

[0329] For example, the depth of the LCU 310 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 CU having a maximum coding unit size, and the SCU may be a CU having a minimum coding unit size.

[0330] Partitioning may begin at the LCU 310, and each time the horizontal and / or vertical dimensions of the CU are reduced by partitioning, the depth of the CU may increase by one.

[0331] For example, for each depth, a non-partitioned CU may have a size of 2N×2N. In addition, when a CU is partitioned, a CU of size 2N×2N may be partitioned into four CUs of size N×N. Whenever the depth increases by 1, the value of N may be halved.

[0332] Reference Figure 3 , an LCU with a depth of 0 may have 64×64 pixels or a 64×64 block. 0 may be the minimum depth. An SCU with a depth of 3 may have 8×8 pixels or an 8×8 block. 3 may be the maximum depth. Here, a CU with a 64×64 block as an LCU may be represented by a depth of 0. A CU with a 32×32 block may be represented by a depth of 1. A CU with a 16×16 block may be represented by a depth of 2. A CU with an 8×8 block as an SCU may be represented by a depth of 3.

[0333] Information about whether the corresponding CU is partitioned can be represented by the partition information of the CU. The partition information can be 1-bit information. All CUs except the SCU can include partition information. For example, the value of the partition information of a non-partitioned CU can be a first value. The value of the partition information of a partitioned CU can be a second value. When the partition information indicates whether the CU is partitioned, the first value can be "0" and the second value can be "1".

[0334] For example, when a single CU is partitioned into four CUs, the horizontal size and vertical size of each of the four CUs generated by partitioning can be half the horizontal size and vertical size of the CU before partitioning. When a CU of size 32×32 is partitioned into four CUs, the size of each of the four partitioned CUs can be 16×16. When a single CU is partitioned into four CUs, it can be considered that the CU has been partitioned in a quadtree structure. In other words, it can be considered that quadtree partitioning has been applied to the CU.

[0335] For example, when a single CU is partitioned into two CUs, the horizontal size or vertical size of each of the two CUs generated by partitioning may be half the horizontal size or vertical size of the CU before partitioning. When a CU of size 32×32 is partitioned vertically into two CUs, the size of each of the two partitioned CUs may be 16×32. When a CU of size 32×32 is partitioned horizontally into two CUs, the size of each of the two partitioned CUs may be 32×16. When a single CU is partitioned into two CUs, it can be considered that the CU has been partitioned in a binary tree structure. In other words, it can be considered that binary tree partitioning has been applied to the CU.

[0336] For example, when a single CU is partitioned (or divided) into three CUs, the original CU before partitioning is partitioned so that its horizontal size or vertical size is divided at a ratio of 1:2:1, thereby enabling the generation of three sub-CUs. For example, when a CU of size 16×32 is partitioned horizontally into three sub-CUs, the three sub-CUs generated by the partitioning may have sizes of 16×8, 16×16, and 16×8, respectively, in the direction from top to bottom. For example, when a CU of size 32×32 is partitioned vertically into three sub-CUs, the three sub-CUs generated by the partitioning may have sizes of 8×32, 16×32, and 8×32, respectively, in the direction from left to right. When a single CU is partitioned into three CUs, the CU may be considered to be partitioned in a ternary tree form. In other words, ternary tree partitioning may be considered to have been applied to the CU.

[0337] Both quadtree partitioning and binary tree partitioning are applied to Figure 3 LCU 310.

[0338] In the encoding apparatus 100, a coding tree unit (CTU) of size 64×64 may be partitioned into multiple smaller CUs using a recursive quadtree structure. A single CU may be partitioned into four CUs of the same size. Each CU may be recursively partitioned and may have a quadtree structure.

[0339] By recursive partitioning of CUs, the optimal partitioning method that incurs the minimum rate-distortion cost can be selected.

[0340] Figure 3 A coding tree unit (CTU) 320 in FIG. 1 is an example of a CTU to which quadtree partitioning, binarytree partitioning, and ternarytree partitioning are all applied.

[0341] As described above, in order to partition a CTU, at least one of quadtree partitioning, binary tree partitioning, and ternary tree partitioning may be applied to the CTU. Partitioning may be applied based on a specific priority.

[0342] For example, quadtree partitioning may be preferentially applied to CTUs. CUs that cannot be further partitioned in quadtree form may correspond to leaf nodes of the quadtree. CUs corresponding to leaf nodes of the quadtree may be root nodes of a binary tree and / or ternary tree. That is, CUs corresponding to leaf nodes of the quadtree may be partitioned in a binary tree or ternary tree form, or may not be further partitioned. In this case, each CU generated by applying binary tree partitioning or ternary tree partitioning to a CU corresponding to a leaf node of the quadtree is prevented from being partitioned again by the quadtree, thereby efficiently performing block partitioning and / or signaling block partition information.

[0343] The quad partition information may be used to signal the partition of the CU corresponding to each node of the quadtree. The quad partition information having a first value (e.g., "1") may indicate that the corresponding CU is partitioned in a quadtree manner. The quad partition information having a second value (e.g., "0") may indicate that the corresponding CU is not partitioned in a quadtree manner. The quad partition information may be a flag having a specific length (e.g., 1 bit).

[0344] There may be no priority between binary tree partitioning and ternary tree partitioning. That is, the CU corresponding to the leaf node of the quadtree may be partitioned in a binary tree form or a ternary tree form. In addition, the CU generated by binary tree partitioning or ternary tree partitioning may be further partitioned in a binary tree form or a ternary tree form, or may not be further partitioned.

[0345] Partitioning performed when there is no priority between binary tree partitioning and ternary tree partitioning may be referred to as "multi-type tree partitioning." That is, the CU corresponding to the leaf node of the quadtree may be the root node of the multi-type tree. The partitioning of the CU corresponding to each node of the multi-type tree may be signaled using at least one of information indicating whether the CU is partitioned according to the multi-type tree, partition direction information, and partition tree information. For the partitioning of the CU corresponding to each node of the multi-type tree, information indicating whether partitioning according to the multi-type tree is performed, partition direction information, and partition tree information may be sequentially signaled.

[0346] For example, information indicating whether a CU is partitioned in a multi-type tree and having a first value (e.g., "1") may indicate that the corresponding CU is partitioned in a multi-type tree form. Information indicating whether a CU is partitioned in a multi-type tree and having a second value (e.g., "0") may indicate that the corresponding CU is not partitioned in a multi-type tree form.

[0347] When a CU corresponding to each node of the multi-type tree is partitioned in the multi-type tree form, the corresponding CU may further include partition direction information.

[0348] The partition direction information may indicate the partition direction of the multi-type tree partition. Partition direction information having a first value (e.g., "1") may indicate that the corresponding CU is partitioned in the vertical direction. Partition direction information having a second value (e.g., "0") may indicate that the corresponding CU is partitioned in the horizontal direction.

[0349] When a CU corresponding to each node of a multi-type tree is partitioned in a multi-type tree form, the corresponding CU may further include partition tree information. The partition tree information may indicate a tree used for multi-type tree partitioning.

[0350] For example, partition tree information having a first value (eg, '1') may indicate that the corresponding CU is partitioned in a binary tree format. Partition tree information having a second value (eg, '0') may indicate that the corresponding CU is partitioned in a ternary tree format.

[0351] Here, each of the above-mentioned information indicating whether partitioning by the multi-type tree is performed, the partition tree information, and the partition direction information may be a flag having a specific length (for example, 1 bit).

[0352] At least one of the four partition information, the information indicating whether partitioning by a multi-type tree is performed, the partition direction information, and the partition tree information may be entropy encoded and / or decoded. To perform entropy encoding / decoding of such information, information of neighboring CUs adjacent to the target CU may be used.

[0353] For example, it can be considered that the partition form (i.e., partitioned / non-partitioned, partition tree, and / or partition direction) of the left CU and / or the upper CU is likely to be similar to the partition form of the target CU. Therefore, based on the information of the neighboring CU, context information for entropy encoding and / or entropy decoding of the information of the target CU can be derived. Here, the information of the neighboring CU may include at least one of the following: 1) four partition information of the neighboring CU, 2) information indicating whether the neighboring CU is partitioned according to a multi-type tree, 3) partition direction information of the neighboring CU, and 4) partition tree information of the neighboring CU.

[0354] In another embodiment of binary tree partitioning and ternary tree partitioning, binary tree partitioning may be performed first. That is, binary tree partitioning may be applied first, and then the CU corresponding to the leaf node of the binary tree may be set as the root node of the ternary tree. In this case, quadtree partitioning or binary tree partitioning may not be performed on the CU corresponding to the node of the ternary tree.

[0355] A CU that is not further partitioned by quadtree partitioning, binary tree partitioning, and / or ternary tree partitioning may be a unit for coding, prediction, and / or transformation. That is, the CU may not be further partitioned for prediction and / or transformation. Therefore, the partition structure for partitioning the CU into prediction units (PUs) and / or transform units (TUs), its partition information, etc. may not be present in the bitstream.

[0356] However, when the size of the CU as the unit of partition is larger than the size of the maximum transform block, the CU may be recursively partitioned until the size of the CU becomes smaller than or equal to the size of the maximum transform block. For example, when the size of the CU is 64×64 and the size of the maximum transform block is 32×32, the CU may be partitioned into four 32×32 blocks to perform transform. For example, when the size of the CU is 32×64 and the size of the maximum transform block is 32×32, the CU may be partitioned into two 32×32 blocks.

[0357] In this case, information indicating whether the CU is partitioned for transform may not be separately signaled. Without signaling, whether the CU is partitioned may be determined by comparing the horizontal size (and / or vertical size) of the CU with the horizontal size (and / or vertical size) of the largest transform block. For example, when the horizontal size of the CU is larger than the horizontal size of the largest transform block, the CU may be divided into two equal parts vertically. Furthermore, when the vertical size of the CU is larger than the vertical size of the largest transform block, the CU may be divided into two equal parts horizontally.

[0358] Information about the maximum size and / or minimum size of a CU and information about the maximum size and / or minimum size of a transform block may be signaled or determined at a level higher than the level of the CU. For example, the higher level may be a sequence level, a picture level, a tile level, a tile group level, or a slice level. For example, the minimum size of a CU may be set to 4×4. For example, the maximum size of a transform block may be set to 64×64. For example, the maximum size of a transform block may be set to 4×4.

[0359] Information about the minimum size of the CU corresponding to the leaf node of the quadtree (i.e., the minimum size of the quadtree) and / or information about the maximum depth of the path from the root node to the leaf node of the multi-type tree (i.e., the maximum depth of the multi-type tree) may be signaled or determined at a level higher than the level of the CU. For example, the higher level may be a sequence level, a picture level, a slice level, a tile group level, or a tile level. Information about the minimum size of the quadtree and / or information about the maximum depth of the multi-type tree may be signaled or determined separately at each of the intra-slice level and the inter-slice level.

[0360] Information about the difference between the size of the CTU and the maximum size of the transform block may be signaled or determined at a level higher than the level of the CU. For example, the higher level may be a sequence level, a picture level, a slice level, a tile group level, or a tile level. Information about the maximum size of the CU corresponding to each node of the binary tree (i.e., the maximum size of the binary tree) may be determined based on the size of the CTU and the information about the difference. The maximum size of the CU corresponding to each node of the ternary tree (i.e., the maximum size of the ternary tree) may have different values ​​depending on the type of slice. For example, the maximum size of the ternary tree at the intra-slice level may be 32×32. For example, the maximum size of the ternary tree at the inter-slice level may be 128×128. For example, the minimum size of the CU corresponding to each node of the binary tree (i.e., the minimum size of the binary tree) and / or the minimum size of the CU corresponding to each node of the ternary tree (i.e., the minimum size of the ternary tree) may be set to the minimum size of the CU.

[0361] In 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. In addition, 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.

[0362] Based on the various block sizes and depths described above, quad partition information, information indicating whether partitioning by a multi-type tree is performed, partition tree information, and / or partition direction information may or may not exist in a bitstream.

[0363] For example, when the size of the CU is not greater than the minimum size of the quadtree, the CU may not include quad partition information, and the quad partition information of the CU may be inferred as the second value.

[0364] For example, when the size (horizontal size and vertical size) of the CU corresponding to each node of the multi-type tree is larger 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 CU may not be partitioned in the binary tree form and / or the ternary tree form. In this determination method, information indicating whether partitioning by the multi-type tree is performed may not be signaled, but may be inferred as the second value.

[0365] Alternatively, when the size (horizontal and vertical dimensions) of the CU corresponding to each node of the multi-type tree is equal to the minimum size (horizontal and vertical dimensions) of the binary tree, or when the size (horizontal and vertical dimensions) of the CU is equal to twice the minimum size (horizontal and vertical dimensions) of the ternary tree, the CU may not be partitioned in a binary tree and / or ternary tree format. With this determination method, information indicating whether partitioning according to the multi-type tree is performed may not be signaled but may be inferred as a second value. This is because when the CU is partitioned in a binary tree and / or ternary tree format, a CU smaller than the minimum size of the binary tree and / or the minimum size of the ternary tree is generated.

[0366] Optionally, binary tree partitioning or ternary tree partitioning may be limited based on the size of the virtual pipeline data unit (i.e., the size of the pipeline buffer). For example, when a CU is partitioned into sub-CUs that do not fit in the size of the pipeline buffer through binary tree partitioning or ternary tree partitioning, the binary tree partitioning or ternary tree partitioning may be limited. The size of the pipeline buffer may be equal to the maximum size of the transform block (e.g., 64×64).

[0367] For example, when the size of the pipeline buffer is 64×64, the following partitions may be limited.

[0368] - Ternary partitioning for N×M CUs (where N and / or M is 128)

[0369] - Horizontal binary tree partitioning for 128×N CUs (where N<=64)

[0370] -Vertical binary tree partitioning for N×128 CUs (where N<=64)

[0371] Alternatively, when the depth of the CU corresponding to each node of the multi-type tree is equal to the maximum depth of the multi-type tree, the CU may not be partitioned in a binary tree form and / or a ternary tree form. In this determination method, information indicating whether partitioning by the multi-type tree is performed may not be signaled but may be inferred as the second value.

[0372] Alternatively, information indicating whether partitioning according to a multi-type tree is performed may be signaled 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 the CU corresponding to each node of the multi-type tree. Otherwise, the CU may not be partitioned in a binary tree form and / or a ternary tree form. With this determination, information indicating whether partitioning according to a multi-type tree is performed may not be signaled but may be inferred as a second value.

[0373] Alternatively, for the CU corresponding to each node of the multi-type tree, partition direction information may be signaled only when both vertical binary tree partitioning and horizontal binary tree partitioning are possible, or only when both vertical ternary tree partitioning and horizontal ternary tree partitioning are possible. Otherwise, the partition direction information may not be signaled but may be inferred as a value indicating the direction in which the CU may be partitioned.

[0374] Alternatively, for a CU corresponding to each node of a multi-type tree, partition tree information may be signaled only when both vertical binary tree partitioning and vertical ternary tree partitioning are possible, or only when both horizontal binary tree partitioning and horizontal ternary tree partitioning are possible. Otherwise, the partition tree information may not be signaled but may be inferred as a value indicating a tree of partitions applicable to the CU.

[0375] Figure 4 is a diagram illustrating a form of prediction units that a coding unit can include.

[0376] In a CU partitioned from an LCU, the CU that is no longer partitioned can be divided into one or more prediction units (PUs). This division is also called "partitioning."

[0377] PU can be a basic unit for prediction. PU can be encoded and decoded in any one of skip mode, inter mode and intra mode. PU can be partitioned into various shapes according to each mode. For example, Figure 1 The target block described above refers to Figure 2 The target blocks described may all be PUs.

[0378] A CU may not be split into PUs. When a CU is not split into PUs, the size of the CU and the size of the PU may be equal to each other.

[0379] In skip mode, partitioning may not exist in a CU.In skip mode, a 2Nx2N mode 410 may be supported without partitioning, wherein in the 2Nx2N mode 410, the size of the PU and the size of the CU are the same as each other.

[0380] In inter mode, eight types of partition shapes may exist in a CU. For example, in inter mode, 2N×2N mode 410, 2N×N mode 415, N×2N mode 420, N×N mode 425, 2N×nU mode 430, 2N×nD mode 435, nL×2N mode 440, and nR×2N mode 445 may be supported.

[0381] In intra mode, 2N×2N mode 410 and N×N mode 425 may be supported.

[0382] In 2N×2N mode 410, a PU of size 2N×2N may be encoded. A PU of size 2N×2N may represent a PU of the same size as a CU. For example, a PU of size 2N×2N may have a size of 64×64, 32×32, 16×16, or 8×8.

[0383] In NxN mode 425, PUs of size NxN may be encoded.

[0384] For example, in intra prediction, when the size of a PU is 8×8, four partitioned PUs may be encoded, and the size of each partitioned PU may be 4×4.

[0385] When encoding a PU in intra mode, any one of multiple intra prediction modes may be used to encode the PU. For example, HEVC technology provides 35 intra prediction modes, and a PU may be encoded in any of the 35 intra prediction modes.

[0386] Which mode of the 2Nx2N mode 410 and the NxN mode 425 is to be used to encode the PU may be determined based on the rate-distortion penalty.

[0387] The encoding device 100 may perform an encoding operation on a PU of size 2N×2N. Here, the encoding operation may be an operation of encoding the PU in each of a plurality of intra-prediction modes that can be used by the encoding device 100. Through the encoding operation, an optimal intra-prediction mode for the PU of size 2N×2N may be derived. The optimal intra-prediction mode may be an intra-prediction mode that, among the plurality of intra-prediction modes that can be used by the encoding device 100, results in a minimum rate-distortion cost when encoding the PU of size 2N×2N.

[0388] In addition, the encoding device 100 may sequentially perform encoding operations on each PU obtained by performing N×N partitioning. Here, the encoding operation may be an operation of encoding the PU in each of a plurality of intra-frame prediction modes that can be used by the encoding device 100. Through the encoding operation, the optimal intra-frame prediction mode for the PU of size N×N may be derived. The optimal intra-frame prediction mode may be the intra-frame prediction mode that produces the minimum rate-distortion cost when encoding the PU of size N×N among the plurality of intra-frame prediction modes that can be used by the encoding device 100.

[0389] The encoding apparatus 100 may determine which of the PU of size 2N×2N and the PU of size N×N to be encoded based on a comparison between the rate-distortion cost of the PU of size 2N×2N and the rate-distortion cost of the PU of size N×N.

[0390] A single CU may be partitioned into one or more PUs, and a PU may be partitioned into multiple PUs.

[0391] For example, when a single PU is partitioned into four PUs, the horizontal and vertical sizes of each of the four PUs generated by partitioning can be half the horizontal and vertical sizes of the PU before partitioning. When a PU of size 32×32 is partitioned into four PUs, the size of each of the four partitioned PUs can be 16×16. When a single PU is partitioned into four PUs, it can be considered that the PU has been partitioned in a quadtree structure.

[0392] For example, when a single PU is partitioned into two PUs, the horizontal size or vertical size of each of the two PUs generated by partitioning may be half the horizontal size or vertical size of the PU before partitioning. When a PU of size 32×32 is partitioned vertically into two PUs, the size of each of the two partitioned PUs may be 16×32. When a PU of size 32×32 is partitioned horizontally into two PUs, the size of each of the two partitioned PUs may be 32×16. When a single PU is partitioned into two PUs, the PU may be considered to have been partitioned in a binary tree structure.

[0393] Figure 5 is a diagram illustrating forms of transformation units that can be included in a coding unit.

[0394] A transform unit (TU) may be a basic unit used for processes such as transform, quantization, inverse transform, inverse quantization, entropy encoding, and entropy decoding in a CU.

[0395] A TU may have a square shape or a rectangular shape. The shape of a TU may be determined based on the size and / or shape of a CU.

[0396] Among the CUs partitioned from the LCU, the CUs that are no longer partitioned into CUs may be partitioned into one or more TUs. Here, the partition structure of the TU may be a quadtree structure. For example, Figure 5 As shown in , a single CU 510 can be partitioned one or more times according to a quadtree structure. Through this partitioning, a single CU 510 can be composed of TUs of various sizes.

[0397] It can be considered that a single CU is recursively split when it is split two or more times. Through splitting, a single CU can be composed of transform units (TUs) of various sizes.

[0398] Alternatively, a single CU may be split into one or more TUs based on the number of vertical lines and / or horizontal lines that partition the CU.

[0399] A CU may be divided into symmetric TUs or asymmetric TUs. To divide into asymmetric TUs, information about the size and / or shape of each TU may be signaled from the encoding apparatus 100 to the decoding apparatus 200. Alternatively, the size and / or shape of each TU may be derived from the information about the size and / or shape of the CU.

[0400] A CU may not be divided into TUs. When a CU is not divided into TUs, the size of the CU and the size of the TU may be equal to each other.

[0401] A single CU may be partitioned into one or more TUs, and a TU may be partitioned into multiple TUs.

[0402] For example, when a single TU is partitioned into four TUs, the horizontal size and vertical size of each of the four TUs generated by partitioning can be half the horizontal size and vertical size of the TU before partitioning. When a 32×32 TU is partitioned into four TUs, the size of each of the four partitioned TUs can be 16×16. When a single TU is partitioned into four TUs, the TU can be considered to have been partitioned in a quadtree structure.

[0403] For example, when a single TU is partitioned into two TUs, the horizontal size or vertical size of each of the two TUs generated by partitioning may be half the horizontal size or vertical size of the TU before partitioning. When a 32×32 TU is partitioned vertically into two TUs, the size of each of the two partitioned TUs may be 16×32. When a 32×32 TU is partitioned horizontally into two TUs, the size of each of the two partitioned TUs may be 32×16. When a single TU is partitioned into two TUs, the TU may be considered to have been partitioned in a binary tree structure.

[0404] Can be used with Figure 5 The CU is divided in different ways as shown in FIG.

[0405] For example, a single CU may be split into three CUs, and the horizontal sizes or vertical sizes of the three CUs generated by the splitting may be 1 / 4, 1 / 2, and 1 / 4 of the horizontal size or vertical size of the original CU before the splitting, respectively.

[0406] For example, when a CU of size 32×32 is vertically split into three CUs, the sizes of the three CUs generated by the splitting may be 8×32, 16×32, and 8×32, respectively. In this way, when a single CU is split into three CUs, it can be considered that the CU is split in the form of a ternary tree.

[0407] One of the exemplary partitioning forms (i.e., quadtree partitioning, binary tree partitioning, and ternary tree partitioning) may be applied to the partitioning of a CU, and multiple partitioning schemes may be combined and used together for the partitioning of a CU. Here, the case where multiple partitioning schemes are combined and used together may be referred to as "composite tree form partitioning."

[0408] Figure 6 The division of blocks according to an example is shown.

[0409] In the video encoding and / or decoding process, such as Figure 6 As shown in , the target block can be divided. For example, the target block can be a CU.

[0410] For splitting of the target block, an indicator indicating splitting information may be signaled from the encoding apparatus 100 to the decoding apparatus 200. The splitting information may be information indicating how the target block is split.

[0411] The split information may be one or more of a split flag (hereinafter referred to as "split_flag"), a quad-binary flag (hereinafter referred to as "QB_flag"), a quadtree flag (hereinafter referred to as "quadtree_flag"), a binary tree flag (hereinafter referred to as "binarytree_flag") and a binary type flag (hereinafter referred to as "Btype_flag").

[0412] "split_flag" may be a flag indicating whether a block is split. For example, a split_flag value of 1 may indicate that the corresponding block is split. A split_flag value of 0 may indicate that the corresponding block is not split.

[0413] "QB_flag" may be a flag indicating which of the quadtree and binary tree forms corresponds to the shape of the block partition. For example, a QB_flag value of 0 may indicate that the block is partitioned in a quadtree form. A QB_flag value of 1 may indicate that the block is partitioned in a binary tree form. Alternatively, a QB_flag value of 0 may indicate that the block is partitioned in a binary tree form. A QB_flag value of 1 may indicate that the block is partitioned in a quadtree form.

[0414] "quadtree_flag" may be a flag indicating whether the block is partitioned in a quadtree form. For example, a quadtree_flag value of 1 may indicate that the block is partitioned in a quadtree form. A quadtree_flag value of 0 may indicate that the block is not partitioned in a quadtree form.

[0415] "binarytree_flag" may be a flag indicating whether the block is partitioned in a binary tree form. For example, a binarytree_flag value of 1 may indicate that the block is partitioned in a binary tree form. A binarytree_flag value of 0 may indicate that the block is not partitioned in a binary tree form.

[0416] "Btype_flag" may be a flag indicating which of vertical and horizontal divisions corresponds to the division direction when the block is divided in a binary tree form. For example, a Btype_flag value of 0 may indicate that the block is divided in the horizontal direction. A Btype_flag value of 1 may indicate that the block is divided in the vertical direction. Alternatively, a Btype_flag value of 0 may indicate that the block is divided in the vertical direction. A Btype_flag value of 1 may indicate that the block is divided in the horizontal direction.

[0417] For example, it may be derived by signaling at least one of quadtree_flag, binarytree_flag, and Btype_flag. Figure 6 The partitioning information of the blocks in is shown in Table 1 below.

[0418] Table 1

[0419]

[0420] For example, the QB_flag may be derived by signaling at least one of the split_flag, QB_flag, and Btype_flag. Figure 6 The partitioning information of the blocks in is shown in Table 2 below.

[0421] Table 2

[0422]

[0423] The partitioning method may be limited to a quadtree or a binary tree depending on the size and / or shape of the block. When this restriction is applied, split_flag may be a flag indicating whether the block is partitioned in a quadtree or a flag indicating whether the block is partitioned in a binary tree. The size and shape of the block may be derived based on the depth information of the block, and the depth information may be signaled from the encoding device 100 to the decoding device 200.

[0424] When the block size falls within a specific range, partitioning only in quadtree form is possible. For example, the specific range may be defined by at least one of a maximum block size and a minimum block size that can be partitioned only in quadtree form.

[0425] Information indicating the maximum block size and the minimum block size that can be divided only in the quadtree form may be signaled from the encoding apparatus 100 to the decoding apparatus 200 through a bitstream. In addition, this information may be signaled for at least one of units such as a video, a sequence, a picture, a parameter, a tile group, and a slice (or segment).

[0426] Alternatively, the maximum block size and / or the minimum block size may be a fixed size predefined by the encoding device 100 and the decoding device 200. For example, when the size of the block is larger than 64×64 and smaller than 256×256, only quadtree division is possible. In this case, split_flag may be a flag indicating whether quadtree division is performed.

[0427] When the size of the block is larger than the maximum size of the transform block, only partitioning in a quadtree form is possible. Here, the sub-block generated by the partitioning may be at least one of a CU and a TU.

[0428] In this case, split_flag may be a flag indicating whether the CU is partitioned in a quadtree form.

[0429] When the size of the block falls within a specific range, it is possible to divide it only in the binary tree form or the ternary tree form. For example, the specific range can be defined by at least one of the maximum block size and the minimum block size that can be divided only in the binary tree form or the ternary tree form.

[0430] Information indicating the maximum block size and / or the minimum block size that can be divided only in a binary tree form or in a ternary tree form may be signaled from the encoding apparatus 100 to the decoding apparatus 200 through a bitstream. In addition, this information may be signaled for at least one of units such as a sequence, a picture, and a slice (or segment).

[0431] Alternatively, the maximum block size and / or the minimum block size may be a fixed size predefined by the encoding device 100 and the decoding device 200. For example, when the size of the block is greater than 8×8 and less than 16×16, only splitting in a binary tree form is possible. In this case, split_flag may be a flag indicating whether to perform splitting in a binary tree form or a ternary tree form.

[0432] The above description of partitioning in a quadtree form may be equally applied to a binary tree form and / or a ternary tree form.

[0433] The partitioning of a block may be limited by the previous partitioning. For example, when a block is partitioned in a specific binary tree format and multiple sub-blocks are generated from the partitioning, each sub-block can be partitioned only in a specific tree format. Here, the specific tree format can be at least one of a binary tree format, a ternary tree format, and a quadtree format.

[0434] When the horizontal size or the vertical size of the partition block is a size that cannot be further divided, the above-mentioned indicator may not be signaled.

[0435] Figure 7 is a diagram for explaining an embodiment of an intra prediction process.

[0436] from Figure 7 The arrow extending radially from the center of the diagram in indicates the prediction direction of the intra prediction mode. Furthermore, numbers appearing near the arrows indicate examples of mode values ​​assigned to the intra prediction mode or the prediction direction of the intra prediction mode.

[0437] exist Figure 7 In the example, number 0 may indicate planar mode, which is a non-directional intra prediction mode, and number 1 may indicate DC mode, which is a non-directional intra prediction mode.

[0438] Intra-frame encoding and / or decoding may be performed using reference samples of neighboring blocks of a target block. The neighboring blocks may be reconstructed neighboring blocks. The reference samples may represent neighboring samples.

[0439] For example, intra encoding and / or decoding may be performed using values ​​of reference samples included in the reconstructed neighboring block or encoding parameters of the reconstructed neighboring block.

[0440] The encoding device 100 and / or the decoding device 200 may generate a prediction block by performing intra-frame prediction on the target block based on information about samples in the target image. When intra-frame prediction is performed, the encoding device 100 and / or the decoding device 200 may generate a prediction block for the target block by performing intra-frame prediction based on information about samples in the target image. When intra-frame prediction is performed, the encoding device 100 and / or the decoding device 200 may perform directional prediction and / or non-directional prediction based on at least one reconstructed reference sample.

[0441] The prediction block may be a block generated as a result of performing intra prediction. The prediction block may correspond to at least one of a CU, a PU, and a TU.

[0442] The prediction block may have a size corresponding to at least one of a CU, a PU, and a TU. The prediction block may have a square shape of 2N×2N or N×N. The N×N size may include sizes of 4×4, 8×8, 16×16, 32×32, 64×64, etc.

[0443] Alternatively, the prediction block may be a square block of size 2×2, 4×4, 8×8, 16×16, 32×32, 64×64, etc. or a rectangular block of size 2×8, 4×8, 2×16, 4×16, 8×16, etc.

[0444] Intra-frame prediction may be performed considering the intra-frame prediction mode used for the target block. The number of intra-frame prediction modes that the target block can have may be a predefined fixed value, or may be a value determined differently according to the properties of the prediction block. For example, the properties of the prediction block may include the size of the prediction block, the type of the prediction block, etc. In addition, the properties of the prediction block may indicate encoding parameters used for the prediction block.

[0445] For example, regardless of the size of the prediction block, the number of intra prediction modes may be fixed to N. Alternatively, the number of intra prediction modes may be 3, 5, 9, 17, 34, 35, 36, 65, 67, or 95, for example.

[0446] The intra prediction mode can be a non-directional mode or a directional mode.

[0447] For example, intra prediction modes may include Figure 7 The numbers 0 to 66 shown in the figure correspond to two non-directional modes and 65 directional modes.

[0448] For example, in the case of using a specific intra prediction method, the intra prediction mode may include Figure 7 The numbers -14 to 80 shown correspond to two non-directional modes and 93 directional modes.

[0449] The two non-directional modes may include a DC mode and a planar mode.

[0450] A directional mode may be a prediction mode with a specific direction or a specific angle. A directional mode may also be referred to as an "angle mode."

[0451] The intra-frame prediction mode may be represented by at least one of a mode number, a mode value, a mode angle, and a mode direction. In other words, the terms "intra-frame prediction mode (mode) number", "intra-frame prediction mode (mode) value", "intra-frame prediction mode (mode) angle", and "intra-frame prediction mode (mode) direction" may be used to have the same meaning and may be used interchangeably.

[0452] The number of intra prediction modes may be M. The value of M may be 1 or greater. In other words, the number of intra prediction modes may be M, where M includes the number of non-directional modes and the number of directional modes.

[0453] Regardless of the size of the block and / or the color components, the number of intra prediction modes may be fixed to M. For example, the number of intra prediction modes may be fixed to any one of 35 and 67 regardless of the size of the block.

[0454] Alternatively, the number of intra prediction modes may differ according to the shape, size, and / or type of color component of a block.

[0455] For example, in Figure 7 , the directional prediction mode shown by the dotted line can be applied only to prediction for non-square blocks.

[0456] For example, the larger the block size, the greater the number of intra-frame prediction modes. Alternatively, the larger the block size, the fewer the number of intra-frame prediction modes. When the block size is 4×4 or 8×8, the number of intra-frame prediction modes may be 67. When the block size is 16×16, the number of intra-frame prediction modes may be 35. When the block size is 32×32, the number of intra-frame prediction modes may be 19. When the block size is 64×64, the number of intra-frame prediction modes may be 7.

[0457] For example, the number of intra prediction modes may differ depending on whether the color component is a luma signal or a chroma signal. Alternatively, the number of intra prediction modes corresponding to a luma component block may be greater than the number of intra prediction modes corresponding to a chroma component block.

[0458] For example, in vertical mode with a mode value of 50, prediction may be performed in the vertical direction based on the pixel values ​​of the reference samples. For example, in horizontal mode with a mode value of 18, prediction may be performed in the horizontal direction based on the pixel values ​​of the reference samples.

[0459] Even in a directional mode other than the above-described modes, the encoding apparatus 100 and the decoding apparatus 200 may perform intra prediction on a target unit using reference samples according to an angle corresponding to the directional mode.

[0460] An intra-frame prediction mode located to the right relative to the vertical mode may be referred to as a "vertical-right mode". An intra-frame prediction mode located below the horizontal mode may be referred to as a "horizontal-below mode". For example, in Figure 7 , the intra prediction mode whose mode value is one of 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, and 66 may be the vertical-right mode. The intra prediction mode whose mode value is one of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, and 17 may be the horizontal-down mode.

[0461] The non-directional mode may include a DC mode and a planar mode. For example, the value of the DC mode may be 1. The value of the planar mode may be 0.

[0462] The directional mode may include an angular mode. Among the plurality of intra prediction modes, the remaining modes except the DC mode and the planar mode may be directional modes.

[0463] When the intra prediction mode is the DC mode, the prediction block may be generated based on the average value of the pixel values ​​of the plurality of reference pixels. For example, the pixel values ​​of the prediction block may be determined based on the average value of the pixel values ​​of the plurality of reference pixels.

[0464] The number of intra prediction modes and the mode values ​​of each intra prediction mode described above are merely exemplary and may be defined differently depending on embodiments, implementations, and / or requirements.

[0465] In order to perform intra prediction on a target block, a step of checking whether samples included in a reconstructed neighboring block can be used as reference samples for the target block may be performed. When a sample that cannot be used as a reference sample for the target block exists among the samples in the neighboring block, a value generated by interpolation and / or replication using at least one sample value among the samples included in the reconstructed neighboring block may replace the sample value of the sample that cannot be used as a reference sample. When the value generated by replication and / or interpolation replaces the sample value of the existing sample, the sample may be used as a reference sample for the target block.

[0466] When intra prediction is used, a filter may be applied to at least one of a reference sample and a prediction sample based on at least one of a size of a target block and an intra prediction mode.

[0467] The type of filter to be applied to at least one of the reference sample and the prediction sample may differ according to at least one of the intra prediction mode of the target block, the size of the target block, and the shape of the target block. The type of filter may be classified according to one or more of the length of the filter tap, the value of the filter coefficient, and the filter strength. The length of the filter tap may indicate the number of filter taps. In addition, the number of filter taps may indicate the length of the filter.

[0468] When the intra prediction mode is the planar mode, the sample value of the predicted target block may be generated by using a weighted sum of the upper reference sample of the target block, the left reference sample of the target block, the upper right reference sample of the target block, and the lower left reference sample of the target block according to the position of the predicted target sample in the prediction block when generating the prediction block of the target block.

[0469] When the intra prediction mode is the DC mode, an average value of reference samples above the target block and reference samples to the left of the target block may be used when generating a prediction block for the target block. Furthermore, filtering using the values ​​of the reference samples may be performed on specific rows or columns in the target block. The specific rows may be one or more rows above the reference sample. The specific columns may be one or more columns to the left of the reference sample.

[0470] When the intra prediction mode is a directional mode, the prediction block may be generated using the upper reference sample, the left reference sample, the upper right reference sample, and / or the lower left reference sample of the target block.

[0471] In order to generate the above-mentioned prediction samples, real number-based interpolation may be performed.

[0472] The intra prediction mode of the target block may be predicted from intra prediction modes of neighboring blocks adjacent to the target block, and information used for the prediction may be entropy encoded / decoded.

[0473] For example, when the intra prediction modes of the target block and the neighboring block are identical to each other, a predefined flag may be used to signal that the intra prediction modes of the target block and the neighboring block are the same.

[0474] For example, an indicator indicating the same intra prediction mode as that of the target block among the intra prediction modes of a plurality of neighboring blocks may be signaled.

[0475] When the intra prediction modes of the target block and the neighboring blocks are different from each other, information about the intra prediction mode of the target block may be encoded and / or decoded using entropy encoding and / or entropy decoding.

[0476] Figure 8 is a diagram showing reference samples used in an intra prediction process.

[0477] The reconstructed reference samples used for intra prediction of the target block may include a lower left reference sample, a left reference sample, an upper left corner reference sample, an upper reference sample, and an upper right reference sample.

[0478] For example, a left reference sample may represent a reconstructed reference pixel adjacent to the left side of the target block. An upper reference sample may represent a reconstructed reference pixel adjacent to the top of the target block. An upper-left corner reference sample may represent a reconstructed reference pixel located at the upper-left corner of the target block. A lower-left corner reference sample may represent a reference sample located below a left sample line among samples located on the same line as a left sample line composed of left reference samples. An upper-right corner reference sample may represent a reference sample located to the right of an upper sample line among samples located on the same line as an upper sample line composed of upper reference samples.

[0479] When the size of the target block is N×N, the numbers of the lower left reference sample, the left reference sample, the upper reference sample, and the upper right reference sample may all be N.

[0480] By performing intra prediction on the target block, a prediction block can be generated. The process of generating the prediction block may include determining the values ​​of the pixels in the prediction block. The target block and the prediction block may be the same size.

[0481] The reference samples used for intra-frame prediction of the target block may change according to the intra-frame prediction mode of the target block. The direction of the intra-frame prediction mode may indicate the dependency between the reference samples and the pixels of the prediction block. For example, the value of a specified reference sample may be used as the value of one or more specified pixels in the prediction block. In this case, the specified reference sample and the one or more specified pixels in the prediction block may be samples and pixels located on a straight line along the direction of the intra-frame prediction mode. In other words, the value of the specified reference sample may be copied as the value of a pixel located in a direction opposite to the direction of the intra-frame prediction mode. Alternatively, the value of a pixel in the prediction block may be the value of a reference sample located in the direction of the intra-frame prediction mode relative to the position of the pixel.

[0482] In this example, when the intra prediction mode of the target block is vertical mode, the upper reference sample can be used for intra prediction. When the intra prediction mode is vertical mode, the value of a pixel in the prediction block can be the value of a reference sample located vertically above the position of the pixel. Therefore, the upper reference sample adjacent to the top of the target block can be used for intra prediction. In addition, the value of the pixels in a row of the prediction block can be the same as the value of the pixels of the upper reference sample.

[0483] In this example, when the intra prediction mode of the target block is horizontal, the left reference sample can be used for intra prediction. When the intra prediction mode is horizontal, the value of a pixel in the prediction block can be the value of a reference sample located horizontally to the left of the pixel. Therefore, the left reference sample adjacent to the left side of the target block can be used for intra prediction. In addition, the value of a pixel in a column of the prediction block can be the same as the value of the pixel of the left reference sample.

[0484] In an example, when the mode value of the intra prediction mode of the current block is 34, at least some of the left reference samples, the upper left reference samples, and at least some of the upper reference samples may be used for intra prediction. When the mode value of the intra prediction mode is 34, the value of a pixel in the prediction block may be the value of a reference sample diagonally located at the upper left corner of the pixel.

[0485] Also, in the case of an intra prediction mode whose mode value is a value ranging from 52 to 66, at least a portion of the upper right reference samples may be used for intra prediction.

[0486] Also, in the case of an intra prediction mode whose mode value is a value ranging from 2 to 17, at least a portion of the lower left reference samples may be used for intra prediction.

[0487] Also, in the case of an intra prediction mode whose mode value is a value ranging from 19 to 49, the top left corner reference sample may be used for intra prediction.

[0488] The number of reference samples used to determine the pixel value of one pixel in the prediction block may be 1 or 2 or more.

[0489] As described above, the pixel value of a pixel in the prediction block may be determined based on the position of the pixel and the position of the reference sample indicated by the direction of the intra prediction mode. When the position of the pixel and the position of the reference sample indicated by the direction of the intra prediction mode are integer positions, the value of one reference sample indicated by the integer position may be used to determine the pixel value of the pixel in the prediction block.

[0490] When the position of a pixel and the position of a reference sample indicated by the direction of the intra-frame prediction mode are not integer positions, an interpolated reference sample based on the two reference samples closest to the position of the reference sample may be generated. The value of the interpolated reference sample may be used to determine the pixel value of a pixel in the prediction block. In other words, when the position of a pixel in the prediction block and the position of a reference sample indicated by the direction of the intra-frame prediction mode indicate a position between two reference samples, an interpolated value based on the values ​​of these two samples may be generated.

[0491] The prediction block generated via prediction may be different from the original target block. In other words, there may be a prediction error, which is the difference between the target block and the prediction block, and there may also be a prediction error between the pixels of the target block and the pixels of the prediction block.

[0492] Hereinafter, the terms “difference”, “error” and “residual” may be used to have the same meaning and may be used interchangeably with each other.

[0493] For example, in the case of directional intra prediction, the longer the distance between the pixels of the prediction block and the reference samples, the greater the prediction error that may occur. This prediction error can lead to discontinuities between the generated prediction block and neighboring blocks.

[0494] To reduce prediction error, a filtering operation may be applied to the prediction block. The filtering operation may be configured to adaptively apply a filter to areas of the prediction block that are considered to have a large prediction error. For example, the areas considered to have a large prediction error may be boundaries of the prediction block. Furthermore, the areas of the prediction block considered to have a large prediction error may vary depending on the intra-frame prediction mode, and the characteristics of the filter may also vary depending on the intra-frame prediction mode.

[0495] like Figure 8 As shown, for intra prediction of a target block, at least one of reference lines 0 to 3 may be used.

[0496] exist Figure 8 Each reference line in may indicate a reference sample line including one or more reference samples. When the number of a reference line is smaller, a reference sample line closer to the target block may be indicated.

[0497] The samples in segments A and F may be obtained by padding using the samples in segments B and E that are closest to the target block, rather than from the reconstructed neighboring blocks.

[0498] Index information indicating a reference sample line to be used for intra prediction of a target block may be signaled. The index information may indicate a reference sample line to be used for intra prediction of the target block among a plurality of reference sample lines. For example, the index information may have a value corresponding to any one of 0 to 3.

[0499] When the upper boundary of the target block is the boundary of the CTU, only reference sample line 0 may be available. Therefore, in this case, index information may not be signaled. When additional reference sample lines other than reference sample line 0 are used, filtering of the prediction block, which will be described later, may not be performed.

[0500] In case of inter-color intra prediction, a prediction block of a target block of a second color component may be generated based on a corresponding reconstructed block of a first color component.

[0501] For example, the first color component may be a luminance component, and the second color component may be a chrominance component.

[0502] To perform inter-color intra prediction, parameters of a linear model between the first color component and the second color component may be derived based on the template.

[0503] The template may include a reference sample above the target block (upper reference sample) and / or a reference sample to the left of the target block (left reference sample), and may include an upper reference sample and / or a left reference sample of a reconstructed block of a first color component corresponding to the reference samples.

[0504] For example, the parameters of the linear model may be derived using the following values: 1) the value of the sample of the first color component having the maximum value among the samples in the template, 2) the value of the sample of the second color component corresponding to the sample of the first color component, 3) the value of the sample of the first color component having the minimum value among the samples in the template, and 4) the value of the sample of the second color component corresponding to the sample of the first color component.

[0505] When the parameters of the linear model are derived, a prediction block of the target block may be generated by applying the corresponding reconstructed block to the linear model.

[0506] Depending on the image format, subsampling may be performed on samples adjacent to a reconstructed block of a first color component and the corresponding reconstructed block of the first color component. For example, when one sample of a second color component corresponds to four samples of a first color component, one corresponding sample may be calculated by subsampling the four samples of the first color component. When subsampling is performed, linear model parameter derivation and inter-color intra prediction may be performed based on the subsampled corresponding samples.

[0507] Information on whether to perform inter-color intra prediction and / or the range of the template may be signaled in the intra prediction mode.

[0508] The target block may be partitioned into two or four sub-blocks in a horizontal direction and / or a vertical direction.

[0509] The subblocks generated by partitioning can be reconstructed sequentially. That is, when intra prediction is performed on each subblock, a subprediction block for the subblock can be generated. Furthermore, when inverse quantization (dequantization) and / or inverse transform is performed on each subblock, a subresidual block for the corresponding subblock can be generated. A reconstructed subblock can be generated by adding the subprediction block to the subresidual block. The reconstructed subblock can be used as a reference sample for intra prediction of a subblock with the next higher priority.

[0510] A subblock may be a block including a specific number (e.g., 16) or more samples. For example, when the target block is an 8×4 block or a 4×8 block, the target block may be partitioned into two subblocks. In addition, when the target block is a 4×4 block, the target block cannot be partitioned into subblocks. When the target block has another size, the target block may be partitioned into four subblocks.

[0511] Information on whether to perform intra prediction based on these subblocks and / or information on the partition direction (horizontal direction or vertical direction) may be signaled.

[0512] This subblock-based intra prediction may be restricted such that it is performed only when reference sample line 0 is used. When subblock-based intra prediction is performed, filtering of the prediction block, which will be described below, may not be performed.

[0513] A final prediction block may be generated by performing filtering on a prediction block generated through intra prediction.

[0514] Filtering may be performed by applying specific weights to filtering target samples, left reference samples, upper reference samples, and / or upper-left reference samples that are targets to be filtered.

[0515] The weight and / or reference samples (eg, range of reference samples, position of reference samples, etc.) used for filtering may be determined based on at least one of block size, intra prediction mode, and position of filtering target samples in a prediction block.

[0516] For example, filtering may be performed only in specific intra prediction modes (eg, DC mode, planar mode, vertical mode, horizontal mode, diagonal mode, and / or adjacent diagonal mode).

[0517] The adjacent diagonal pattern may be a pattern with a number obtained by adding k to the diagonal pattern number, or may be a pattern with a number obtained by subtracting k from the diagonal pattern number. In other words, the number of the adjacent diagonal pattern may be the sum of the diagonal pattern number and k, or may be the difference between the diagonal pattern number and k. For example, k may be a positive integer of 8 or less.

[0518] The intra prediction mode of the target block may be derived using intra prediction modes of neighboring blocks occurring around the target block, and such derived intra prediction mode may be entropy encoded and / or entropy decoded.

[0519] For example, when the intra prediction mode of the target block is the same as that of the neighboring block, specific flag information may be used to signal information indicating that the intra prediction mode of the target block is the same as that of the neighboring block.

[0520] Also, for example, indicator information of a neighboring block having the same intra prediction mode as the intra prediction mode of the target block among intra prediction modes of a plurality of neighboring blocks may be signaled.

[0521] For example, when the intra-frame prediction mode of the target block is different from the intra-frame prediction mode of the neighboring block, entropy encoding and / or entropy decoding can be performed on information about the intra-frame prediction mode of the target block by performing entropy encoding and / or entropy decoding based on the intra-frame prediction mode of the neighboring block.

[0522] Figure 9 is a diagram for explaining an embodiment of an inter-frame prediction process.

[0523] Figure 9 The rectangle shown in can represent an image (or picture). Figure 9 In the example, an arrow may indicate a prediction direction. An arrow pointing from a first picture to a second picture indicates that the second picture refers to the first picture. That is, each picture may be encoded and / or decoded according to the prediction direction.

[0524] Images can be classified into intra-frame pictures (I pictures), uni-predictive pictures or predictive coded pictures (P pictures), and bi-predictive pictures or bi-predictive coded pictures (B pictures) according to the coding type. Each picture can be encoded and / or decoded according to the coding type of each picture.

[0525] When the target image to be encoded is an I picture, the target image can be encoded using data included in the image itself without performing inter-frame prediction with reference to other images. For example, the I picture can be encoded only via intra-frame prediction.

[0526] When the target image is a P picture, the target image may be encoded through inter-frame prediction using a reference picture existing in one direction. Here, the one direction may be a forward direction or a backward direction.

[0527] When the target image is a B picture, the image may be encoded by inter-frame prediction using reference pictures existing in both directions, or may be encoded by inter-frame prediction using reference pictures existing in one of the forward direction and the backward direction. Here, the two directions may be the forward direction and the backward direction.

[0528] P-pictures and B-pictures encoded and / or decoded using reference pictures may be regarded as images using inter-frame prediction.

[0529] Hereinafter, inter prediction in inter mode according to an embodiment will be described in detail.

[0530] Inter-frame prediction or motion compensation may be performed using a reference image and motion information.

[0531] In the inter mode, the encoding apparatus 100 may perform inter prediction and / or motion compensation on the target block, and the decoding apparatus 200 may perform inter prediction and / or motion compensation corresponding to the inter prediction and / or motion compensation performed by the encoding apparatus 100 on the target block.

[0532] The motion information of the target block may be derived separately during inter prediction by the encoding apparatus 100 and the decoding apparatus 200. The motion information may be derived using the motion information of the reconstructed neighboring block, the motion information of the col block, and / or the motion information of the blocks adjacent to the col block.

[0533] For example, the encoding apparatus 100 or the decoding apparatus 200 may perform prediction and / or motion compensation by using the motion information of the spatial candidate and / or the temporal candidate as the motion information of the target block. The target block may represent a PU and / or a PU partition.

[0534] The spatial candidate may be a reconstructed block that is spatially adjacent to the target block.

[0535] The temporal candidate may be a reconstructed block corresponding to the target block in a previously reconstructed co-located picture (col picture).

[0536] In inter-frame prediction, the encoding device 100 and the decoding device 200 can improve encoding efficiency and decoding efficiency by utilizing motion information of spatial candidates and / or temporal candidates. The motion information of the spatial candidate may be referred to as "spatial motion information." The motion information of the temporal candidate may be referred to as "temporal motion information."

[0537] Hereinafter, the motion information of a spatial candidate may be the motion information of a PU including the spatial candidate. The motion information of a temporal candidate may be the motion information of a PU including the temporal candidate. The motion information of a candidate block may be the motion information of a PU including the candidate block.

[0538] Inter prediction may be performed using reference pictures.

[0539] The reference picture may be at least one of a picture before the target picture and a picture after the target picture.The reference picture may be an image used for prediction of the target block.

[0540] In inter prediction, a region in a reference picture may be specified using a reference picture index (or refIdx) indicating a reference picture, a motion vector to be described later, etc. Here, the region specified in the reference picture may indicate a reference block.

[0541] Inter prediction can select a reference picture and further select a reference block corresponding to the target block from the reference picture. In addition, inter prediction can use the selected reference block to generate a prediction block for the target block.

[0542] Motion information may be derived by each of the encoding apparatus 100 and the decoding apparatus 200 during inter prediction.

[0543] A spatial candidate may be a block that 1) exists in the target picture 2) has been previously reconstructed via encoding and / or decoding and 3) is adjacent to the target block or is located at a corner of the target block. Here, a "block located at a corner of the target block" may be a block vertically adjacent to a neighboring block horizontally adjacent to the target block, or a block horizontally adjacent to a neighboring block vertically adjacent to the target block. In addition, a "block located at a corner of the target block" may have the same meaning as a "block adjacent to a corner of the target block." The meaning of a "block located at a corner of the target block" may be included in the meaning of a "block adjacent to the target block."

[0544] For example, the spatial candidate can be a reconstructed block located to the left of the target block, a reconstructed block located above the target block, a reconstructed block located at the lower left corner of the target block, a reconstructed block located at the upper right corner of the target block, or a reconstructed block located at the upper left corner of the target block.

[0545] Each of the encoding apparatus 100 and the decoding apparatus 200 may identify a block existing in the col picture at a position spatially corresponding to the target block. The position of the target block in the target picture and the position of the identified block in the col picture may correspond to each other.

[0546] Each of the encoding apparatus 100 and the decoding apparatus 200 may determine a col block existing at a predefined relative position with respect to the identified block as a temporal candidate. The predefined relative position may be a position existing inside and / or outside the identified block.

[0547] For example, the col block may include a first col block and a second col block. When the coordinates of the identified block are (xP, yP) and the size of the identified block is represented by (nPSW, nPSH), the first col block may be a block located at the coordinates (xP+nPSW, yP+nPSH). The second col block may be a block located at the coordinates (xP+(nPSW>>1), yP+(nPSH>>1)). When the first col block is unavailable, the second col block may be selectively used.

[0548] The motion vector of the target block may be determined based on the motion vector of the col block. Each of the encoding apparatus 100 and the decoding apparatus 200 may scale the motion vector of the col block. The scaled motion vector of the col block may be used as the motion vector of the target block. In addition, the motion vector of the motion information of the temporal candidate stored in the list may be the scaled motion vector.

[0549] The ratio of the motion vector of the target block to the motion vector of the col block may be the same as the ratio of the first temporal distance to the second temporal distance. The first temporal distance may be the distance between the reference picture and the target picture of the target block. The second temporal distance may be the distance between the reference picture and the col picture of the col block.

[0550] The scheme for deriving motion information may vary depending on the inter-frame prediction mode of the target block. For example, as inter-frame prediction modes applied to inter-frame prediction, there may be Advanced Motion Vector Predictor (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 Inter mode, Current picture reference mode, etc. Merge mode may also be referred to as "motion merge mode." Each mode will be described in detail below.

[0551] 1) AMVP model

[0552] When using the AMVP mode, the encoding device 100 may search for similar blocks in the vicinity of the target block. The encoding device 100 may obtain a prediction block by performing prediction on the target block using the motion information of the found similar blocks. The encoding device 100 may encode a residual block that is the difference between the target block and the prediction block.

[0553] 1-1) Create a list of predicted motion vector candidates

[0554] When the AMVP mode is used as the prediction mode, each of the encoding device 100 and the decoding device 200 can create a list of prediction motion vector candidates using the motion vector of the spatial candidate, the motion vector of the temporal candidate, and the zero vector. The prediction motion vector candidate list may include one or more prediction motion vector candidates. At least one of the motion vector of the spatial candidate, the motion vector of the temporal candidate, and the zero vector may be determined and used as the prediction motion vector candidate.

[0555] Hereinafter, the terms "prediction motion vector (candidate)" and "motion vector (candidate)" may be used to have the same meaning and may be used interchangeably with each other.

[0556] Hereinafter, the terms “predicted motion vector candidate” and “AMVP candidate” may be used to have the same meaning and may be used interchangeably with each other.

[0557] Hereinafter, the terms “motion vector prediction candidate list” and “AMVP candidate list” may be used to have the same meaning and may be used interchangeably with each other.

[0558] The spatial candidate may include a reconstructed spatial neighboring block. In other words, the motion vector of the reconstructed neighboring block may be referred to as a "spatial prediction motion vector candidate."

[0559] The temporal candidate may include the col block and blocks adjacent to the col block. In other words, the motion vector of the col block or the motion vector of the block adjacent to the col block may be referred to as a "temporal prediction motion vector candidate".

[0560] The zero vector may be the (0,0) motion vector.

[0561] The predicted motion vector candidate may be a motion vector predictor for predicting a motion vector. In addition, in the encoding apparatus 100, each predicted motion vector candidate may be an initial search position for a motion vector.

[0562] 1-2) Searching for motion vector using a list of predicted motion vector candidates

[0563] The encoding device 100 may determine a motion vector to be used for encoding the target block within the search range using the list of predicted motion vector candidates. In addition, the encoding device 100 may determine a predicted motion vector candidate to be used as the predicted motion vector of the target block from among the predicted motion vector candidates in the predicted motion vector candidate list.

[0564] The motion vector to be used for encoding the target block may be a motion vector that can be encoded at a minimum cost.

[0565] Also, the encoding apparatus 100 may determine whether to encode the target block using the AMVP mode.

[0566] 1-3) Transmission of inter-frame prediction information

[0567] The encoding apparatus 100 may generate a bitstream including inter prediction information required for inter prediction, and the decoding apparatus 200 may perform inter prediction on a target block using the inter prediction information of the bitstream.

[0568] The inter prediction information may include 1) mode information indicating whether the AMVP mode is used, 2) a prediction motion vector index, 3) a motion vector difference (MVD), 4) a reference direction, and 5) a reference picture index.

[0569] Hereinafter, the terms “predicted motion vector index” and “AMVP index” may be used to have the same meaning and may be used interchangeably with each other.

[0570] Furthermore, the inter prediction information may include a residual signal.

[0571] When the mode information indicates that the AMVP mode is used, the decoding apparatus 200 may acquire a predicted motion vector index, an MVD, a reference direction, and a reference picture index from a bitstream through entropy decoding.

[0572] The predicted motion vector index may indicate a predicted motion vector candidate to be used for prediction of the target block among the predicted motion vector candidates included in the predicted motion vector candidate list.

[0573] 1-4) Inter-frame prediction in AMVP mode using inter-frame prediction information

[0574] The decoding apparatus 200 may derive a prediction motion vector candidate using the prediction motion vector candidate list, and may determine motion information of a target block based on the derived prediction motion vector candidate.

[0575] The decoding apparatus 200 may determine a motion vector candidate for the target block from among the predicted motion vector candidates included in the predicted motion vector candidate list using the predicted motion vector index. The decoding apparatus 200 may select the predicted motion vector candidate indicated by the predicted motion vector index from among the predicted motion vector candidates included in the predicted motion vector candidate list as the predicted motion vector of the target block.

[0576] The encoding apparatus 100 may generate an entropy-coded predicted motion vector index by applying entropy coding to the predicted motion vector index, and may generate a bitstream including the entropy-coded predicted motion vector index. The entropy-coded predicted motion vector index may be signaled from the encoding apparatus 100 to the decoding apparatus 200 through the bitstream. The decoding apparatus 200 may extract the entropy-coded predicted motion vector index from the bitstream, and may obtain the predicted motion vector index by applying entropy decoding to the entropy-coded predicted motion vector index.

[0577] The motion vector actually used for inter-frame prediction of the target block may not match the predicted motion vector. MVD may be used to indicate the difference between the motion vector actually used for inter-frame prediction of the target block and the predicted motion vector. The encoding apparatus 100 may derive a predicted motion vector similar to the motion vector actually used for inter-frame prediction of the target block in order to use the smallest possible MVD.

[0578] The motion vector difference (MVD) may be the difference between the motion vector of the target block and the predicted motion vector. The encoding apparatus 100 may calculate the MVD and may generate an entropy-encoded MVD by applying entropy encoding to the MVD. The encoding apparatus 100 may generate a bitstream including the entropy-encoded MVD.

[0579] The MVD may be transmitted from the encoding apparatus 100 to the decoding apparatus 200 through a bitstream. The decoding apparatus 200 may extract the entropy-encoded MVD from the bitstream, and may obtain the MVD by applying entropy decoding to the entropy-encoded MVD.

[0580] The decoding apparatus 200 may derive a motion vector of the target block by summing the MVD and the predicted motion vector. In other words, the motion vector of the target block derived by the decoding apparatus 200 may be the sum of the MVD and the motion vector candidate.

[0581] In addition, the encoding device 100 may generate entropy-encoded MVD resolution information by applying entropy encoding to the calculated MVD resolution information, and may generate a bitstream including the entropy-encoded MVD resolution information. The decoding device 200 may extract the entropy-encoded MVD resolution information from the bitstream, and may obtain the MVD resolution information by applying entropy decoding to the entropy-encoded MVD resolution information. The decoding device 200 may use the MVD resolution information to adjust the resolution of the MVD.

[0582] In addition, the encoding apparatus 100 may calculate MVD based on an affine model. The decoding apparatus 200 may derive an affine controlled motion vector of a target block by summing MVD and an affine controlled motion vector candidate, and may derive a motion vector of a subblock using the affine controlled motion vector.

[0583] The reference direction may indicate a list of reference pictures to be used for prediction of the target block. For example, the reference direction may indicate one of reference picture list L0 and reference picture list L1.

[0584] The reference direction only indicates the reference picture list to be used for prediction of the target block and may not mean that the direction of the reference picture is limited to the forward direction or the backward direction. In other words, each of the reference picture lists L0 and L1 may include pictures in the forward direction and / or the backward direction.

[0585] A unidirectional reference direction may mean that a single reference picture list is used. A bidirectional reference direction may mean that two reference picture lists are used. In other words, the reference direction may indicate one of the following cases: a case where only reference picture list L0 is used, a case where only reference picture list L1 is used, or a case where two reference picture lists are used.

[0586] The reference picture index may indicate a reference picture used to predict the target block among the reference pictures in the reference picture list. The encoding apparatus 100 may generate an entropy-coded reference picture index by applying entropy coding to the reference picture index, and may generate a bitstream including the entropy-coded reference picture index. The entropy-coded reference picture index may be signaled from the encoding apparatus 100 to the decoding apparatus 200 via the bitstream. The decoding apparatus 200 may extract the entropy-coded reference picture index from the bitstream, and may obtain the reference picture index by applying entropy decoding to the entropy-coded reference picture index.

[0587] When two reference picture lists are used to predict a target block, a single reference picture index and a single motion vector can be used for each reference picture list. Furthermore, when two reference picture lists are used to predict a target block, two prediction blocks can be specified for the target block. For example, the (final) prediction block for the target block can be generated using the average or weighted sum of the two prediction blocks for the target block.

[0588] The motion vector of the target block may be derived through the predicted motion vector index, MVD, reference direction, and reference picture index.

[0589] The decoding apparatus 200 may generate a prediction block for the target block based on the derived motion vector and the reference picture index. For example, the prediction block may be a reference block indicated by the derived motion vector in the reference picture indicated by the reference picture index.

[0590] Since the predicted motion vector index and the MVD are encoded but the motion vector itself of the target block is not encoded, the number of bits transmitted from the encoding apparatus 100 to the decoding apparatus 200 may be reduced and encoding efficiency may be improved.

[0591] For the target block, the motion information of reconstructed neighboring blocks can be used. In certain inter-frame prediction modes, the encoding device 100 may not separately encode the actual motion information of the target block. Instead of encoding the motion information of the target block, additional information may be encoded, where the additional information enables the motion information of the target block to be derived using the motion information of the reconstructed neighboring blocks. Since this additional information is encoded, the number of bits transmitted to the decoding device 200 can be reduced, and encoding efficiency can be improved.

[0592] For example, as an inter-frame prediction mode in which the motion information of the target block is not directly encoded, a skip mode and / or a merge mode may exist. Here, each of the encoding device 100 and the decoding device 200 may use an identifier and / or an index indicating a unit among the reconstructed neighboring units whose motion information is to be used as the motion information of the target unit.

[0593] 2) Merge mode

[0594] Merge is a scheme for deriving motion information of a target block. The term "merge" may mean combining the motions of multiple blocks. "Merge" may also mean that the motion information of one block is also applied to other blocks. In other words, merge mode may be a mode in which the motion information of a target block is derived from the motion information of neighboring blocks.

[0595] When using merge mode, the encoding device 100 may use the motion information of the spatial candidate and / or the motion information of the temporal candidate to predict the motion information of the target block. The spatial candidate may include a reconstructed spatial neighboring block that is spatially adjacent to the target block. The spatial neighboring block may include a left neighboring block and an upper neighboring block. The temporal candidate may include a col block. The terms "spatial candidate" and "spatial merge candidate" may be used to have the same meaning and may be used interchangeably with each other. The terms "temporal candidate" and "temporal merge candidate" may be used to have the same meaning and may be used interchangeably with each other.

[0596] The encoding apparatus 100 may obtain a prediction block through prediction. The encoding apparatus 100 may encode a residual block that is a difference between the target block and the prediction block.

[0597] 2-1) Create a merge candidate list

[0598] When using merge mode, each of the encoding device 100 and the decoding device 200 can use the motion information of the spatial candidate and / or the motion information of the temporal candidate to create a merge candidate list. The motion information may include 1) a motion vector, 2) a reference picture index, and 3) a reference direction. The reference direction can be unidirectional or bidirectional. The reference direction may indicate an inter-frame prediction indicator.

[0599] The merge candidate list may include a merge candidate. The merge candidate may be motion information. In other words, the merge candidate list may be a list storing multiple pieces of motion information.

[0600] The merge candidate may be a plurality of pieces of motion information of temporal candidates and / or spatial candidates. In other words, the merge candidate list may include motion information of temporal candidates and / or spatial candidates, etc.

[0601] In addition, the merge candidate list may include new merge candidates generated by combining merge candidates already in the merge candidate list. In other words, the merge candidate list may include new motion information generated by combining multiple pieces of motion information previously in the merge candidate list.

[0602] In addition, the merge candidate list may include a history-based merge candidate. The history-based merge candidate may be motion information of a block that was encoded and / or decoded before the target block.

[0603] Furthermore, the merge candidate list may include a merge candidate based on an average of two merge candidates.

[0604] A merge candidate may be a specific mode for deriving inter-frame prediction information. The merge candidate may be information indicating a specific mode for deriving inter-frame prediction information. Inter-frame prediction information for a target block may be derived according to the specific mode indicated by the merge candidate. Furthermore, the specific mode may include a process for deriving a series of inter-frame prediction information. This specific mode may be an inter-frame prediction information deriving mode or a motion information deriving mode.

[0605] Inter prediction information of the target block may be derived according to a mode indicated by a merge candidate selected from among merge candidates in the merge candidate list by a merge index.

[0606] For example, the motion information derivation mode in the merge candidate list may be at least one of the following modes: 1) a motion information derivation mode for a sub-block unit and 2) an affine motion information derivation mode.

[0607] In addition, the merge candidate list may include motion information of a zero vector. A zero vector may also be referred to as a "zero merge candidate."

[0608] In other words, the multiple motion information in the merge candidate list can be at least one of the following information: 1) motion information of spatial candidates, 2) motion information of temporal candidates, 3) motion information generated by combining multiple motion information previously existing in the merge candidate list, and 4) zero vector.

[0609] Motion information may include 1) a motion vector, 2) a reference picture index, and 3) a reference direction. The reference direction may also be referred to as an "inter prediction indicator." The reference direction may be unidirectional or bidirectional. A unidirectional reference direction may indicate L0 prediction or L1 prediction.

[0610] A merge candidate list may be created before performing prediction in merge mode.

[0611] The number of merge candidates in the merge candidate list may be predefined. Each of the encoding device 100 and the decoding device 200 may add merge candidates to the merge candidate list according to a predefined scheme and a predefined priority, so that the merge candidate list has a predefined number of merge candidates. The merge candidate list of the encoding device 100 and the merge candidate list of the decoding device 200 may be made identical to each other using the predefined scheme and the predefined priority.

[0612] Merging may be applied on a CU or PU basis. When performing merging on a CU or PU basis, the encoding device 100 may transmit a bitstream including predefined information to the decoding device 200. For example, the predefined information may include 1) information indicating whether merging is performed for each block partition, and 2) information on blocks on which merging is to be performed among blocks that are spatial candidates and / or temporal candidates for the target block.

[0613] 2-2) Searching for motion vectors using the merge candidate list

[0614] The encoding device 100 may determine a merge candidate to be used for encoding the target block. For example, the encoding device 100 may perform prediction on the target block using a merge candidate in the merge candidate list and may generate a residual block for the merge candidate. The encoding device 100 may encode the target block using the merge candidate that generates the minimum cost in encoding the prediction and residual block.

[0615] Also, the encoding apparatus 100 may determine whether to encode the target block using the merge mode.

[0616] 2-3) Transmission of inter-frame prediction information

[0617] The encoding device 100 may generate a bitstream including inter-frame prediction information required for inter-frame prediction. The encoding device 100 may generate entropy-coded inter-frame prediction information by performing entropy coding on the inter-frame prediction information, and may transmit the bitstream including the entropy-coded inter-frame prediction information to the decoding device 200. The entropy-coded inter-frame prediction information may be transmitted from the encoding device 100 to the decoding device 200 by signaling through the bitstream. The decoding device 200 may extract the entropy-coded inter-frame prediction information from the bitstream, and may obtain the inter-frame prediction information by applying entropy decoding to the entropy-coded inter-frame prediction information.

[0618] The decoding apparatus 200 may perform inter prediction on a target block using inter prediction information of a bitstream.

[0619] The inter prediction information may include 1) mode information indicating whether a merge mode is used, 2) a merge index, and 3) correction information.

[0620] Furthermore, the inter prediction information may include a residual signal.

[0621] The decoding apparatus 200 may acquire a merge index from a bitstream only when the mode information indicates that the merge mode is used.

[0622] The mode information may be a merge flag. The unit of the mode information may be a block. The information about the block may include the mode information, and the mode information may indicate whether the merge mode is applied to the block.

[0623] The merge index may indicate a merge candidate to be used for predicting the target block among the merge candidates included in the merge candidate list. Alternatively, the merge index may indicate a block to be merged with the target block among neighboring blocks that are spatially or temporally adjacent to the target block.

[0624] The encoding apparatus 100 may select a merge candidate having the highest encoding performance among the merge candidates included in the merge candidate list, and may set a value of the merge index to indicate the selected merge candidate.

[0625] The correction information may be information for correcting a motion vector. The encoding apparatus 100 may generate the correction information. The decoding apparatus 200 may correct a motion vector of a merge candidate selected by a merge index based on the correction information.

[0626] The correction information may include at least one of information indicating whether correction is to be performed, correction direction information, and correction size information. A prediction mode that corrects a motion vector based on signaled correction information may be referred to as a 'merge mode with motion vector difference'.

[0627] 2-4) Inter-frame prediction using merge mode of inter-frame prediction information

[0628] The decoding apparatus 200 may perform prediction on the target block using a merge candidate indicated by a merge index among merge candidates included in the merge candidate list.

[0629] The motion vector of the target block may be specified by the motion vector of the merge candidate indicated by the merge index, the reference picture index, and the reference direction.

[0630] 3) Skip mode

[0631] Skip mode can be a mode in which the motion information of a spatial candidate or a temporal candidate is applied to the target block without change. In addition, skip mode can be a mode in which a residual signal is not used. In other words, when skip mode is used, the reconstructed block can be the same as the predicted block.

[0632] The difference between the merge mode and the skip mode is whether the residual signal is transmitted or used. That is, the skip mode may be similar to the merge mode except that the residual signal is not transmitted or used.

[0633] When the skip mode is used, the encoding device 100 may transmit information about a block whose motion information is to be used as the motion information of the target block among blocks that are spatial candidates or temporal candidates to the decoding device 200 through a bitstream. The encoding device 100 may generate entropy-coded information by performing entropy coding on the information, and may signal the entropy-coded information to the decoding device 200 through a bitstream. The decoding device 200 may extract the entropy-coded information from the bitstream, and may obtain information by applying entropy decoding to the entropy-coded information.

[0634] In addition, when the skip mode is used, the encoding device 100 may not transmit other syntax information (such as MVD) to the decoding device 200. For example, when the skip mode is used, the encoding device 100 may not signal syntax elements related to at least one of the MVD, the coded block flag, and the transform coefficient level to the decoding device 200.

[0635] 3-1) Create a merge candidate list

[0636] The merge candidate list can also be used in skip mode. In other words, the merge candidate list can be used in both merge mode and skip mode. In this regard, the merge candidate list can also be referred to as a "skip candidate list" or a "merge / skip candidate list."

[0637] Alternatively, the skip mode may use an additional candidate list different from the candidate list of the merge mode. In this case, in the following description, the merge candidate list and the merge candidate may be replaced by the skip candidate list and the skip candidate, respectively.

[0638] A merge candidate list may be created before performing prediction in skip mode.

[0639] 3-2) Searching for motion vectors using the merge candidate list

[0640] The encoding device 100 may determine a merge candidate to be used for encoding the target block. For example, the encoding device 100 may perform prediction on the target block using the merge candidate in the merge candidate list. The encoding device 100 may encode the target block using the merge candidate that generates the minimum cost in prediction.

[0641] Also, the encoding apparatus 100 may determine whether to encode the target block using the skip mode.

[0642] 3-3) Transmission of inter-frame prediction information

[0643] The encoding apparatus 100 may generate a bitstream including inter prediction information required for inter prediction, and the decoding apparatus 200 may perform inter prediction on a target block using the inter prediction information of the bitstream.

[0644] The inter prediction information may include 1) mode information indicating whether the skip mode is used and 2) a skip index.

[0645] The skip index may be the same as the merge index described above.

[0646] When skip mode is used, the target block may be encoded without using a residual signal. The inter prediction information may not include a residual signal. Alternatively, the bitstream may not include a residual signal.

[0647] The decoding device 200 may obtain the skip index from the bitstream only when the mode information indicates that the skip mode is used. As described above, the merge index and the skip index may be the same as each other. The decoding device 200 may obtain the skip index from the bitstream only when the mode information indicates that the merge mode or the skip mode is used.

[0648] The skip index may indicate a merge candidate to be used for prediction of a target block among merge candidates included in the merge candidate list.

[0649] 3-4) Inter-frame prediction in skip mode using inter-frame prediction information

[0650] The decoding apparatus 200 may perform prediction on the target block using the merge candidate indicated by the skip index among the merge candidates included in the merge candidate list.

[0651] The motion vector of the target block may be specified by the motion vector of the merge candidate indicated by the skip index, the reference picture index, and the reference direction.

[0652] 4) Current picture reference mode

[0653] The current picture reference mode may denote a prediction mode that uses a previously reconstructed area in a target picture to which the target block belongs.

[0654] A motion vector for specifying a previously reconstructed area may be used.A reference picture index of a target block may be used to determine whether the target block has been encoded in a current picture reference mode.

[0655] A flag or index indicating whether the target block is a block encoded in the current picture reference mode may be signaled by the encoding apparatus 100 to the decoding apparatus 200. Alternatively, whether the target block is a block encoded in the current picture reference mode may be inferred through the reference picture index of the target block.

[0656] When the target block is encoded in the current picture reference mode, the current picture may exist at a fixed position or an arbitrary position in the reference picture list for the target block.

[0657] For example, the fixed position may be a position where the value of the reference picture index is 0 or a last position.

[0658] When a target picture exists at an arbitrary position in a reference picture list, an additional reference picture index indicating such an arbitrary position may be signaled by the encoding apparatus 100 to the decoding apparatus 200 .

[0659] 5) Sub-block merging mode

[0660] The subblock merge mode may be a mode in which motion information is derived from subblocks of a CU.

[0661] When the sub-block merge mode is applied, the sub-block merge candidate list may be generated using the motion information of the col-sub-block of the target sub-block in the reference image (i.e., the sub-block based temporal merge candidate) and / or the affine control point motion vector merge candidate.

[0662] 6) Triangle partitioning mode

[0663] In the triangular partitioning mode, the target block can be partitioned in a diagonal direction, and sub-target blocks generated by the partitioning can be generated. For each sub-target block, motion information of the corresponding sub-target block can be derived, and the derived motion information can be used to derive prediction samples for each sub-target block. The prediction samples of the target block can be derived by taking the weighted sum of the prediction samples of the sub-target blocks generated by the partitioning.

[0664] 7) Combined inter-intra prediction mode

[0665] The combined inter-intra prediction mode may be a mode in which a prediction sample of a target block is derived using a weighted sum of a prediction sample generated via inter prediction and a prediction sample generated via intra prediction.

[0666] In the above-described mode, the decoding device 200 may autonomously correct the derived motion information. For example, the decoding device 200 may search for motion information having the minimum sum of absolute differences (SAD) in a specific area based on a reference block indicated by the derived motion information, and may derive the found motion information as corrected motion information.

[0667] In the above-described mode, the decoding apparatus 200 may use optical flow to compensate for prediction samples derived through inter-frame prediction.

[0668] In the above-described AMVP mode, merge mode, skip mode, etc., index information of a list may be used to specify motion information to be used for prediction of a target block among a plurality of pieces of motion information in the list.

[0669] To improve encoding efficiency, the encoding apparatus 100 may signal only the index of the element generating the minimum cost in inter-frame prediction of the target block among the elements in the list. The encoding apparatus 100 may encode the index and signal the encoded index.

[0670] Therefore, the encoding device 100 and the decoding device 200 must be able to derive the above-described lists (i.e., the predicted motion vector candidate list and the merge candidate list) based on the same data using the same scheme. Here, the same data may include reconstructed pictures and reconstructed blocks. In addition, in order to specify an element using an index, the order of the elements in the list must be fixed.

[0671] Figure 10 Illustrated are spatial candidates according to an embodiment.

[0672] exist Figure 10 , the positions of the spatial candidates are shown.

[0673] The large block in the center of the figure may represent the target block, and the five small blocks may represent spatial candidates.

[0674] The coordinates of the target block may be (xP, yP), and the size of the target block may be represented by (nPSW, nPSH).

[0675] The spatial candidate A0 may be a block adjacent to the lower left corner of the target block. A0 may be a block occupying a pixel located at coordinates (xP-1, yP+nPSH).

[0676] Spatial candidate A1 may be a block adjacent to the left side of the target block. A1 may be the lowest block among the blocks adjacent to the left side of the target block. Alternatively, A1 may be a block adjacent to the top of A0. A1 may be a block occupying the pixel at coordinates (xP-1, yP+nPSH-1).

[0677] The spatial candidate B0 may be a block adjacent to the upper right corner of the target block. B0 may be a block occupying a pixel located at coordinates (xP+nPSW, yP-1).

[0678] Spatial candidate B1 may be a block adjacent to the top of the target block. B1 may be the rightmost block among the blocks adjacent to the top of the target block. Alternatively, B1 may be a block adjacent to the left of B0. B1 may be a block occupying the pixel located at coordinates (xP+nPSW-1, yP-1).

[0679] The spatial candidate B2 may be a block adjacent to the upper left corner of the target block. B2 may be a block occupying a pixel located at coordinates (xP-1, yP-1).

[0680] Determination of availability of spatial and temporal candidates

[0681] In order to include the motion information of the spatial candidate or the motion information of the temporal candidate in the list, it is necessary to determine whether the motion information of the spatial candidate or the motion information of the temporal candidate is available.

[0682] Hereinafter, candidate blocks may include spatial candidates and temporal candidates.

[0683] For example, the determination may be performed by sequentially applying the following steps 1) to 4) below.

[0684] Step 1) When the PU including the candidate block is located outside the boundary of the picture, the availability of the candidate block may be set to “false.” The expression “availability is set to false” may have the same meaning as “set to unavailable.”

[0685] Step 2) When the PU including the candidate block is located outside the boundary of the slice, the availability of the candidate block may be set to “false.” When the target block and the candidate block are located in different slices, the availability of the candidate block may be set to “false.”

[0686] Step 3) When the PU including the candidate block is located outside the boundary of the tile, the availability of the candidate block may be set to “false.” When the target block and the candidate block are located in different tiles, the availability of the candidate block may be set to “false.”

[0687] Step 4) When the prediction mode of the PU including the candidate block is intra prediction mode, the availability of the candidate block may be set to “false.” When the PU including the candidate block does not use inter prediction, the availability of the candidate block may be set to “false.”

[0688] Figure 11 An order in which motion information of spatial candidates is added to a merge list according to an embodiment is shown.

[0689] like Figure 11As shown in , when multiple pieces of motion information of spatial candidates are added to the merge list, the order of A1, B1, B0, A0, and B2 may be used. That is, multiple pieces of motion information of available spatial candidates may be added to the merge list in the order of A1, B1, B0, A0, and B2.

[0690] Methods for exporting merge lists in merge mode and skip mode

[0691] As described above, the maximum number of merge candidates in a merge list can be set. The set maximum number can be indicated by "N". The set number can be transmitted from the encoding device 100 to the decoding device 200. The slice header of the slice may include N. In other words, the maximum number of merge candidates in the merge list for the target block of the slice can be set by the slice header. For example, the value of N can basically be 5.

[0692] A plurality of pieces of motion information (ie, merge candidates) may be added to the merge list in the order of the following steps 1) to 4) below.

[0693] Step 1) Among the space candidates, available space candidates can be added to the merge list. Figure 11 The plurality of pieces of motion information of the available spatial candidates are added to the merge list in the order shown in FIG. Here, when the motion information of the available spatial candidate overlaps with other motion information already in the merge list, the motion information of the available spatial candidate may not be added to the merge list. The operation of checking whether the corresponding motion information overlaps with other motion information in the list may be simply referred to as "overlap check."

[0694] The maximum number of pieces of motion information added may be N.

[0695] Step 2) When the number of motion information items in the merge list is less than N and a temporal candidate is available, the motion information of the temporal candidate may be added to the merge list. Here, when the motion information of the available temporal candidate overlaps with other motion information already in the merge list, the motion information of the available temporal candidate may not be added to the merge list.

[0696] Step 3) When the number of pieces of motion information in the merge list is less than N and the type of the target slice is 'B', combined motion information generated by combining bidirectional predictions (bi-prediction) may be added to the merge list.

[0697] The target slice may be a slice including the target block.

[0698] The combined motion information may be a combination of L0 motion information and L1 motion information. The L0 motion information may be motion information that refers only to the L0 reference picture list. The L1 motion information may be motion information that refers only to the L1 reference picture list.

[0699] In the merge list, there may be one or more pieces of L0 motion information. In addition, in the merge list, there may be one or more pieces of L1 motion information.

[0700] The combined motion information may include one or more pieces of combined motion information. When generating the combined motion information, the L0 motion information and the L1 motion information to be used in the step of generating the combined motion information may be predefined among the one or more pieces of L0 motion information and the one or more pieces of L1 motion information. The one or more pieces of combined motion information may be generated in a predefined order by combining bidirectional prediction using a pair of different motion information in a merge list. One piece of the pair of different motion information may be L0 motion information, and the other piece of the pair of different motion information may be L1 motion information.

[0701] For example, the combined motion information added with the highest priority may be a combination of L0 motion information with a merge index of 0 and L1 motion information with a merge index of 1. When the motion information with a merge index of 0 is not L0 motion information or when the motion information with a merge index of 1 is not L1 motion information, the combined motion information may be neither generated nor added. Next, the combined motion information added with the next highest priority may be a combination of L0 motion information with a merge index of 1 and L1 motion information with a merge index of 0. The subsequent detailed combinations may conform to other combinations in the field of video encoding / decoding.

[0702] Here, when the combined motion information overlaps with other motion information already present in the merge list, the combined motion information may not be added to the merge list.

[0703] Step 4) When the number of pieces of motion information in the merge list is less than N, motion information of a zero vector may be added to the merge list.

[0704] The zero-vector motion information may be motion information in which a motion vector is a zero vector.

[0705] The number of pieces of zero-vector motion information may be one or more. The reference picture indexes of one or more pieces of zero-vector motion information may be different from each other. For example, the reference picture index value of the first zero-vector motion information may be 0. The reference picture index value of the second zero-vector motion information may be 1.

[0706] The number of pieces of zero-vector motion information may be the same as the number of reference pictures in the reference picture list.

[0707] The reference direction of the zero-vector motion information can be bidirectional. Both motion vectors can be zero vectors. The number of pieces of zero-vector motion information can be the smaller of the number of reference pictures in reference picture list L0 and the number of reference pictures in reference picture list L1. Alternatively, when the number of reference pictures in reference picture list L0 and the number of reference pictures in reference picture list L1 are different, a unidirectional reference direction can be used for a reference picture index applicable only to a single reference picture list.

[0708] The encoding apparatus 100 and / or the decoding apparatus 200 may then add zero-vector motion information to the merge list while changing the reference picture index.

[0709] When the zero-vector motion information overlaps with other motion information already in the merge list, the zero-vector motion information may not be added to the merge list.

[0710] The order of steps 1) to 4) is merely exemplary and may be changed. In addition, some of the steps above may be omitted according to predefined conditions.

[0711] Method for deriving a candidate list of predicted motion vectors in AMVP mode

[0712] The maximum number of motion vector predictor candidates in the motion vector predictor candidate list may be predefined. The predefined maximum number may be indicated by N. For example, the predefined maximum number may be 2.

[0713] A plurality of pieces of motion information (ie, prediction motion vector candidates) may be added to the prediction motion vector candidate list in the order of steps 1) to 3) below.

[0714] Step 1) Available spatial candidates among the spatial candidates may be added to the motion vector prediction candidate list.The spatial candidates may include a first spatial candidate and a second spatial candidate.

[0715] The first spatial candidate may be one of A0, A1, scaled A0, and scaled A1. The second spatial candidate may be one of B0, B1, B2, scaled B0, scaled B1, and scaled B2.

[0716] Multiple pieces of motion information of available spatial candidates may be added to the motion vector prediction candidate list in the order of the first spatial candidate and the second spatial candidate. In this case, if the motion information of an available spatial candidate overlaps with other motion information already in the motion vector prediction candidate list, the motion information of the available spatial candidate may not be added to the motion vector prediction candidate list. In other words, when the value of N is 2, if the motion information of the second spatial candidate is the same as the motion information of the first spatial candidate, the motion information of the second spatial candidate may not be added to the motion vector prediction candidate list.

[0717] The maximum number of pieces of motion information added may be N.

[0718] Step 2) When the number of pieces of motion information in the motion vector predictor candidate list is less than N and a temporal candidate is available, the motion information of the temporal candidate may be added to the motion vector predictor candidate list. In this case, if the motion information of the available temporal candidate overlaps with other motion information already in the motion vector predictor candidate list, the motion information of the available temporal candidate may not be added to the motion vector predictor candidate list.

[0719] Step 3) When the number of pieces of motion information in the motion vector predictor candidate list is less than N, zero vector motion information may be added to the motion vector predictor candidate list.

[0720] The zero-vector motion information may include one or more pieces of zero-vector motion information. Reference picture indices of the one or more pieces of zero-vector motion information may be different from each other.

[0721] The encoding apparatus 100 and / or the decoding apparatus 200 may sequentially add a plurality of pieces of zero-vector motion information to the prediction motion vector candidate list while changing the reference picture index.

[0722] When the zero-vector motion information overlaps with other motion information already existing in the motion vector predictor candidate list, the zero-vector motion information may not be added to the motion vector predictor candidate list.

[0723] The above description of the zero vector motion information in conjunction with the merge list is also applicable to the zero vector motion information, and its repeated description will be omitted.

[0724] The order of steps 1) to 3) described above is merely exemplary and may be changed. In addition, some of the steps may be omitted according to predefined conditions.

[0725] Figure 12 Transformation and quantization processes according to examples are shown.

[0726] like Figure 12As shown in , the quantized levels may be generated by performing a transform and / or quantization process on the residual signal.

[0727] The residual signal may be generated as a difference between the original block and the predicted block.Here, the predicted block may be a block generated via intra prediction or inter prediction.

[0728] The residual signal may be transformed into a signal in the frequency domain through a transform process as part of the quantization process.

[0729] The transform kernel used for the transform may include various DCT kernels such as discrete cosine transform (DCT) type 2 (DCT-II) and discrete sine transform (DST) kernels.

[0730] These transform kernels may perform separable transform or two-dimensional (2D) non-separable transform on the residual signal. The separable transform may be a transform indicating that a one-dimensional (1D) transform is performed on the residual signal in each of horizontal and vertical directions.

[0731] The DCT type and DST type adaptively used for 1D transform may include DCT-V, DCT-VIII, DST-I, and DST-VII in addition to DCT-II, as shown in each of Table 3 and Table 4 below.

[0732] Table 3

[0733]

[0734]

[0735] Table 4

[0736] Transformation Set Transformation Candidates 0 DST-VII,DCT-VIII,DST-I 1 DST-VII,DST-I,DCT-VIII 2 DST-VII,DCT-V,DST-I

[0737] As shown in Tables 3 and 4, when deriving the DCT type or DST type to be used for transformation, a transform set can be used. Each transform set may include multiple transform candidates. Each transform candidate may be a DCT type or a DST type.

[0738] Table 5 below shows an example of a transform set to be applied to a horizontal direction and a transform set to be applied to a vertical direction according to an intra prediction mode.

[0739] Table 5

[0740] Intra prediction mode 0 1 2 3 4 5 6 7 8 9 Vertical Transform Set 2 1 0 1 0 1 0 1 0 1 Horizontal Transform Set 2 1 0 1 0 1 0 1 0 1 Intra prediction mode 10 11 12 13 14 15 16 17 18 19 Vertical Transform Set 0 1 0 1 0 0 0 0 0 0 Horizontal Transform Set 0 1 0 1 2 2 2 2 2 2 Intra prediction mode 20 21 22 23 24 25 26 27 28 29 Vertical Transform Set 0 0 0 1 0 1 0 1 0 1 Horizontal Transform Set 2 2 2 1 0 1 0 1 0 1 Intra prediction mode 30 31 32 33 34 35 36 37 38 39 Vertical Transform Set 0 1 0 1 0 1 0 1 0 1 Horizontal Transform Set 0 1 0 1 0 1 0 1 0 1 Intra prediction mode 40 41 42 43 44 45 46 47 48 49 Vertical Transform Set 0 1 0 1 0 1 2 2 2 2 Horizontal Transform Set 0 1 0 1 0 1 0 0 0 0 Intra prediction mode 50 51 52 53 54 55 56 57 58 59 Vertical Transform Set 2 2 2 2 2 1 0 1 0 1 Horizontal Transform Set 0 0 0 0 0 1 0 1 0 1 Intra prediction mode 60 61 62 63 64 65 66 Vertical Transform Set 0 1 0 1 0 1 0 Horizontal Transform Set 0 1 0 1 0 1 0

[0741] In Table 5, the numbers of a vertical transform set and a horizontal transform set to be applied to a horizontal direction of a residual signal according to an intra prediction mode of a target block are shown.

[0742] like Figure 4 and Figure 5As illustrated in , a transform set to be applied in the horizontal and vertical directions may be predefined according to the intra prediction mode of the target block. The encoding apparatus 100 may perform transform and inverse transform on the residual signal using the transform included in the transform set corresponding to the intra prediction mode of the target block. In addition, the decoding apparatus 200 may perform inverse transform on the residual signal using the transform included in the transform set corresponding to the intra prediction mode of the target block.

[0743] In the transformation and inverse transformation, the transform set to be applied to the residual signal may be determined and may not be signaled as illustrated in Tables 3, 4, and 5. Transform indication information may be signaled from the encoding apparatus 100 to the decoding apparatus 200. The transform indication information may be information indicating which of a plurality of transform candidates included in the transform set to be applied to the residual signal is to be used.

[0744] For example, when the target block size is 64×64 or smaller, a transform set each having three transforms can be configured according to the intra prediction mode. An optimal transform method can be selected from a total of nine multi-transform methods generated by combining three transforms in the horizontal direction and three transforms in the vertical direction. Using such an optimal transform method, the residual signal can be encoded and / or decoded, thereby improving coding efficiency.

[0745] Here, information indicating which of the multiple transforms belonging to each transform set has been used for at least one of the vertical transform and the horizontal transform may be entropy encoded and / or entropy decoded. Here, truncated unary binarization may be used to encode and / or decode such information.

[0746] As described above, methods using various transforms may be applied to a residual signal generated via intra prediction or inter prediction.

[0747] The transform may include at least one of a first transform and a secondary transform. A transform coefficient may be generated by performing a first transform on the residual signal, and a secondary transform coefficient may be generated by performing a secondary transform on the transform coefficient.

[0748] The first transform may be referred to as a “primary transform.” Furthermore, the first transform may also be referred to as an “adaptive multi-transform (AMT) scheme.” As described above, AMT may indicate that different transforms are applied to each 1D direction (ie, vertical and horizontal directions).

[0749] The secondary transform may be a transform for increasing the energy concentration of the transform coefficients generated by the first transform. Similar to the first transform, the secondary transform may be a separable transform or a non-separable transform. Such a non-separable transform may be a non-separable secondary transform (NSST).

[0750] The first transform may be performed using at least one of a plurality of predefined transform methods, such as discrete cosine transform (DCT), discrete sine transform (DST), Karhunen-Loeve transform (KLT), and the like.

[0751] Furthermore, the first transform may be a transform having various types according to a kernel function defining discrete cosine transform (DCT) or discrete sine transform (DST).

[0752] For example, the transform type may be determined based on at least one of the following: 1) a prediction mode of the target block (e.g., one of intra-frame prediction and inter-frame prediction), 2) a size of the target block, 3) a shape of the target block, 4) an intra-frame prediction mode of the target block, 5) a component of the target block (e.g., one of a luminance component and a chrominance component), and 6) a partition type applied to the target block (e.g., one of a quadtree, a binary tree, and a ternary tree).

[0753] For example, according to the transform kernels presented in Table 6 below, the first transform may include transforms such as DCT-2, DCT-5, DCT-7, DST-7, DST-1, DST-8, and DCT-8. In Table 6 below, various transform types and transform kernel functions for multi-transform selection (MTS) are exemplified.

[0754] MTS may refer to the selection of a combination of one or more DCT and / or DST kernels to transform the residual signal in the horizontal and / or vertical directions.

[0755] Table 6

[0756]

[0757] In Table 6, i and j may be integer values ​​equal to or greater than 0 and less than or equal to N-1.

[0758] A secondary transform may be performed on transform coefficients generated by performing the first transform.

[0759] As in the first transform, a transform set may also be defined in a secondary transform.The method for deriving and / or determining the above-mentioned transform set may be applied not only to the first transform but also to the secondary transform.

[0760] The primary transform and the secondary transform may be determined for a specific goal.

[0761] For example, the first transform and the secondary transform may be applied to signal components corresponding to one or more of a luminance (luma) component and a chrominance (chroma) component. Whether to apply the first transform and / or the secondary transform may be determined based on at least one of the coding parameters for the target block and / or the neighboring blocks. For example, whether to apply the first transform and / or the secondary transform may be determined based on the size and / or shape of the target block.

[0762] In the encoding apparatus 100 and the decoding apparatus 200 , transform information indicating a transform method to be used for a target may be derived by using the designation information.

[0763] For example, the transform information may include a transform index to be used for the primary transform and / or the secondary transform. Alternatively, the transform information may indicate that the primary transform and / or the secondary transform is not to be used.

[0764] For example, when the target of the primary transform and the secondary transform is a target block, the transform method to be applied to the primary transform and / or the secondary transform indicated by the transform information can be determined based on at least one of the encoding parameters for the target block and / or blocks adjacent to the target block.

[0765] Alternatively, transformation information indicating a transformation method for a specific target may be signaled from the encoding apparatus 100 to the decoding apparatus 200 .

[0766] For example, for a single CU, whether a primary transform is used, an index indicating the primary transform, whether a secondary transform is used, and an index indicating the secondary transform may be derived as transform information by the decoding apparatus 200. Alternatively, for a single CU, transform information indicating whether a primary transform is used, an index indicating the primary transform, whether a secondary transform is used, and an index indicating the secondary transform may be signaled.

[0767] A quantized transform coefficient (ie, a quantization level) may be generated by performing quantization on a result generated by performing the first transform and / or the sub-transform or performing quantization on a residual signal.

[0768] Figure 13 A diagonal scan according to an example is shown.

[0769] Figure 14 A horizontal scan according to an example is shown.

[0770] Figure 15 A vertical scan according to an example is shown.

[0771] The quantized transform coefficients may be scanned via at least one of (upper right) diagonal scanning, vertical scanning, and horizontal scanning according to at least one of an intra prediction mode, a block size, and a block shape. The block may be a transform unit (TU).

[0772] Each scan may be initiated at a specific start point and may be terminated at a specific end point.

[0773] For example, by using Figure 13 The coefficients of the block are scanned by diagonal scanning to change the quantized transform coefficients into a 1D vector form. Optionally, the quantized transform coefficients can be used according to the size of the block and / or the intra prediction mode. Figure 14 Horizontal scan or Figure 15 vertical scanning instead of diagonal scanning.

[0774] Vertical scanning may be an operation of scanning 2D block type coefficients in a column direction, and horizontal scanning may be an operation of scanning 2D block type coefficients in a row direction.

[0775] In other words, which of diagonal scanning, vertical scanning, and horizontal scanning is to be used may be determined according to the size of a block and / or an inter prediction mode.

[0776] like Figure 13 、 Figure 14 and Figure 15 As shown in , the quantized transform coefficients may be scanned along a diagonal direction, a horizontal direction, or a vertical direction.

[0777] The quantized transform coefficients can be represented by a block shape. Each block can include multiple sub-blocks. Each sub-block can be defined according to a minimum block size or a minimum block shape.

[0778] In the scanning, a scanning order according to a type or direction of scanning may be first applied to a subblock. In addition, a scanning order according to a direction of scanning may be applied to quantized transform coefficients in each subblock.

[0779] For example, Figure 13 、 Figure 14 and Figure 15 As shown in , when the size of the target block is 8×8, quantized transform coefficients can be generated through the first transform, the secondary transform, and the quantization of the residual signal of the target block. Therefore, one of three types of scanning orders can be applied to the four 4×4 sub-blocks, and the quantized transform coefficients can also be scanned for each 4×4 sub-block according to the scanning order.

[0780] The encoding apparatus 100 may generate entropy-encoded quantized transform coefficients by performing entropy encoding on the scanned quantized transform coefficients, and may generate a bitstream including the entropy-encoded quantized transform coefficients.

[0781] The decoding apparatus 200 may extract entropy-encoded quantized transform coefficients from a bitstream and may generate quantized transform coefficients by performing entropy decoding on the entropy-encoded quantized transform coefficients. The quantized transform coefficients may be arranged in the form of 2D blocks via inverse scanning. Here, as the inverse scanning method, at least one of upper right diagonal scanning, vertical scanning, and horizontal scanning may be performed.

[0782] In the decoding apparatus 200, inverse quantization may be performed on the quantized transform coefficients. Depending on whether the inverse secondary transform is to be performed, a secondary inverse transform may be performed on the result generated by performing the inverse quantization. In addition, depending on whether the first inverse transform is to be performed, a first inverse transform may be performed on the result generated by performing the secondary inverse transform. A reconstructed residual signal may be generated by performing the first inverse transform on the result generated by performing the secondary inverse transform.

[0783] For luma components reconstructed via intra prediction or inter prediction, inverse mapping with a dynamic range may be performed before loop filtering.

[0784] The dynamic range can be divided into 16 equal segments and the mapping function of the corresponding segments can be signaled.Such mapping function can be signaled at the slice level or tile group level.

[0785] An inverse mapping function for performing inverse mapping may be derived based on the mapping function.

[0786] Loop filtering, storage of reference pictures, and motion compensation may be performed in the inverse mapping area.

[0787] The prediction block generated by inter prediction can be transformed to the mapping area by mapping using a mapping function, and the transformed prediction block can be used to generate a reconstructed block. However, since intra prediction is performed in the mapping area, the prediction block generated by intra prediction can be used to generate a reconstructed block without the need for mapping and / or inverse mapping.

[0788] For example, when the target block is a residual block of a chroma component, the residual block may be transformed to the inverse mapping area by scaling the chroma component of the mapping area.

[0789] Whether scaling is available can be signaled at the slice level or tile group level.

[0790] For example, scaling may be applied only if mapping is available for luma components and the partitions of the luma components and the partitions of the chroma components follow the same tree structure.

[0791] Scaling may be performed based on an average value of the values ​​of samples in the luma prediction block corresponding to the chroma prediction block. Here, when the target block uses inter prediction, the luma prediction block may refer to a mapped luma prediction block.

[0792] The value required for scaling may be derived by referring to a lookup table using an index of a segment to which the average value of the sample values ​​of the luma prediction block belongs.

[0793] The residual block may be transformed to the inverse mapping region by scaling the residual block using the final derived value. Thereafter, for the block of the chroma component, reconstruction, intra prediction, inter prediction, loop filtering, and storage of reference pictures may be performed in the inverse mapping region.

[0794] For example, information indicating whether mapping and / or inverse mapping of luma components and chroma components is available may be signaled through a sequence parameter set.

[0795] A prediction block for a target block may be generated based on a block vector. The block vector may indicate a displacement between the target block and a reference block. The reference block may be a block in the target image.

[0796] In this manner, a prediction mode that generates a prediction block by referring to a target image may be referred to as an 'intra block copy (IBC) mode'.

[0797] The IBC mode can be applied to a CU of a specific size. For example, the IBC mode can be applied to a CU of M×N size. Here, M and N can be less than or equal to 64.

[0798] IBC modes may include skip mode, merge mode, AMVP mode, and the like. In the case of skip mode or merge mode, a merge candidate list may be configured, and a merge index may be signaled, thereby specifying a single merge candidate among the merge candidates present in the merge candidate list. The block vector of the specified merge candidate may be used as the block vector of the target block.

[0799] In the case of AMVP mode, a differential block vector can be signaled. In addition, a prediction block vector can be derived from the left neighboring block and the upper neighboring block of the target block. In addition, an index indicating which neighboring block to use can be signaled.

[0800] The prediction block in IBC mode can be included in the target CTU or the left CTU and can be limited to blocks within the previously reconstructed area. For example, the value of the block vector can be limited so that the prediction block of the target block is located in a specific area. The specific area can be an area defined by three 64×64 blocks that are encoded and / or decoded before the 64×64 block including the target block. By limiting the value of the block vector in this way, the memory consumption and device complexity caused by the implementation of the IBC mode can be reduced.

[0801] Figure 16 is a configuration diagram of an encoding device according to an embodiment.

[0802] The encoding apparatus 1600 may correspond to the encoding apparatus 100 described above.

[0803] The encoding apparatus 1600 may include a processing unit 1610, a memory 1630, a user interface (UI) input device 1650, a UI output device 1660, and a storage 1640 communicating with each other through a bus 1690. The encoding apparatus 1600 may further include a communication unit 1620 connected to a network 1699.

[0804] The processing unit 1610 may be a central processing unit (CPU) or a semiconductor device for executing processing instructions stored in the memory 1630 or the storage 1640. The processing unit 1610 may be at least one hardware processor.

[0805] The processing unit 1610 may generate and process signals, data, or information input to, output from, or used in the encoding device 1600, and may perform checks, comparisons, determinations, etc. related to the signals, data, or information. In other words, in embodiments, the generation and processing of data or information, as well as checks, comparisons, and determinations related to the data or information, may be performed by the processing unit 1610.

[0806] The processing unit 1610 may include an inter-frame prediction unit 110, 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.

[0807] At least some of the inter-frame prediction unit 110, the intra-frame prediction unit 120, the switch 115, the subtractor 125, the transform unit 130, the quantization unit 140, the entropy encoding unit 150, the inverse quantization unit 160, the inverse transform unit 170, the adder 175, the filter unit 180, and the reference picture buffer 190 may be program modules and may communicate with an external device or system. The program modules may be included in the encoding apparatus 1600 in the form of an operating system, an application module, or other program modules.

[0808] The program modules may be physically stored in various types of well-known storage devices. In addition, at least some of the program modules may also be stored in a remote storage device that can communicate with the encoding device 1600.

[0809] Program modules may include, but are not limited to, routines, subroutines, programs, objects, components, and data structures for performing functions or operations according to embodiments or for implementing abstract data types according to embodiments.

[0810] The program modules may be implemented using instructions or codes executed by at least one processor of the encoding device 1600 .

[0811] The processing unit 1610 can execute instructions or codes in the inter-frame prediction unit 110, the intra-frame prediction unit 120, the switch 115, the subtractor 125, the transform unit 130, the quantization unit 140, the entropy coding unit 150, the inverse quantization unit 160, the inverse transform unit 170, the adder 175, the filter unit 180 and the reference picture buffer 190.

[0812] The storage unit may represent the memory 1630 and / or the storage 1640. Each of the memory 1630 and the storage 1640 may be any of various types of volatile or non-volatile storage media. For example, the memory 1630 may include at least one of a read-only memory (ROM) 1631 and a random access memory (RAM) 1632.

[0813] The storage unit may store data or information used for the operation of the encoding apparatus 1600. In an embodiment, data or information of the encoding apparatus 1600 may be stored in the storage unit.

[0814] For example, the storage unit may store pictures, blocks, lists, motion information, inter-frame prediction information, bitstreams, and the like.

[0815] The encoding device 1600 may be implemented in a computer system including a computer-readable storage medium.

[0816] The storage medium may store at least one module required for the operation of the encoding apparatus 1600. The memory 1630 may store at least one module and may be configured such that the at least one module is executed by the processing unit 1610.

[0817] Functions related to communication of data or information of the encoding apparatus 1600 may be performed through the communication unit 1620 .

[0818] For example, the communication unit 1620 may transmit the bitstream to the decoding apparatus 1700 which will be described later.

[0819] Figure 17 is a configuration diagram of a decoding device according to an embodiment.

[0820] The decoding device 1700 may correspond to the decoding device 200 described above.

[0821] The decoding apparatus 1700 may include a processing unit 1710, a memory 1730, a user interface (UI) input device 1750, a UI output device 1760, and a storage 1740 communicating with each other through a bus 1790. The decoding apparatus 1700 may also include a communication unit 1720 connected to a network 1799.

[0822] The processing unit 1710 may be a central processing unit (CPU) or a semiconductor device for executing processing instructions stored in the memory 1730 or the storage 1740. The processing unit 1710 may be at least one hardware processor.

[0823] The processing unit 1710 may generate and process signals, data, or information input to, output from, or used in the decoding device 1700, and may perform checks, comparisons, determinations, etc. related to the signals, data, or information. In other words, in embodiments, the generation and processing of data or information, as well as checks, comparisons, and determinations related to the data or information, may be performed by the processing unit 1710.

[0824] The processing unit 1710 may include an entropy decoding unit 210 , an inverse quantization unit 220 , an inverse transform unit 230 , an intra prediction unit 240 , an inter prediction unit 250 , a switch 245 , an adder 255 , a filter unit 260 , and a reference picture buffer 270 .

[0825] At least some of the entropy decoding unit 210, the inverse quantization unit 220, the inverse transform unit 230, the intra-frame prediction unit 240, the inter-frame prediction unit 250, the adder 255, the switch 245, the filter unit 260, and the reference picture buffer 270 of the decoding device 200 may be program modules and may communicate with an external device or system. The program modules may be included in the decoding device 1700 in the form of an operating system, an application module, or other program modules.

[0826] The program modules may be physically stored in various types of well-known storage devices. In addition, at least some of the program modules may also be stored in a remote storage device that can communicate with the decoding device 1700.

[0827] Program modules may include, but are not limited to, routines, subroutines, programs, objects, components, and data structures for performing functions or operations according to embodiments or for implementing abstract data types according to embodiments.

[0828] The program modules may be implemented using instructions or codes executed by at least one processor of the decoding device 1700 .

[0829] The processing unit 1710 can execute instructions or codes in the entropy decoding unit 210, the inverse quantization unit 220, the inverse transform unit 230, the intra-frame prediction unit 240, the inter-frame prediction unit 250, the switch 245, the adder 255, the filter unit 260 and the reference picture buffer 270.

[0830] The storage unit may represent the memory 1730 and / or the storage 1740. Each of the memory 1730 and the storage 1740 may be any of various types of volatile or non-volatile storage media. For example, the memory 1730 may include at least one of the ROM 1731 and the RAM 1732.

[0831] The storage unit may store data or information used for the operation of the decoding apparatus 1700. In an embodiment, data or information of the decoding apparatus 1700 may be stored in the storage unit.

[0832] For example, the storage unit may store pictures, blocks, lists, motion information, inter-frame prediction information, bitstreams, and the like.

[0833] The decoding device 1700 may be implemented in a computer system including a computer-readable storage medium.

[0834] The storage medium may store at least one module required for the operation of the decoding apparatus 1700. The memory 1730 may store at least one module and may be configured such that the at least one module is executed by the processing unit 1710.

[0835] Functions related to communication of data or information of the decoding apparatus 1700 may be performed through the communication unit 1720 .

[0836] For example, the communication unit 1720 may receive a bitstream from the encoding apparatus 1700 .

[0837] Hereinafter, a processing unit may refer to the processing unit 1610 of the encoding apparatus 1600 and / or the processing unit 1710 of the decoding apparatus 1700. For example, with respect to functions related to prediction, a processing unit may refer to the switch 115 and / or the switch 245. With respect to functions related to inter-frame prediction, a processing unit may refer to the inter-frame prediction unit 110, the subtractor 125, and the adder 175, and may refer to the inter-frame prediction unit 250 and the adder 255. With respect to functions related to intra-frame prediction, a processing unit may refer to the intra-frame prediction unit 120, the subtractor 125, and the adder 175, and may refer to the intra-frame prediction unit 240 and the adder 255. With respect to functions related to transform, a processing unit may refer to the transform unit 130 and the inverse transform unit 170, and may refer to the inverse transform unit 230. With respect to functions related to quantization, a processing unit may refer to the quantization unit 140 and the inverse quantization unit 160, and may refer to the inverse quantization unit 220. Regarding functions related to entropy encoding and / or entropy decoding, the processing unit may indicate the entropy encoding unit 150 and / or the entropy decoding unit 210. Regarding functions related to filtering, the processing unit may indicate the filter unit 180 and / or the filter unit 260. Regarding functions related to reference pictures, the processing unit may indicate the reference picture buffer 190 and / or the reference picture buffer 270.

[0838] In typical image encoding / decoding methods, decoder-side motion information derivation methods may be used to a limited extent. Therefore, the improvement in coding efficiency attributable to decoder-side motion information derivation methods may also be limited.

[0839] In an embodiment, in order to improve encoding efficiency in inter-frame prediction, an encoding / decoding method, apparatus, and storage medium using a motion information search method may be provided.

[0840] The processing unit may perform inter prediction on the target block. Here, when the prediction of the target block is performed by the motion information search method, the processing unit may derive motion information of the target block from available information.

[0841] With the help of derivation of motion information, coding efficiency can be improved by minimizing the number of bits used to signal the coding information required for decoding of the target block.

[0842] In an embodiment, the encoding information may include information required for decoding of the target block described in the embodiment (sent from the encoding apparatus 1600 to the decoding apparatus 1700). For example, the encoding information may include encoding parameters.

[0843] In an embodiment, the information described as being signaled via a bitstream may be included in the encoding information. In addition, the encoding information may be signaled via a bitstream.

[0844] In an embodiment, the available information may refer to information that is previously decoded (or reconstructed) before performing prediction on the target block.

[0845] In an embodiment, the available information may include at least one of encoding parameters, motion information, predicted sample information, reconstructed sample information, and loop-filtered decoded sample information at a specific sample position in the target picture.

[0846] In an embodiment, the available information may include at least one of encoding parameters, motion information, prediction sample information, reconstruction sample information, and loop filtering decoded sample information of a specific block including a specific sample position in the target picture.

[0847] In an embodiment, the specific sample point position may refer to a position of a neighboring sample point adjacent to the left, above, and / or above-left of the target block.

[0848] In an embodiment, the available information may include at least one of encoding parameters, motion information, prediction sample information, reconstruction sample information, and loop-filtered reconstruction sample information at a specific sample position in a reference picture for the target block.

[0849] In an embodiment, the available information may include at least one of the following: encoding parameters of a block including a specific sample position in a reference picture for the target block, motion information, prediction sample information, reconstruction sample information, and loop filtering reconstruction sample information.

[0850] For example, the specific sample point position may be indicated by the motion information of the target block.

[0851] In an embodiment, the available information may include at least one of encoding parameters, motion information, prediction sample information, reconstruction sample information, and loop-filtered reconstruction sample information at a specific sample position in the col picture for the target block.

[0852] In an embodiment, the available information may include at least one of the following: encoding parameters of a block including a specific sample position in a col picture for the target block, motion information, prediction sample information, reconstruction sample information, and loop filtering reconstruction sample information.

[0853] For example, the specific sample point position may be indicated by the motion information of the target block.

[0854] For example, the specific sample point position may refer to the position of a neighboring sample point adjacent to the left, above, and / or above-left of the co1 block.

[0855] Hereinafter, an embodiment in which the motion information search method is implemented will be described.

[0856] Figure 18 is a flowchart illustrating a target block prediction method and a bitstream generation method according to an embodiment.

[0857] The target block prediction method and the bitstream generation method according to the embodiment may be performed by the encoding apparatus 1600. This embodiment may be a part of the target block encoding method or the video encoding method.

[0858] The prediction may be one of the prediction methods described in the above embodiments. For example, the prediction may be inter-frame prediction or intra-frame prediction.

[0859] At step 1810 , the processing unit 1610 may determine prediction information to be used for encoding of a target block.

[0860] The prediction information may include the information used for prediction described in the embodiments.

[0861] For example, the prediction information may include inter-frame prediction information. For example, the prediction information may include intra-frame prediction information.

[0862] For example, the prediction information may include a motion information candidate list and final motion information, which will be described later. The determination of the prediction information may include configuration of the motion information candidate list and determination of the final motion information.

[0863] In step 1820, encoded encoding information may be generated by performing encoding on the encoding information.

[0864] The encoding information may refer to information signaled / encoded / decoded as described in the embodiments. In other words, the encoding information may be information for allowing the decoding apparatus 1700 to perform prediction corresponding to the prediction performed by the encoding apparatus 1600.

[0865] At step 1830 , the processing unit 1610 may generate a bitstream.

[0866] The bitstream may include information about the target block.In addition, the bitstream may include the information described above in the embodiments.

[0867] For example, the bitstream may include encoded encoding information or coding information.

[0868] For example, the bitstream may include coding parameters related to the target block and / or properties of the target block.

[0869] The information included in the bitstream may be generated in step 1820 , or may be at least partially generated in steps 1810 and 1820 .

[0870] The processing unit 1610 may store the generated bitstream in the memory 1640. Alternatively, the communication unit 1620 may transmit the bitstream to the decoding apparatus 1700.

[0871] The bitstream may include encoding information about the target block. The processing unit 1610 may generate the encoding information about the target block by performing entropy encoding on the information about the target block.

[0872] In step 1840 , the processing unit 1610 may perform prediction on the target block using the information about the target block and the prediction information.

[0873] The processing unit 1610 may use the encoding information when performing prediction on the target block. Alternatively, the encoding information may be generated to correspond to information used for prediction of the target block.

[0874] A prediction block may be generated by predicting a target block. A residual block may be generated as a difference between the target block and the prediction block. Information about the target block may be generated by applying transformation and quantization to the residual block.

[0875] The information about the target block may include transform and quantization coefficients for the target block. A reconstructed residual block may be generated by applying inverse quantization and inverse transformation to the transform and quantization coefficients of the target block. A reconstructed block may be generated as the sum of the prediction block and the reconstructed residual block.

[0876] Figure 19is a flowchart illustrating a target block prediction method using a bitstream according to an embodiment.

[0877] The target block prediction method using a bitstream according to an embodiment may be performed by the decoding apparatus 1700. This embodiment may be a part of a target block decoding method or a video decoding method.

[0878] The prediction may be one of the prediction methods described above in the embodiments. For example, the second prediction may be inter-frame prediction or intra-frame prediction.

[0879] In step 1910, the communication unit 1720 may acquire a bitstream. The communication unit 1720 may receive the bitstream from the encoding apparatus 1600. The processing unit 1710 may store the acquired bitstream in the memory 1740.

[0880] The processing unit 1710 may read the bitstream from the memory 1740 .

[0881] The bitstream may include information about the target block.

[0882] The information about the target block may include transform and quantization coefficients for the target block.

[0883] Furthermore, the bitstream may include the information described above in the embodiments.

[0884] For example, the bitstream may include encoded encoding information or coding information.

[0885] For example, the bitstream may include coding parameters related to the target block and / or properties of the target block.

[0886] The computer-readable storage medium may include a bitstream, and prediction and decoding of the target block may be performed using information about the target block included in the bitstream.

[0887] The computer-readable storage medium may be a non-transitory computer-readable storage medium.

[0888] The bitstream may include encoding information about the target block. The processing unit 1710 may generate information about the target block by performing entropy decoding on the encoding information about the target block.

[0889] At step 1920 , the processing unit 1710 may obtain encoding information from the bitstream.

[0890] The processing unit 1710 may generate encoding information by performing decoding on the encoding information encoded by the bitstream.

[0891] The encoding information may refer to information signaled / encoded / decoded as described in the embodiments. In other words, the encoding information may be information for allowing the decoding apparatus 1700 to perform prediction corresponding to the prediction performed by the encoding apparatus 1600.

[0892] In step 1930 , the processing unit 1710 may determine prediction information to be used for decoding of the target block.

[0893] The prediction information may include the information used for prediction described in the embodiments.

[0894] The prediction information in the decoding apparatus 1700 may be the same as the prediction information in the encoding apparatus 1600. In other words, the processing unit 1710 may generate the same prediction information as the prediction information used in step 1840 in order to perform the same prediction as that performed in step 1840.

[0895] For example, the prediction information may include inter-frame prediction information. For example, the prediction information may include intra-frame prediction information.

[0896] For example, the prediction information may include a motion information candidate list and final motion information, which will be described later. The determination of the prediction information may include configuration of the motion information candidate list and determination of the final motion information.

[0897] The processing unit 1710 may determine the prediction information using the method used in the embodiment.

[0898] The processing unit 1710 may determine prediction information for a target block based on prediction method related information obtained from a bitstream.

[0899] The prediction information may include inter-frame prediction information. The prediction information may include intra-frame prediction information.

[0900] In step 1940 , the processing unit 1710 may perform prediction on the target block using the information about the target block and the prediction information.

[0901] The processing unit 1710 may use the encoding information when performing prediction on the target block. A prediction block may be generated by predicting the target block.

[0902] The information about the target block may include transform and quantization coefficients for the target block. A reconstructed residual block may be generated by applying inverse quantization and inverse transformation to the transform and quantization coefficients for the target block. A reconstructed block may be generated as the sum of the prediction block and the reconstructed residual block.

[0903] Intra Block Copy (IBC)

[0904] An intra block copy (IBC) mode may refer to a mode in which a region indicated by a block vector of a target block is used as a prediction block of the target block.

[0905] The target block may be encoded / decoded in one of an intra prediction mode, an inter prediction mode, and an intra block copy mode.

[0906] 1) When the luminance component and the chrominance component have independent block partition structures (i.e., when a dual-tree structure is used), and 2) when the luminance component and the chrominance component have the same block partition structure (i.e., when a single-tree structure is used), an encoding / decoding method based on prediction using intra-block copying can be used.

[0907] The intra block copy mode may be a method for deriving a block (eg, a reference block or a prediction block) from a previously encoded / decoded region in a target image using a derived block vector (BV).

[0908] The target image may be an image including the target block. Here, because the block is derived from the image including the target block, the intra block copy may correspond to the intra prediction.

[0909] The block vector may refer to an intra block vector.

[0910] The previously encoded / decoded area may be an area in a reconstructed image or a decoded image for a target picture (target image). Here, the area in the reconstructed image may be referred to as a reconstructed area. The area in the decoded image may be referred to as a decoded area.

[0911] The previously encoded / decoded region in the target image may be a reconstructed region to which the at least one loop filtering scheme was not applied.

[0912] In an embodiment, loop filtering may include 1) chroma scaling and luma mapping, 2) deblocking filtering, adaptive sample offset (ASO), and adaptive loop filtering.

[0913] In an embodiment, the previously encoded / decoded region in the target image may be a reconstructed / decoded region on which at least one loop filtering scheme is performed.

[0914] Adaptive Motion Vector Resolution (AMVR)

[0915] In an embodiment, the term "resolution" may refer to the term "motion vector resolution".

[0916] In adaptive motion vector resolution, the resolution of motion vector differences can be adjusted on a block basis.

[0917] The adaptive motion vector resolution information may indicate the resolution of the motion vector difference. The resolution of the motion vector difference of the target block may be determined by signaling / encoding / decoding the adaptive motion vector resolution information.

[0918] The motion vector resolutions applicable to blocks may be the same as or different from each other.

[0919] For example, the resolution of the motion vector applicable to the target block may be determined based on at least one of an encoding parameter, motion information, and mode information of the target block.

[0920] Adaptive motion vector resolution can improve coding efficiency by adjusting the resolution of motion vector differences.

[0921] For example, the adjusted resolution may be one of 16 pixels (pixels), 8 pixels, 4 pixels, full pixels, half pixels, and quarter pixels, and is not limited to the pixel values ​​listed above.

[0922] When the value of the component of the motion vector difference is changed by 1 when the adjusted resolution is n pixels, the position indicated by the motion vector difference may be changed by n pixels. In other words, when the adjusted resolution of the target block is n pixels, each component of the motion vector difference may indicate a reference block in units of n pixels.

[0923] For example, when the motion vector difference actually applied to the target block is (a, b) and the adjusted resolution is p pixels, (a / p, b / p) in addition to (a, b) may be encoded. That is, the motion vector difference signaled / encoded by the encoding device 1600 may be (a / p, b / p). The decoding device 1700 may derive the original motion vector difference (a, b) by multiplying p by the signaled motion vector difference (a / p, b / p).

[0924] Decoder side motion vector derivation

[0925] In an embodiment, a decoder-side motion information derivation method may derive motion information of a target block by 1) deriving initial motion information of the target block and 2) performing refinement on the initial motion information using a predefined operation.

[0926] In an embodiment, refinement of specific information may involve modifying, correcting, or updating the specific information. In an embodiment, the terms "refine," "revise," and "correct" may be used interchangeably. Refinement information may be generated by performing refinement on specific information.

[0927] For example, the decoder-side motion information derivation method 1) can derive motion information offsets and / or motion vector differences for a second direction by applying a predefined operation to the motion information offsets and / or motion vector differences for a first direction, and 2) can refine the motion information by adding the derived motion information offsets and / or derived motion vector differences to the motion information in the second direction.

[0928] For example, the predefined operations may include at least one of mirroring, scaling, and copying, and are not limited to the operations listed above.

[0929] For example, when mirroring is applied to a specific motion vector MV, the result of the mirroring may be -MV.

[0930] For example, when scaling is applied to a specific motion vector MV, the scaled motion vector may be derived by changing the size of the specific motion vector MV based on 1) a picture order count (POC) interval between an image including the specific motion vector MV and a reference image indicated by the MV and 2) a POC interval between an image including a target block and a reference image for the target block. The direction of the specific motion vector MV and the direction of the scaled motion vector generated by applying scaling to the MV may be the same as each other.

[0931] For example, when copying is applied to a specific motion vector MV, the result of the copying may be MV.

[0932] In an embodiment, the decoder-side motion information derivation method may refine the motion information by searching for a position indicated by the initial motion information.

[0933] In an embodiment, a decoder-side motion information derivation method may refine a motion information candidate list by searching an initial motion information candidate list. The initial motion information candidate list may include multiple (initial) motion information. Through the refinement, at least some of the multiple motion information in the motion information candidate list may be refined.

[0934] For example, each piece of motion information in the refined motion information candidate list may be one of 1) initial motion information or 2) refined motion information generated by refining the initial motion information using a decoder-side motion information derivation method. Furthermore, each piece of motion information is not limited to 1) initial motion information and 2) refined motion information. The initial motion information may refer to one piece of motion information among a plurality of pieces included in the initial motion information candidate list.

[0935] In an embodiment, a decoder-side motion information derivation method may 1) compare the matching costs of multiple candidates in an (initial) motion information candidate list, and 2) adjust the order of the multiple candidates in the motion information candidate list based on the matching costs of the multiple candidates. In other words, the multiple candidates in the motion information candidate list may be reordered based on their matching costs. The multiple candidates may be multiple pieces of motion information.

[0936] For example, the plurality of candidates in the (initial) motion information candidate list may be sorted in ascending order of the matching costs of the plurality of candidates.

[0937] In an embodiment, the initial motion information may be 1) the motion information to which the decoder-side motion information derivation method is applied and / or 2) the motion information of each search step to which the decoder-side motion information derivation method is applied.

[0938] For example, the initial motion information may be at least one of a motion vector predictor (MVP), a motion information candidate, and motion information of a neighboring block.

[0939] At least one of the following processes may be applied to the initial motion information. The motion information to which the following processes are applied may be used as the initial motion information of the decoder-side motion information derivation method.

[0940] Process 1: The decoder-side motion information derivation method can be used to refine the initial motion information.

[0941] Process 2: Motion information offset can be used to refine the initial motion information.

[0942] For example, the motion vector difference may be added to the initial motion information.

[0943] For example, a motion information offset may be added to the initial motion information.

[0944] For example, the initial motion information or a portion of the initial motion information may be changed to the same value as the motion information offset.

[0945] For example, when an advanced motion vector prediction (AMVP) mode is used for a target block, the initial motion information may be a motion vector predictor.

[0946] For example, when the AMVP mode is used for the target block, the initial motion information may be the sum of a motion vector predictor (MVP) and a motion vector difference.

[0947] For example, when performing the decoder-side motion information derivation method, at least one motion information offset may be added to at least one of the plurality of pieces of refined motion information generated at the searching step.

[0948] For example, when the decoder-side motion information derivation method consists of N search steps, first refined motion information may be generated in the first search step. A first motion information offset may be added to the first refined motion information. A second search step may be performed on the first refined motion information. Thereafter, a second motion information offset may be added to the (N-1)th refined motion information generated in the (N-1)th search step. The (N)th search step may be performed on the (N-1)th refined motion information.

[0949] For example, information about the motion information offset may be signaled / encoded / decoded.

[0950] The motion information offset can be determined by a rate-distortion optimization process, in which case the coding efficiency of the target block can be improved.

[0951] For example, the motion information offset may be a predefined value. The predefined value may be 0. The predefined value may be (0, 0).

[0952] For example, the motion information offset may refer to a motion vector difference.

[0953] For example, when bidirectional inter prediction is performed on a target block, at least one of the motion information offsets may be added to each of the motion information in the L0 direction and the motion information in the L1 direction.

[0954] For example, at least one of the motion information offsets may be added only to the motion information in the LX direction.

[0955] For example, when bidirectional inter prediction is performed on a target block, the corresponding motion information offset may be added only to the motion information in the LX direction.

[0956] X can be 0, 1, or a positive integer.

[0957] X may be a predefined value.

[0958] For example, the predefined value may be 0.

[0959] For example, the predefined value may be a value indicating a direction having a lower matching cost between the L0 direction and the L1 direction. Here, the matching cost may be a matching cost of motion information.

[0960] For example, the predefined value may be a value indicating a direction having a higher matching cost between the L0 direction and the L1 direction. Here, the matching cost may be a matching cost of motion information.

[0961] When the predefined value X is used, the decoding device 1700 can determine the motion information or direction (the motion information offset will be added to the motion information or direction) without signaling / encoding / decoding the predefined value. Through this determination, the number of bits required for signaling can be reduced.

[0962] Information indicating the predefined value X may be signaled / encoded / decoded. When the information indicating the predefined value X is signaled / encoded / decoded, the predefined value X used in each block may be determined through rate-distortion optimization. This determination may improve prediction performance of the target block.

[0963] For example, the motion information offset may be a motion vector.

[0964] For example, a motion information offset may be determined using a combination of an angle in the angle list and a distance offset in the distance offset list. The angle list may include multiple angles. The distance offset list may include multiple distance offsets.

[0965] In an embodiment, the angle list may be a list of angles with respect to the X-axis / Y-axis.

[0966] In an embodiment, “X axis / Y axis” may be replaced with “width axis / height axis” or “horizontal axis / vertical axis”.

[0967] The angle list may be configured to include angles having a value of “i_ANGLE×π / NUM_ANGLE.” i_ANGLE may be an integer of 0, 1, ..., or (2×NUM_ANGLE−1).

[0968] The number of angles in the angle list may be 2×NUM_ANGLE.

[0969] NUM_ANGLE can be a predefined value. NUM_ANGLE can be 4, 8, or 16. NUM_ANGLE can be a positive integer.

[0970] NUM_ANGLE and / or the angles constituting the angle list may be determined based on encoding parameters of the target block. For example, the encoding parameters may include motion information.

[0971] For example, when the affine mode is used for the target block, the value of NUM_ANGLE may be 8 (or the first value). When the affine mode is not used for the target block, the value of NUM_ANGLE may be 16 (or the second value).

[0972] For example, when the affine mode is used for the current prediction block, the value of NUM_ANGLE may be 8 (or the third value). When the affine mode is not used for the current prediction block, the value of NUM_ANGLE may be 4 (or the fourth value).

[0973] The value of NUM_ANGLE described above may be only an example. Each of the first value, the second value, the third value, and the fourth value may be a specific integer of 1 or greater.

[0974] For example, the distance offset list may be a list indicating distances from a position indicated by the current motion vector. Alternatively, the distance offset list may be a list indicating positions relative to a position indicated by the current motion vector.

[0975] For example, the current motion vector may refer to a motion vector of a target block.

[0976] For example, when the motion vector (of the target block) indicates a first position, the second position can be specified by an angle in the angle list and a distance offset in the distance offset list. Here, the angle in the angle list may indicate the angle between a first line and a second line. The first line may be the x-axis or the y-axis. The second line may be a straight line passing through the first position and the second position. The distance offset in the distance offset list may indicate the distance between the first position and the second position.

[0977] For example, the distance offset list may be configured to include at least one of 0, 1, 2, 4, and 8. In other words, the distance offsets in the distance offset list may be some or all of 0, 1, 2, 4, and 8.

[0978] For example, the distance offset list may include offset -The value corresponding to the pixel is used as the distance offset. Each distance offset in the distance offset list can represent multi offset - The value of the pixel.

[0979] For example, multi offset Can be 1 / 8, 1 / 4, 1 / 2, 1, 2, 4, 8, 16, 32, or a positive integer.

[0980] For example, the distance offsets constituting the distance offset list and the size of the distance offset list may be determined based on an encoding parameter of the target block or motion information of the target block.

[0981] In an embodiment, the size of a list may represent the number of elements (or candidates) in the corresponding list.

[0982] For example, when the affine mode is not used for the target block, the distance offset list may be {4, 8, 16, 32, 64, 128}. When the affine mode is used for the target block, the distance offset list may be {1, 2, 4, 8, 16}.

[0983] The values ​​of the distance offsets in the aforementioned distance offset list may be only examples. The value of each distance offset may be a specific integer of 1 or greater.

[0984] For example, a position determined using a combination of a specific angle θ and a specific distance offset ρ may indicate a position at a distance ρ in a direction at an angle θ from the X-axis / Y-axis.

[0985] For example, a motion vector determined using a combination of a specific angle θ and a specific distance offset ρ may refer to a motion vector indicating a position at a distance ρ in a direction at an angle θ from the X-axis / Y-axis.

[0986] For example, a motion information offset may be added to the reference picture index.

[0987] For example, when the value of the reference image index of the target block is a first value and the value of the motion information offset added to the reference image index is a second value, the value of the reference image index of the target block may be changed to the first value + the second value. The first value may be 0, 1, or a positive integer. The second value may be -2, -1, 0, 1, 2, or an integer.

[0988] For example, when the value of the reference image index of the target block is a first value and the value of the motion information offset added to the reference image index is a second value, the value of the reference image index of the target block may be changed to the second value. The first value may be 0, 1, or a positive integer. The second value may be 0, 1, or a positive integer.

[0989] For example, when the decoder-side motion information derivation method is performed, at least one of the plurality of pieces of refined motion information generated in the corresponding search step may be changed to have a value equal to a value of the at least one motion information offset.

[0990] For example, when the decoder-side motion information derivation method consists of N search steps, the motion vector of the refined motion information generated in the first search step may be changed to the first motion information offset. A second search step may be performed on the motion information with the changed motion vector. Thereafter, the reference image index of the refined motion information generated in the (N-1)th search step may be changed to the second motion information offset. The Nth search step may be performed on the motion information with the changed reference image index.

[0991] For example, the motion information offset may refer to a coding parameter. Alternatively, the motion information offset may refer to motion information.

[0992] For example, the motion information offset may refer to a motion vector or a reference picture index.

[0993] For example, when bidirectional inter prediction is used for a target block, motion information in an L0 direction and motion information in an L1 direction may be changed to values ​​equal to motion information offsets in corresponding directions.

[0994] For example, only the motion information in the LX direction among the plurality of pieces of motion information in the plurality of directions may be changed to a value equal to the motion information offset.

[0995] For example, when bidirectional inter prediction is used for a target block, only motion information in an LX direction among a plurality of pieces of motion information in a plurality of directions may be changed to a value equal to the motion information offset.

[0996] X can be 0, 1, or a positive integer.

[0997] X may be a predefined value.

[0998] For example, the predefined value may be 0.

[0999] For example, the predefined value may be a value corresponding to a direction having a lower matching cost of motion information, in the L0 direction and the L1 direction.

[1000] For example, the predefined value may be a value corresponding to a direction having a higher matching cost of motion information, in the L0 direction and the L1 direction.

[1001] When a predefined value X is used, the decoder side can determine X without signaling / encoding / decoding information indicating the predefined value. Through this determination, the number of bits required for signaling can be reduced.

[1002] Information indicating X may be signaled / encoded / decoded. When information indicating X is signaled / encoded / decoded, X used in each block may be determined by rate-distortion optimization. This determination may improve prediction performance of the target block.

[1003] For example, the search in an embodiment may be performed using calculation of a cost function for determining similarities between NUM_TEMPLATE_COMPARE templates.

[1004] Searching in the embodiments may refer to a process of determining motion information that satisfies specific conditions within a predefined search range.

[1005] For example, the motion information satisfying a specific condition may refer to motion information having the lowest matching cost among a plurality of pieces of motion information within a search range.

[1006] In an embodiment, NUM_TEMPLATE_COMPARE may be 0, 1, 2, or a positive integer.

[1007] Refined motion information may refer to motion information that meets specific conditions.

[1008] The refined motion information may refer to at least one of: 1) motion information determined by a decoder-side motion information derivation method, and 2) motion information determined at each search step of the decoder-side motion information derivation method.

[1009] In an embodiment, the search range may be a specific range with the position indicated by the initial motion information as the center. In other words, the center of the search range may be the position indicated by the initial motion information. The specific range may be a range with a predefined area.

[1010] For example, the center of the search range may be a point indicated by the initial motion information. The search range may have a rectangular shape with a horizontal length of SR_X and a vertical length of SR_Y. Alternatively, the search range may have a diamond shape with a horizontal length of SR_X and a vertical length of SR_Y. However, the shape and size of the search range are not limited to the above embodiment.

[1011] Each of SR_X and SR_Y may be a predefined positive integer.

[1012] Each template may be a subset of pixels included in a specific area.

[1013] For example, the template of the target block may include one or more of the following: 1) a subset of pixels included in a TMSIZE_LEFT line adjacent to the left side of the target block and 2) a subset of pixels included in a TMSIZE_ABOVE line adjacent to the top (top) of the target block. However, the relationship between the position of the template and the position of the target block or the template configuration method is not limited to the above embodiment.

[1014] For example, the template of the reference block may include one or more of the following: 1) a subset of pixels included in a TMSIZE_LEFT line adjacent to the left side of the reference block, and 2) a subset of pixels included in a TMSIZE_ABOVE line adjacent to the top of the reference block. However, the relationship between the position of the template and the position of the reference block or the template configuration method is not limited to the above embodiment.

[1015] Each of TMSIZE_LEFT and TMSIZE_ABOVE may be a predefined integer of 0 or greater. TMSIZE_LEFT and TMSIZE_ABOVE may be the same as or different from each other.

[1016] For example, a cost function of the template of the target block and a cost function of the template of the reference block in the L0 direction and / or the L1 direction may be calculated.

[1017] For example, a cost function of the template of the reference block in the L0 direction and a cost function of the template of the reference block in the L1 direction may be calculated.

[1018] In an embodiment, the cost function may be one or more of the sum of absolute differences (SAD), the sum of absolute transformed differences (SATD), the mean removed sum of absolute differences (MR-SAD), the mean square error (MSE), and the sum of squared errors (SSE). However, the cost function may not be limited to the items listed above.

[1019] When a decoder-side motion information derivation method is performed on a target block, a cost function used in the decoder-side motion information derivation method may be predefined.

[1020] When a decoder-side motion information derivation method is performed on a target block, information about a cost function used in the decoder-side motion information derivation method may be signaled / encoded / decoded.

[1021] The template may be all or part of block information of one or more blocks included in a specific area.

[1022] For example, the template of the target block may include one or more of the following: 1) a subset of block information of blocks included in a TMSIZE_LEFT line adjacent to the left side of the target block and 2) a subset of block information of blocks included in a TMSIZE_ABOVE line adjacent to the top of the target block. However, the relationship between the positions of the template and the target block or the template configuration method is not limited to the above embodiment.

[1023] For example, the template of the reference block may include one or more of the following: 1) a subset of block information of blocks included in a TMSIZE_LEFT line adjacent to the left side of the reference block and 2) a subset of block information of blocks included in a TMSIZE_ABOVE line adjacent to the top of the reference block. However, the relationship between the positions of the template and the reference block or the template configuration method is not limited to the above embodiment.

[1024] Each of TMSIZE_LEFT and TMSIZE_ABOVE may be a predefined integer of 0 or greater.

[1025] In an embodiment, the block information may refer to at least one of information about a target block, information about a neighboring block, information about a reference block, and information about a col block.

[1026] Furthermore, the block information may include at least one of the encoding parameters.

[1027] Also, the block information may include at least one of a plurality of pieces of information used in inter prediction, intra prediction, transformation, inverse transformation, quantization, inverse quantization, entropy encoding / decoding, and loop filtering.

[1028] That is, the block information may include at least one value, combination value, or statistical value of the following: block size, block depth, block partition information, block shape (square or non-square), information indicating whether the block is partitioned in a quadtree form, information indicating whether the block is partitioned in a binary tree form, partition direction in the binary tree form (horizontal or vertical), partition shape in the binary tree form (symmetric or asymmetric), prediction mode (intra-frame prediction or inter-frame prediction), intra-frame luma prediction mode / direction, intra-frame chroma prediction mode / direction, intra-frame partition information, inter-frame partition information, coding block partition flag, block partition flag, transform block partition flag, reference sample filter taps, reference sample filter coefficients, block filter taps, block filter coefficients, block boundary filter taps, block boundary filter coefficients, motion vector (L0, L 1, L2, and L3), motion vector difference (motion vector difference of at least one of L0, L1, L2, and L3), inter prediction direction (inter prediction direction of at least one of unidirectional prediction and bidirectional prediction), reference image (picture) index (reference image index of at least one of L0, L1, L2, and L3), inter prediction indicator, prediction list utilization flag, reference image list, motion vector prediction index, motion vector prediction candidates, motion information candidate list, information indicating whether merge mode is used, merge index, merge candidate, merge candidate list, information indicating whether skip mode is used, interpolation filter type, interpolation filter taps, interpolation filter coefficients, motion vector magnitude, precision of motion vector representation (unit or resolution of representing motion vectors),such as integer samples, 1 / 2 samples, 1 / 4 samples, 1 / 8 samples, 1 / 16 samples, and 1 / 32 samples), transform type, transform size, information indicating whether primary transform is used, information indicating whether secondary transform is used, primary transform index, secondary transform index, information indicating whether a residual signal exists, coding block pattern, coding block flag, quantization parameter, residual quantization parameter, quantization matrix, information indicating whether a loop filter is applied, loop filter coefficients, loop filter taps, loop filter shape / type, information indicating whether a deblocking filter is applied, deblocking filter coefficient number, deblocking filter taps, deblocking filter strength, deblocking filter shape / type, information indicating whether adaptive sample offset is applied, adaptive sample offset value, adaptive sample offset category, adaptive sample offset type, information indicating whether adaptive loop filter is applied, adaptive loop filter coefficient, adaptive loop filter taps, adaptive loop filter shape / type, binarization / debinarization method, context model determination method, context model update method, information indicating whether normal mode is performed, information indicating whether bypass mode is performed, context binary bits, bypass Binary bit, significant coefficient flag, last significant coefficient flag, flag indicating whether encoding / decoding of the unit of the coefficient group is applied, position of the last significant coefficient, flag indicating whether the coefficient value is greater than 1, flag indicating whether the coefficient value is greater than 2, flag indicating whether the coefficient value is greater than 3, residual coefficient value information, positive and negative sign information, reconstructed luminance samples, reconstructed chrominance samples, residual luminance samples, residual chrominance samples, luminance transform coefficient, chrominance transform coefficient, luminance quantization level, chrominance quantization level, transform coefficient horizontal scanning method, size of the motion vector search range at the decoder side, at the decoder side The shape of the motion vector search range, the number of motion vector search iterations at the decoder side, CTU size information, minimum block size information, maximum block size information, maximum block depth information, minimum block depth information, slice identification information, slice partition information, tile identification information, tile block type, tile block partition information, input sample bit depth, reconstructed sample bit depth, residual sample bit depth, transform coefficient bit depth, quantization level bit depth, indicator for overlapped block motion compensation (OBMC) mode, indicator for local illumination compensation (LIC) mode, and motion information offset.

[1029] For example, the cost function may be a function that determines similarities between pieces of motion information of templates and / or between encoding parameters.

[1030] In an embodiment, the similarity between the first value and the second value may be determined using at least one of: 1) a difference between the two values, 2) a ratio of the two values, and 3) an operation of comparing the difference between the two values ​​to a specific value.

[1031] For example, the fact that multiple pieces of motion information (or encoding parameters) of the templates are similar to each other may mean that the difference between the multiple pieces of motion information (or encoding parameters) of the templates is less than or equal to the predefined positive number THRES_PARAMETER. However, the criteria based on which similarity is determined are not limited to the above comparison.

[1032] For example, the fact that multiple pieces of motion information (or encoding parameters) of the templates are similar to each other may mean that the sum of the differences between the multiple pieces of motion information (or encoding parameters) of the templates is less than or equal to the predefined positive number THRES_PARAMETER. However, the criteria based on which similarity is determined are not limited to the above comparison.

[1033] For example, the fact that multiple pieces of motion information (or encoding parameters) of the templates are similar to each other may mean that the ratio of the multiple pieces of motion information (or encoding parameters) of the templates is less than or equal to the predefined positive number WEIGHT_THRES_PARAMETER. However, the criteria based on which similarity is determined are not limited to the above comparison.

[1034] For example, the fact that multiple pieces of motion information (or encoding parameters) of the templates are similar to each other may mean that the ratio of the multiple pieces of motion information (or encoding parameters) of the templates is equal to or greater than a predefined positive number WEIGHT_THRES_PARAMETER. However, the criteria based on which similarity is determined are not limited to the above comparison.

[1035] For example, the fact that multiple motion information (or encoding parameters) of a template are similar to each other may mean that the number of multiple identical motion information (or encoding parameters) in the template is equal to or greater than a predefined positive number THRES_PARAMETER, but the criteria for determining the similarity are not limited to the above comparison.

[1036] Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER may be a predefined positive integer.

[1037] Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER can be set based on motion information (or encoding parameters). Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER can vary according to the motion information (or encoding parameters). Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER can be set based on the type of motion information (or encoding parameters). Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER can vary according to the type of motion information (or encoding parameters).

[1038] For example, the refined motion information candidate list may include DMVG_NUM pieces of motion information.

[1039] DMVG_NUM can be a predefined positive integer of 1 or greater.

[1040] Information indicating DMVG_NUM may be signaled / encoded / decoded.

[1041] DMVG_NUM may be equal to the size of the initial motion information candidate list.

[1042] DMVG_NUM may be different from the size of the initial motion information candidate list.

[1043] For example, DMVD_NUM may be smaller than the size of the initial motion information candidate list.

[1044] Each candidate in the refined motion information candidate list may be initial motion information or refined motion information generated based on the initial motion information.

[1045] For example, the refined motion information candidate list may be configured to include DMVG_NUM pieces of initial motion information having the lowest matching cost among the candidates in the initial motion information candidate list.

[1046] In an embodiment, “a plurality of pieces of information having the lowest matching costs” may refer to “a plurality of pieces of information selected in ascending order of matching costs”.

[1047] For example, a plurality of pieces of refined motion information may be generated by applying refinement to candidates in the initial motion information candidate list.The refined motion information candidate list may be configured to include DMVG_NUM pieces of refined motion information having the lowest matching cost among the plurality of pieces of refined motion information.

[1048] For example, in a case where the motion information in the LX direction has been determined as MVP_LX when refining the initial motion information candidate list for the L(1-X) direction, each candidate in the initial motion information candidate list for the L(1-X) direction may be refined to minimize the bilateral matching cost with MVP_LX.

[1049] For example, in a case where the motion information in the LX direction has been determined as MVP_LX when refining the initial motion information candidate list for the L(1-X) direction, the order of the candidates in the initial motion information candidate list for the L(1-X) direction may be reordered in ascending order of bilateral matching costs with MVP_LX.

[1050] X can be 0, 1, or a positive integer.

[1051] X may be a predefined value. The predefined value may be 0.

[1052] For example, the predefined value may refer to a value indicating a direction having a lower matching cost between the L0 direction and the L1 direction. Here, the matching cost of a specific direction may be a matching cost of motion information in the specific direction.

[1053] For example, the predefined value may refer to a value indicating a direction having a higher matching cost between the L0 direction and the L1 direction. Here, the matching cost of a specific direction may be a matching cost of motion information in the specific direction.

[1054] Information indicating X may be signaled / encoded / decoded.

[1055] The information indicating X can be determined through rate-distortion optimization processing. Through this determination, the encoding efficiency of the target block can be improved.

[1056] For example, when performing searches in a decoder-side motion information derivation method, a position-based weight w can be used to match the cost of motion information in each search process.

[1057] For example, the matching cost of specific motion information in each search process may be refined by an operation using a weight w. The operation may be at least one of four arithmetic operations.

[1058] Whether to use the weight w and / or the value of the weight w may be determined based on the position indicated by the initial motion information.

[1059] The initial motion information may refer to initial motion information of a decoder-side motion information derivation method. Alternatively, the initial motion information may refer to motion information that is a target for which refinement is to be performed in each search process of the decoder-side motion information derivation method.

[1060] In an embodiment, the weight w may be a value by which the matching cost of specific motion information is multiplied in each search process.

[1061] For example, the weight w may be determined based on the distance between the position indicated by the specific motion information and the position indicated by the initial motion information.

[1062] For example, whether the weight w has a value other than 1 during the search process may be determined based on the distance between the position indicated by the specific motion information and the position indicated by the initial motion information.

[1063] For example, whether the weight w is multiplied by the matching cost of the specific motion information during the search process may be determined based on the distance between the position indicated by the specific motion information and the position indicated by the initial motion information.

[1064] The weight w may be predefined.

[1065] For example, the matching cost at the position indicated by the initial motion information may be multiplied by the predefined weight W. Here, W may be a value less than or equal to 1, or a value equal to or greater than 1.

[1066] For example, when the AMVP mode is used for the target block, W at the position indicated by the initial motion information may be a value equal to or greater than 1.

[1067] For example, when the merge mode is used for the target block, W at the position indicated by the initial motion information may be a value less than or equal to 1.

[1068] For example, when the AMVP mode is used for the target block, the weight W may have a larger value since the specific position is closer to the position indicated by the initial motion information.

[1069] For example, when the merge mode is used for the target block, the weight W may have a smaller value since the specific position is closer to the position indicated by the initial motion information.

[1070] For example, when the value of the weight W multiplied by the matching cost of the specific motion information is less than or equal to 1 when performing a specific search step in the decoder-side motion information derivation method, the possibility that the initial motion information of the current search step will be refined into the specific motion information may increase.

[1071] For example, when the value of the weight W multiplied by the matching cost of the specific motion information is equal to or greater than 1 when performing a specific search step in the decoder-side motion information derivation method, the possibility that the initial motion information of the current search step will be refined into the specific motion information can be reduced.

[1072] The decoder-side motion information derivation method may comprise one or more search steps.

[1073] At each search step, refined motion information may be determined by searching from the initial motion information at the search step.

[1074] The two different search steps may differ from each other in at least one of initial motion information, template configuration method, type of cost function for determining similarity, unit for performing refinement of motion information, search range, search pattern, search resolution, and search method.

[1075] For example, each search step of the decoder-side motion information derivation method may be considered as an independent decoder-side motion information derivation method implemented using one search step.

[1076] For example, a first decoder-side motion information derivation method consisting of L search steps and a second decoder-side motion information derivation method consisting of M search steps can be considered as one decoder-side motion information derivation method consisting of N search steps. N can be the sum of L and M.

[1077] The types of search methods may differ from each other according to one or more of the following: 1) search pattern, 2) search resolution, 3) search range, and 4) unit for deriving motion information. However, the criteria based on which the types of search methods are classified are not limited to such conditions.

[1078] The search resolution may be one of 4 pixels, full pixels, half pixels, and quarter pixels. However, the search resolution is not limited to the above pixel values.

[1079] The search pattern may be one of a diamond pattern, a cross pattern, and a full search pattern. However, the search pattern is not limited to the patterns listed above.

[1080] Searching using a diamond pattern may refer to searching for one or more positions of (0, 2×RR), (RR, RR), (2×RR, 0), (RR, -RR), (0, -RR), (-RR, -RR), (-RR, 0), (-RR, RR), and (0, 0) when (0, 0) represents the position indicated by the initial motion information. RR may refer to a search resolution and may be a predefined positive number.

[1081] Searching using a cross pattern may be an operation of searching for one or more of the positions of (0, RR), (RR, 0), (0, -RR), (-RR, 0), and (0, 0) when (0, 0) represents the position indicated by initial motion information.

[1082] Searching using the full search style can be an operation that searches all locations within a predefined search scope.

[1083] For example, assuming that FS_i has a value ranging from -FS_X to FS_X and FS_j has a value ranging from -FS_Y to FS_Y, searching using the full search pattern may refer to an operation of searching for a position corresponding to (FS_i × RR, FX_j × RR). Here, (0, 0) may be the position indicated by the initial motion information. However, the search range is not limited to the above positions. Each of FS_X and FS_Y may be a predefined positive number.

[1084] For example, the decoder-level motion information derivation method may be configured in the following order: 1) a step of deriving motion information for an entire block, and 2) a step of deriving motion information for a sub-block of the block. However, the method of deriving motion information performed at the respective steps and the order of the respective steps are not limited to the method and order described above in the embodiment.

[1085] For example, the decoder-side motion information derivation method may consist of N search steps. In a first search step, motion information for a first block may be derived. In a second search step, subsequent to the first search step, motion information for a second block may be derived. Here, the units of the second block may be smaller than the units of the first block. In other words, the second block may be a subblock of the first block. Alternatively, the units of the first block and the second block may be identical. In other words, the first block and the second block may be identical.

[1086] For example, assuming that derivation of motion information of the entire block is performed in the first search step, derivation of motion information of the entire block or motion information of sub-blocks of the block may be performed in the second search step.

[1087] The search method of each step may be predefined based on the encoding parameters or motion information of the target block.

[1088] When the decoder-side motion information derivation method consists of two or more search steps, initial motion information inputted in a specific search step after the first search step may be motion information refined in a search step before the specific search step.

[1089] For example, the decoder-side motion information derivation method can consist of three search steps. 1) In the first search step, the motion information can be refined using bilateral matching for the entire block. 2) In the second search step, the motion information can be refined using bilateral matching on a sub-block basis, with each sub-block having a size of 16×16. 3) In the third search step, the motion information can be refined using optical flow-based motion information prediction on a sub-block basis, with each sub-block having a size of 8×8.

[1090] The motion information refined by the decoder-side motion information derivation method according to an embodiment may replace the initial motion information. For example, the refined motion information may be used to determine the motion information for inter-frame prediction in the L0 direction and / or L1 direction of the target block.

[1091] The decoder-side motion information derivation method according to an embodiment may be one or more of template matching (TM), bilateral matching (BM), decoder-side motion vector refinement (DMVR), symmetric motion vector difference (SMVD), and optical flow-based motion information prediction. However, the decoder-side motion information derivation method is not limited to the above methods.

[1092] In an embodiment, a description of executing at least one of the methods of TM, BM, DMVR, SMVD, and optical flow-based motion information prediction in the decoder-side motion information derivation invention when a specific condition is met may include a description of not executing the remaining methods of the decoder-side motion information derivation invention except for at least one of TM, BM, DMVR, SMVD, and optical flow-based motion information prediction when the specific condition is met. For example, a description of executing TM when a specific condition is met may mean that BM is not executed when the specific condition is met. A description of executing BM when a specific condition is met may mean that TM is not executed when the specific condition is met.

[1093] Symmetrical Motion Vector Difference (SMVD) mode

[1094] Figure 20 Motion vector differences in a symmetric motion vector difference pattern according to an example are shown.

[1095] The SMVD mode may be a prediction mode using bidirectional prediction and may be one of the AMVP modes.

[1096] For example, when bidirectional inter prediction is used for a target block, the symmetric motion vector difference mode may be signaled as a sub-mode of the AMVP mode.

[1097] For example, when the symmetric motion vector difference mode is used, the reference image index may be determined based on a predefined method.

[1098] For example, the signaling of the reference picture (picture) list may be skipped.

[1099] For example, when using the SMVD mode, a pair of reference images may be specified in the reference image list for the L0 direction and the reference image list for the L1 direction. Here, the specified reference images may be the first reference image in the reference image list L0 and the second reference image in the reference image list L1. The first POC interval may be the difference between the POC of the target image and the POC of the first reference image. The second POC interval may be the difference between the POC of the target image and the POC of the second reference image. The target image may be an image including a target block. The first POC interval and the second POC interval may be equal to each other. The direction of the first reference image relative to the target image and the direction of the second reference image relative to the target image may be opposite to each other.

[1100] In other words, the reference picture index used in the SMVD mode may be a reference picture index of a reference picture having the same POC interval.

[1101] When using SMVD mode, a motion vector prediction index can be signaled / encoded / decoded for the reference image in each direction, and information about one motion vector difference can be signaled / encoded / decoded. In SMVD mode, two MVP indices and one MVD can be signaled.

[1102] When the SMVD mode is used, the motion vector difference of the reference image in the corresponding direction may include a first motion vector difference that is encoded / decoded and a second motion vector difference that is symmetrical to the first motion vector difference.

[1103] For example, when only the motion vector difference MVD0 in the L0 direction is signaled / encoded / decoded, the motion vector difference MVD1 in the L1 direction may be -MVD0.

[1104] Figure 21 Syntax elements signaled / encoded / decoded in symmetric motion vector difference mode are shown according to an example.

[1105] x0 and y0 may mean that the upper left position of the luma block in the target block is (x0, y0).

[1106] sps_smvd_enabled_flag may be a syntax element determined in a sequence parameter set and may be an indicator indicating whether a symmetric motion vector difference mode is to be enabled in a target sequence.

[1107] When the value of sps_smvd_enabled_flag is the first value, the symmetric motion vector difference mode may be disabled in the target sequence. The first value may be 0 or false.

[1108] When the value of sps_smvd...

Claims

1. An image encoding method, comprising: determining prediction information to be used for encoding of a target block; generating encoded encoded information by performing encoding on the encoded information; as well as A bitstream including the encoded encoding information is generated.

2. The image encoding method according to claim 1, wherein: The prediction information includes a motion information candidate list and final motion information.

3. The image encoding method according to claim 1, further comprising: Prediction is performed on the target block using inter prediction, intra block copy, or intra template matching prediction.

4. The image encoding method according to claim 3, wherein: A motion vector search method for prediction is used, and The motion vector search method is template matching.

5. The image encoding method according to claim 4, wherein: Inter prediction is bidirectional prediction, and The motion vector is fixed for one of the two directions in bidirectional prediction.

6. The image encoding method according to claim 5, wherein: A motion search is performed for the remaining one of the two directions.

7. The image encoding method according to claim 3, wherein: The motion vector search method for inter prediction is used, and The motion vector search method is bilateral matching.

8. An image decoding method, comprising: obtaining a bitstream including encoded encoding information; generating coded information by performing decoding on the coded coded information; as well as Prediction information to be used for decoding of the target block is determined.

9. The image decoding method according to claim 8, wherein: The prediction information includes a motion information candidate list and final motion information.

10. The image decoding method according to claim 8, further comprising: Prediction is performed on the target block based on the prediction information using inter prediction, intra block copy, or intra template matching prediction.

11. The image decoding method according to claim 10, wherein: A motion vector search method for prediction is used, and The motion vector search method is template matching.

12. The image decoding method according to claim 11, wherein: Inter prediction is bidirectional prediction, and The motion vector is fixed for one of the two directions of bidirectional prediction.

13. The image decoding method according to claim 12, wherein: A motion search is performed for the remaining one of the two directions.

14. The image decoding method according to claim 10, wherein: The motion vector search method for inter prediction is used, and The motion vector search method is bilateral matching.

15. A computer-readable storage medium for storing a bit stream for image decoding, wherein: The bitstream includes encoding information encoded, The encoding information is generated by performing decoding on the encoded encoding information, and Prediction information to be used for decoding of the target block is determined.

16. The computer-readable storage medium of claim 15, wherein: The prediction information includes a motion information candidate list and final motion information.

17. The computer-readable storage medium of claim 15, wherein: Prediction of the target block is performed based on the prediction information using inter prediction, intra block copy, or intra template matching prediction.

18. The computer-readable storage medium of claim 17, wherein: A motion vector search method for prediction is used, and The motion vector search method is template matching.

19. The computer-readable storage medium of claim 18, wherein: Inter prediction is bidirectional prediction, and The motion vector is fixed for one of the two directions of bidirectional prediction.

20. The computer-readable storage medium of claim 19, wherein: A motion search is performed for the remaining one of the two directions.

Images