KR20230046265A - Method, apparatus and recording medium for encoding/decoding image - Google Patents

Abstract

A method, device, and recording medium for video encoding/decoding are disclosed. In a typical video encoding/decoding method, a decoder-side motion information derivation method may be used in a limited manner. Therefore, improvements in encoding efficiency due to the decoder-side motion information derivation method may also be limited. In embodiments, a motion information search method used in inter-prediction is disclosed. Encoding efficiency in inter-prediction can be improved by using various motion detection methods.

Description

Method, apparatus and recording medium for video encoding/decoding {METHOD, APPARATUS AND RECORDING MEDIUM FOR ENCODING/DECODING IMAGE}

The present invention relates to a method, apparatus and recording medium for video encoding/decoding.

Through the continuous development of the information communication industry, broadcasting services having high definition (HD) resolution have been spread worldwide. Through this proliferation, many users have become accustomed to high-resolution and high-definition images and/or videos.

In order to satisfy users' demand for high image quality, many organizations are accelerating the development of next-generation video devices. Users' interest in Ultra High Definition (UHD) TV, which has four times higher resolution than FHD TV, as well as High Definition TV (HDTV) and Full HD (FHD) TV. has increased, and according to this increase in interest, an image encoding/decoding technology for an image having a higher resolution and quality is required.

As video compression technology, there are various technologies such as inter prediction technology, intra prediction technology, transformation and quantization technology, and entropy encoding technology.

The inter-prediction technique is a technique of predicting a value of a pixel included in a current picture using pictures before and/or after the current picture. Intra-prediction technology is a technology of predicting a value of a pixel included in a current picture by using information about pixels in the current picture. The transformation and quantization technique is a technique for compressing the energy of the residual image. The entropy coding technique assigns short codes to values with a high frequency of occurrence and assigns long codes to values with a low frequency of occurrence.

Using this video compression technology, data for video can be effectively compressed, transmitted, and stored.

An embodiment may provide an apparatus, method, and recording medium for performing encoding/decoding of a target block using inter prediction.

In one aspect, determining prediction information to be used for encoding of a target block; generating encoded coding information by performing encoding on the coding information; and generating a bitstream including the encoded coding information.

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

The image encoding method may further include performing inter prediction on the target block.

The motion vector search method for the inter prediction may be used.

The motion vector search method may be template matching.

The inter prediction may be bidirectional prediction.

A motion vector may be fixed for one of the directions of the bi-directional prediction.

Motion search may be performed in the other one of the directions.

The motion vector search method for the inter prediction may be used.

The motion vector retrieval method may be bilateral matching.

In another aspect, obtaining a bitstream including coded coding information; generating coding information by decoding the encoded coding information; and determining prediction information to be used for decoding a target block.

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

The image decoding method may further include performing inter prediction on the target block using the prediction information.

The motion vector search method for the inter prediction may be used.

The motion vector search method may be template matching.

The inter prediction may be bidirectional prediction.

A motion vector may be fixed for one of the directions of the bi-directional prediction.

Motion search may be performed in the other one of the directions.

The motion vector search method for the inter prediction may be used.

The motion vector retrieval method may be bilateral matching.

On the other hand, in a computer-readable recording medium storing a bitstream for video decoding, the bitstream includes encoded coding information, and decoding of the encoded coding information is performed so that the coding information is A computer readable recording medium in which prediction information to be generated and used for decoding a target block is determined.

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

Inter prediction may be performed on the target block using the prediction information.

The motion vector search method for the inter prediction may be used.

The motion vector search method may be template matching.

The inter prediction may be bidirectional prediction.

A motion vector may be fixed for one of the directions of the bi-directional prediction.

Motion search may be performed in the other one of the directions.

An apparatus, method, and recording medium for performing encoding/decoding of a target block using inter prediction are provided.

1 is a block diagram showing a configuration according to an embodiment of an encoding device to which the present invention is applied.
2 is a block diagram showing a configuration according to an embodiment of a decoding device to which the present invention is applied.
3 is a diagram schematically illustrating a division structure of an image when encoding and decoding an image.
4 is a diagram illustrating a form of a prediction unit that a coding unit may include.
5 is a diagram illustrating a form of a transform unit that may be included in a coding unit.
6 shows division of a block according to an example.
7 is a diagram for explaining an embodiment of an intra prediction process.
8 is a diagram for explaining reference samples used in an intra prediction process.
9 is a diagram for explaining an embodiment of an inter prediction process.
10 shows spatial candidates according to an example.
11 illustrates an order of adding motion information of spatial candidates to a merge list according to an example.
12 illustrates a process of transformation and quantization according to an example.
13 illustrates diagonal scanning according to an example.
14 illustrates horizontal scanning according to an example.
15 illustrates vertical scanning according to an example.
16 is a structural diagram of an encoding device according to an embodiment.
17 is a structural diagram of a decryption device according to an embodiment.
18 is a flowchart of a method of predicting a target block and generating a bitstream according to an embodiment.
19 is a flowchart of a method of predicting a target block using a bitstream according to an embodiment.
20 illustrates motion vector differences in a symmetric motion vector difference mode according to an example.
21 illustrates syntax elements signaled/encoded/decoded in a symmetric motion vector differential mode according to an example.
22 illustrates matching on both sides according to an example.
23 illustrates template matching according to an embodiment.
24 illustrates a first relationship between search patterns and resolutions according to an example.
25 illustrates a second relationship between search patterns and resolutions according to an example.
26 illustrates a third relationship between search patterns and resolutions according to an example.
27 illustrates a fourth relationship between search patterns and resolutions according to an example.
28 illustrates a fifth relationship between search patterns and resolutions according to an example.
29 illustrates a sixth relationship between search patterns and resolutions according to an example.
30 illustrates a method of constructing a first template in an affine mode according to an example.
31 illustrates a second template configuration method in an affine mode according to an example.
32 is a first flowchart illustrating a method of selecting a method of constructing a motion information candidate list according to an example.
33 is a second flowchart illustrating a method of selecting a method of constructing a motion information candidate list according to an example.
34 is a third flowchart illustrating a method of selecting a method of constructing a motion information candidate list according to an example.
35 is a fourth flowchart illustrating a method of selecting a method of configuring a motion information candidate list according to an example.
36 illustrates reconstruction of a motion information candidate list according to an embodiment.
37 illustrates setting of flags according to values of motion information indices according to an example.
38 illustrates a first setting for an index according to values of motion information indexes according to an example.
39 illustrates a second setting for an index according to values of motion information indexes according to an example.
40 illustrates a first setting for an index according to values of reference picture indexes according to an example.
41 illustrates a second setting for an index according to values of reference picture indexes according to an example.
42 illustrates setting of flags according to values of reference picture indexes according to an example.
43 illustrates a method of determining a sign for each component of a motion vector difference according to an embodiment.
44 illustrates a first code for signaling/encoding/decoding motion vector differences according to an example.
45 illustrates a second code for signaling/encoding/decoding motion vector differences according to an example.
46 illustrates a third code for signaling/encoding/decoding motion vector differences according to an example.
47 illustrates a fourth code for signaling/encoding/decoding motion vector differences according to an example.
48 illustrates a fifth code for signaling/encoding/decoding motion vector differences according to an example.
49 illustrates a signaling method of information for a motion vector search mode according to an example.
50 may be a first code representing coding information signaled in relation to a block division structure according to an example.
51 may be a second code representing coding information signaled in relation to a block division structure according to an example.

Since the present invention can make various changes and have various embodiments, specific embodiments will be illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS For detailed descriptions of exemplary embodiments described below, reference is made to the accompanying drawings, which illustrate specific embodiments by way of example. These embodiments are described in sufficient detail to enable those skilled in the art to practice the embodiments. It should be understood that the various embodiments are different, but need not be mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be implemented in one embodiment in another embodiment without departing from the spirit and scope of the invention. Additionally, it should be understood that the location or arrangement of individual components within each disclosed embodiment may be changed without departing from the spirit and scope of the embodiment. Accordingly, the detailed description set forth below is not to be taken in a limiting sense, and the scope of the exemplary embodiments, if properly described, is limited only by the appended claims, along with all equivalents as claimed by those claims.

Like reference numbers in the drawings indicate the same or similar function throughout the various aspects. The shapes and sizes of elements in the drawings may be exaggerated for clarity.

In the present invention, terms such as first and second may be used to describe various components, but the components should not be limited by the terms. These terms are only used for the purpose of distinguishing one component from another. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element, without departing from the scope of the present invention. The term "and/or" may include any combination of a plurality of related listed items or any of a plurality of related listed items.

When a component is referred to as being “connected” or “connected” to another component, the two components may be directly connected or connected to each other, but It will be understood that other components may exist in the middle of the two components. On the other hand, when a component is referred to as "directly connected" or "directly connected" to another component, it should be understood that no other component exists in the middle of the above two components. something to do.

Elements appearing in the embodiments are shown independently to represent different characteristic functions, and do not mean that each element is composed of separate hardware or a single software unit. That is, each component is listed and included as each component for convenience of description, and at least two of each component are combined to form one component, or one component is divided into a plurality of components to perform functions. It can be performed, and integrated embodiments and separate embodiments of each of these components are also included in the scope of the present invention as long as they do not depart from the essence of the present invention.

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 context clearly dictates otherwise. In embodiments, the terms "comprise" or "have" are intended to indicate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, but one or more other It should be understood that the presence or addition of features, numbers, steps, operations, components, parts, or combinations thereof is not precluded. That is, the description of "including" a specific configuration in the embodiments does not exclude configurations other than the corresponding configuration, and means that additional configurations may also be included in the practice of the present invention or the scope of the technical idea of the present invention. .

In embodiments, the term “at least one” may mean one or more numbers such as 1, 2, 3, and 4. In embodiments, the term “a plurality of” may mean one of two or more numbers, such as 2, 3, and 4.

Some of the components of the embodiments are not essential components that perform essential functions in the present invention, but may be optional components for improving performance. Embodiments may be implemented by including only essential components in implementing the essence of the embodiments, excluding components used for performance improvement. A structure including only essential components excluding optional components used for performance improvement is also included in the scope of the embodiments.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily practice the embodiments. In describing the embodiments, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present specification, the detailed description will be omitted. In addition, the same reference numerals are used for the same components in the drawings, and duplicate descriptions of the same components are omitted.

Hereinafter, an image may mean one picture constituting a video, and may also indicate the video itself. For example, "encoding and/or decoding an image" may mean "encoding and/or decoding a video", and may mean "encoding and/or decoding one of images constituting a video". may be

Hereinafter, the terms “video” and “motion picture(s)” may be used interchangeably and may be used interchangeably.

Hereinafter, the target image may be an encoding target image that is an encoding target and/or a decoding target image that is a decoding target. Also, the target image may be an input image input to an encoding device or an input image input to a decoding device. Also, the target image may be a current image that is a target of current encoding and/or decoding. For example, the terms “target image” and “current image” may be used interchangeably and may be used interchangeably.

Hereinafter, the terms “image”, “picture”, “frame” and “screen” may be used interchangeably and interchangeably.

Hereinafter, the target block may be an encoding target block that is an encoding target and/or a decoding target block that is a decoding target. Also, the target block may be a current block that is a target of current encoding and/or decoding. For example, the terms “target block” and “current block” may be used interchangeably and may be used interchangeably. The current block may refer to a coding target block that is an encoding target during encoding and/or a decoding target block that is a decoding target during decoding. Also, the current block may be at least one of a coding block, a prediction block, a residual block, and a transform block.

Hereinafter, the terms “block” and “unit” may be used interchangeably and may be used interchangeably. Or "block" may represent a specific unit.

In the following, the terms "region" and "segment" may be used interchangeably.

In embodiments, each of the specified information, data, flag, index and element, attribute, etc. may have a value. The value "0" of information, data, flags, indexes, elements, and attributes may represent false, logical false, or a first predefined value. That is to say, the value "0", false, logic false and the first predefined value may be used interchangeably. The value "1" of information, data, flags, indexes, elements, and attributes may represent true, logical true, or a second predefined value. In other words, the value "1", true, logically true, and the second predefined value may be used interchangeably.

When a variable such as i or j is used to indicate a row, column, or index, the value of i may be an integer greater than or equal to 0, or may be an integer greater than or equal to 1. That is to say, row, column, index, etc. may be counted from 0, in embodiments, may be counted from 1.

In embodiments, the term “one or more” or the term “at least one” may mean the term “plurality”. “One or more” or “at least one” may be used interchangeably with “plural”.

Below, terms used in the embodiments are explained.

Encoder: An encoder may mean a device that performs encoding. In other words, an encoder may mean an encoding device.

Decoder: A decoder may refer to a device that performs decoding. In other words, a decryptor may mean a decryption device.

Unit: A unit may represent a unit of encoding and/or decoding of an image. The terms "unit" and "block" may be used interchangeably and may be used interchangeably.

- A unit may be an MxN array of samples. M and N may each be a positive integer. A unit may refer to an array of samples in a two-dimensional form.

- In encoding and decoding of an image, a unit may be a region created by dividing one image. In other words, a unit may be a specified area within one image. One image may be divided into a plurality of units. Alternatively, the unit may refer to the divided parts when one image is divided into subdivided parts and encoding or decoding of the divided parts is performed.

- In encoding and decoding of an image, a predefined process for a unit may be performed according to a unit type.

- Depending on the function, the type of unit is a macro unit, a coding unit (CU), a prediction unit (PU), a residual unit, and a transform unit (TU), etc. can be classified as Or, depending on the function, the unit is a block, macroblock, coding tree unit, coding tree block, coding unit, coding block, or prediction unit. It may mean a prediction unit, a prediction block, a residual unit, a residual block, a transform unit, a transform block, and the like. For example, the target unit may be at least one of a CU, a PU, a residual unit, and a TU to be encoded and/or decoded.

- A unit may refer to information including a luma component block, a chroma component block corresponding to the luma component block, and a syntax element for each block, in order to be referred to as a block.

- Units may vary in size and shape. Also, units can have various sizes and shapes. In particular, the shape of the unit may include not only a square but also a two-dimensional geometric figure such as a rectangle, a trapezoid, a triangle, and a pentagon.

- In addition, the unit information may include at least one or more of a unit type, a size of a unit, a depth of a unit, a coding order of a unit, a decoding order of a unit, and the like. For example, the unit type may indicate one of CU, PU, residual unit, and TU.

- One unit can be further divided into sub-units having a smaller size than the unit.

Depth: Depth may mean the degree of division of a unit. Also, the depth of a unit may indicate a level at which the unit exists when the unit(s) are expressed as a tree structure.

- Unit division information may include depth about the depth of the unit. Depth may indicate the number and/or extent to which a unit is divided.

- In the tree structure, the root node has the shallowest depth and the leaf node has the deepest depth. The root node may be the highest node. A leaf node may be the lowest node.

- One unit may be hierarchically divided into a plurality of sub-units while having depth information based on a tree structure. In other words, a unit and a sub-unit generated by division of the unit may correspond to a node and a child node of the node, respectively. Each divided sub-unit may have a depth. Since depth represents the number and/or degree of division of a unit, division information of a sub-unit may include information about the size of the sub-unit.

- In the tree structure, the highest node may correspond to the first non-split unit. The highest node may be referred to as a root node. Also, the highest node may have the smallest depth value. In this case, the highest node may have a depth of level 0.

- A node with a depth of level 1 may represent a unit created as the original unit is split once. A node with a depth of level 2 may represent a unit created as the original unit is split twice.

- A node with a depth of level n may represent a unit created as the initial unit is split n times.

- A leaf node may be the lowest node and may be a node that cannot be further divided. The depth of a leaf node may be the maximum level. For example, the predefined value of the maximum level may be 3.

-QT depth may indicate the depth for quad division. BT depth may indicate a depth for binary partitioning. The TT depth may indicate a depth for ternary division.

Sample: A sample may be a base unit constituting a block. A sample may be represented as values from 0 to 2 ^Bd -1 according to a bit depth (Bd).

- A sample can be a pixel or a pixel value.

- In the following, the terms "pixel", "pixel" and "sample" may be used in the same meaning and may be used interchangeably.

Coding Tree Unit (CTU): A CTU may consist of one luma component (Y) coding tree block and two chroma component (Cb, Cr) coding tree blocks related to the luma component coding tree block. there is. In addition, the CTU may mean including syntax elements for the above blocks and each block of the above blocks.

- Each coding tree unit is a quad tree (QT), binary tree (BT), and ternary tree (TT) to construct sub units such as a coding unit, a prediction unit, and a transform unit. It can be segmented using one or more segmentation schemes. A quad tree may mean a quarternary tree. In addition, each coding tree unit may be split using a MultiType Tree (MTT) using one or more splitting schemes.

- CTU may be used as a term to refer to a pixel block, which is a processing unit in the process of decoding and encoding an image, as in segmentation of an input image.

Coding Tree Block (CTB): A coding tree block may be used as a term to refer to any one of a Y coding tree block, a Cb coding tree block, and a Cr coding tree block.

Neighbor Block: A neighboring block may mean a block adjacent to a target block. A neighboring block may mean a reconstructed neighboring block.

- In the following, the terms "neighboring block" and "adjacent block" may be used in the same meaning and may be used interchangeably.

- A neighboring block may mean a reconstructed neighboring block.

Spatial neighbor block: A spatial neighbor block may be a block that is spatially adjacent to the target block. Neighboring blocks may include spatial neighboring blocks.

- The target block and spatial neighboring blocks may be included in the target picture.

- A spatial neighboring block may mean a block whose boundary meets the target block or a block located within a predetermined distance from the target block.

- A spatial neighboring block may mean a block adjacent to a vertex of a target block. Here, the block adjacent to the vertex 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.

Temporal neighbor block: A temporal neighbor block may be a block that is temporally adjacent to the target block. Neighboring blocks may include temporal neighboring blocks.

- A temporal neighboring block may include a co-located block (col block).

- A collocated block may be a block in an already reconstructed co-located picture (col picture). A position of a collocated block in a collocated picture may correspond to a position of a target block in a target picture. Alternatively, the position of the collocated block in the collocated picture may be the same as the position of the target block in the target picture. A collocated picture may be a picture included in a reference picture list.

- A temporal neighboring block may be a block temporally adjacent to a spatial neighboring block of a target block.

Prediction mode: The prediction mode may be information indicating a mode used for intra prediction or a mode used for inter prediction.

Prediction unit: A prediction unit may mean a base unit for prediction such as inter prediction, intra prediction, inter compensation, intra compensation, and motion compensation.

- One prediction unit may be divided into a plurality of smaller-sized partitions or sub-prediction units. A plurality of partitions may also be a basis unit in performing prediction or compensation. A partition generated by division of a prediction unit may also be a prediction unit.

Prediction unit partition (prediction unit partition): A prediction unit partition may mean a form in which a prediction unit is divided.

Reconstructed neighboring unit: A reconstructed neighboring unit may be a unit that has already been decoded and reconstructed in a neighbor of a target unit.

- The reconstructed neighbor unit may be a spatial neighbor unit or a temporal neighbor unit to the target unit.

- The reconstructed spatial neighboring unit may be a unit in the target picture and already reconstructed through encoding and/or decoding.

- The reconstructed temporal neighbor unit may be a unit in the reference picture and a unit that has already been reconstructed through encoding and/or decoding. A position of the reconstructed temporal neighboring unit within the reference image may be the same as a position within the target unit's target picture or may correspond to a position within the target unit's target picture. Alternatively, the reconstructed temporal neighboring unit may be a neighboring block of a corresponding block in the reference picture. 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 correspondence of the positions of the blocks may mean that the positions of the blocks are the same, and may mean that one block is included in another block, and one block occupies a specified position of the other block. can mean doing

Sub-picture: A picture can be divided into one or more sub-pictures. A sub-picture may consist of one or more tile rows and one or more tile columns.

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

- A sub-picture may be a rectangular area of one or more slices within one picture.

- One sub-picture may include one or more tiles, one or more bricks, and/or one or more slices.

Tile: A tile can be a square or rectangular (ie, non-square) area within a picture.

- A tile may contain one or more CTUs.

- A tile can be divided into one or more bricks.

Brick: A brick may mean one or more CTU rows within a tile.

- A tile can be divided into one or more bricks. Each brick may contain one or more CTU rows.

- A tile that is not divided into two or more can also mean a brick.

Slice: A slice can include one or more tiles within a picture. Or, a slice may include one or more bricks within a tile.

- A sub-picture may contain one or more slices that collectively cover a rectangular area within the picture. Accordingly, each sub-picture boundary may always be a slice boundary. Also, each vertical sub-picture boundary may always be a vertical tile boundary.

Parameter set: A parameter set may correspond to header information among structures in a bitstream.

- Parameter sets include a Video Parameter Set (VPS), a Sequence Parameter Set (SPS), a Picture Parameter Set (PPS), an Adaptation Parameter Set (APS), and a decoding parameter It may include at least one of a set (Decoding Parameter Set; DPS), and the like.

Information signaled through a parameter set may be applied to pictures referring to the parameter set. For example, information in the VPS may be applied to pictures referring to the VPS. Information in the SPS may be applied to pictures referring to the SPS. Information in the PPS may be applied to pictures referring to the PPS.

A parameter set may refer to an upper parameter set. For example, PPS may refer to SPS. SPS may refer to VPS.

- In addition, the parameter set may include tile group, slice header information, and tile header information. A tile group may refer to a group including a plurality of tiles. Also, the meaning of a tile group may be the same as that of a slice.

Rate-distortion optimization: The encoding device uses a combination of the size of the coding unit, the prediction mode, the size of the prediction unit, motion information, and the size of the conversion unit to provide high encoding efficiency. Distortion optimization can be used.

- The rate-distortion optimization method may calculate a rate-distortion cost of each combination in order to select an optimal combination among the above combinations. The rate-distortion cost can be calculated using the formula “D+λ*R”. In general, a combination that minimizes the rate-distortion cost according to the equation “D+λ*R” can be selected as an optimal combination in the rate-distortion optimization method.

- D can represent distortion. D may be the mean square error of difference values between the original transform coefficients and the reconstructed transform coefficients within the transform unit.

- R can represent a rate. R may represent a bit rate using related context information.

- λ may represent a Lagrangian multiplier. R may include not only coding parameter information such as a prediction mode, motion information, and a coded block flag, but also bits generated by encoding transform coefficients.

- The encoding device may perform processes such as inter prediction, intra prediction, transformation, quantization, entropy encoding, inverse quantization, and/or inverse transformation to calculate accurate D and R. These processes can greatly increase complexity in an encoding device.

Bitstream: A bitstream may mean a string of bits including coded image information.

Parsing: Parsing may mean determining a value of a syntax element by entropy decoding a bitstream. Alternatively, parsing may mean entropy decoding itself.

Symbol: may mean 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. Also, a symbol may mean an object of entropy encoding or a result of entropy decoding.

Reference picture: A reference picture may refer to an image that a unit refers to for inter prediction or motion compensation. Alternatively, the reference picture may be an image including a reference unit referred to by a target unit for inter prediction or motion compensation.

Hereinafter, the terms “reference picture” and “reference image” may be used interchangeably and may be used interchangeably.

Reference picture list: The reference picture list may be a list including one or more reference pictures used for inter prediction or motion compensation.

- The types of reference picture lists are List Combined (LC), List 0 (L0), List 1 (L1), List 2 (List 2; L2), and List 3 (List 3; L3). ), etc. may be present.

- One or more reference picture lists may be used for inter prediction.

Inter prediction indicator: The inter prediction indicator may indicate the direction of inter prediction for a target unit. Inter prediction can be one of uni-prediction and bi-prediction, etc. Alternatively, the inter prediction indicator may indicate the number of reference pictures used when generating a prediction unit of a target unit. Alternatively, the inter prediction indicator may indicate the number of prediction blocks used for inter prediction or motion compensation of the target unit.

Prediction list utilization flag: The prediction list utilization flag may indicate whether a prediction unit is generated using at least one reference picture in a specific reference picture list.

- An inter prediction indicator can be derived using the prediction list utilization flag. Conversely, the prediction list utilization flag can be derived using the inter prediction indicator. For example, when the prediction list utilization flag indicates a first value of 0, it may indicate that a prediction block is not generated using a reference picture in the reference picture list for the target unit. When the prediction list utilization flag indicates the second value of 1, it may indicate that a prediction unit is generated using the reference picture list for the target unit.

Reference picture index: The reference picture index may be an index indicating a specific reference picture in a reference picture list.

Picture Order Count (POC): The POC of a picture may indicate a display order of pictures.

Motion Vector (MV): A motion vector may be a two-dimensional vector used in inter prediction or motion compensation. A motion vector may mean an offset between a target image and a reference image.

- For example, MV can be expressed as (mv _x , mv _y ). mv _x may represent a horizontal component, and mv _y may represent a vertical component.

Search range: The search range may be a two-dimensional area where MVs are searched during inter prediction. For example, the size of the search area may be MxN. M and N may each be a positive integer.

Motion vector candidate: A motion vector candidate may mean a block as a prediction candidate or a motion vector of a block as a prediction candidate when predicting a motion vector.

- A motion vector candidate may be included in a motion vector candidate list.

Motion vector candidate list: The motion vector candidate list may refer to a list constructed using one or more motion vector candidates.

Motion vector candidate index: The motion vector candidate index may mean an indicator 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.

Motion information: Motion information includes not only motion vectors, reference picture indices and inter prediction indicators, but also reference picture list information, reference pictures, motion vector candidates, motion vector candidate indices, merge candidates and merge indices, etc. It may mean information including at least one of

Merge candidate list: A merge candidate list may refer to a list constructed using one or more merge candidates.

Merge candidate: A merge candidate is a spatial merge candidate, a temporal merge candidate, a combined merge candidate, a combined bi-prediction merge candidate, a candidate based on history, a candidate based on an average of two candidates, and zero It may mean a merge candidate and the like. A merge candidate may include an inter prediction indicator, and may include motion information such as a reference picture index for each list, a motion vector, a prediction list utilization flag, and an inter prediction indicator.

Merge index: A merge index may be an indicator pointing to a merge candidate in a merge candidate list.

- The merge index may indicate a reconstructed unit that derives a merge candidate from among reconstructed units spatially adjacent to the target unit and reconstructed units temporally adjacent to the target unit.

- The merge index may indicate at least one piece of motion information of a merge candidate.

Transform unit: A transform unit may be a basic unit in residual signal encoding and/or residual signal decoding, such as transform, inverse transform, quantization, inverse quantization, transform coefficient encoding, and transform coefficient decoding. One transform unit may be divided into a plurality of sub-transform units having smaller sizes. Here, the transformation may include one or more of a first-order transformation and a second-order transformation, and the inverse transformation may include one or more of a first-order inverse transformation and a second-order inverse transformation.

Scaling: Scaling may refer to a process of multiplying a transform coefficient level by a factor.

- As a result of the scaling to the transform coefficient level, a transform coefficient can be generated. Scaling may be referred to as dequantization.

Quantization Parameter (QP): A quantization parameter may mean a value used when generating a transform coefficient level for a transform coefficient in quantization. Alternatively, the quantization parameter may refer to a value used when generating a transform coefficient by scaling a transform coefficient level in inverse quantization. Alternatively, the quantization parameter may be a value mapped to a quantization step size.

Delta quantization parameter: The delta quantization parameter may refer to a difference value between a predicted quantization parameter and a quantization parameter of a target unit.

Scan: A scan may refer to a method of arranging the order of coefficients within a unit, block or matrix. For example, arranging a 2D array into a 1D array may be referred to as a scan. Alternatively, arranging a one-dimensional array into a two-dimensional array may also be referred to as scan or inverse scan.

Transform coefficient: A transform coefficient may be a coefficient value generated by performing transformation in an encoding device. Alternatively, the transform coefficient may be a coefficient value generated by performing at least one of entropy decoding and inverse quantization in the decoding apparatus.

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

Quantized level: A quantized level may refer to a value generated by performing quantization on a transform coefficient or a residual signal in an encoding device. Alternatively, the quantized level may mean a value to be subjected to inverse quantization when the decoding apparatus performs inverse quantization.

- A quantized transform coefficient level, which is a result of transform and quantization, may also be included in the meaning of the quantized level.

Non-zero transform coefficient: A non-zero transform coefficient may mean a transform coefficient having a non-zero value or a transform coefficient level having a non-zero value. Alternatively, the non-zero transform coefficient may refer to a transform coefficient whose value is not 0 or a transform coefficient level whose value is not 0.

Quantization matrix: A quantization matrix may mean a matrix used in a quantization process or an inverse quantization process to improve subjective or objective picture quality of an image. A quantization matrix may also be referred to as a scaling list.

Quantization matrix coefficient: A quantization matrix coefficient may mean each element in a quantization matrix. Quantization matrix coefficients may also be referred to as matrix coefficients.

Default matrix: The default matrix may be a quantization matrix predefined in an encoding device and a decoding device.

Non-default matrix: The non-default matrix may be a quantization matrix that is not predefined in the encoding device and the decoding device. The non-default matrix may refer to a quantization matrix signaled from an encoding device to a decoding device by a user.

Most Probable Mode (MPM): MPM may indicate an intra prediction mode that is highly likely to be used for intra prediction of a target block.

- The encoding device and the decoding device may determine one or more MPMs based on a coding parameter related to the target block and an attribute of an object related to the target block.

- The encoding device and the decoding device may determine one or more MPMs based on the intra prediction mode of the reference block. Reference blocks may be plural. The plurality of reference blocks may include a spatial neighboring block adjacent to the left side of the target block and a spatial neighboring block adjacent to the top of the target block. In other words, one or more different MPMs may be determined depending on which intra prediction modes are used for reference blocks.

- One or more MPMs may be determined in the same way in the encoding device and the decoding device. In other words, the encoding device and the decoding device may share an MPM list including one or more identical MPMs.

MPM List: An MPM list can be a list containing one or more MPMs. The number of one or more MPMs in the MPM list may be predefined.

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

- Since the MPM list is determined in the same way in the encoding device and the decoding device, the MPM list itself may not need to be transmitted from the encoding device to the decoding device.

- The MPM indicator may be signaled from the encoding device to the decoding device. As the MPM indicator is signaled, the decoding apparatus may determine an MPM to be used for intra prediction of the target block among MPMs in the MPM list.

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

- The MPM use indicator may be signaled from the encoding device to the decoding device.

Signaling: Signaling may indicate that information is transmitted from an encoding device to a decoding device. Alternatively, signaling may mean that an encoding device includes information in a bitstream or a recording medium. Information signaled by the encoding device may be used by the decoding device.

- The encoding device may generate encoded information by performing encoding on signaled information. Encoded information may be transmitted from an encoding device to a decoding device. The decoding apparatus may obtain information by decoding the transmitted encoded information. Here, encoding may be entropy encoding, and decoding may be entropy decoding.

Optional Signaling: Information can be selectively signaled. Selective signaling of information may mean that an encoding device selectively includes information in a bitstream or a recording medium (according to specific conditions). Selective signaling of information may mean that a decoding apparatus selectively extracts information from a bitstream (according to a specific condition).

Omission of signaling: Signaling of information may be omitted. Omission of signaling of information about information may mean that an encoding device does not include information in a bitstream or a recording medium (according to a specific condition). Omission of signaling for information may mean that the decoding apparatus does not extract information from the bitstream (according to a specific condition).

Statistical value: Variables, coding parameters and constants, etc. can have values that can be computed. A statistical value may be a value generated by an operation on the values of these specified objects. For example, the statistical value is an average value, a weighted average value, a weighted sum, a minimum value, a maximum value, and a mode for values such as a specified variable, a specified coding parameter, and a specified constant. It can be one or more of a value, a median value, and an interpolated value.

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

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. The encoding apparatus 100 may sequentially encode one or more images of a video.

Referring to FIG. 1, an encoding apparatus 100 includes an inter prediction unit 110, an intra prediction unit 120, a switch 115, a subtractor 125, a transform unit 130, a quantization unit 140, and entropy encoding. It may include a 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.

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 for the target block may be one of an intra mode and an inter mode.

Hereinafter, the terms “intra mode”, “intra prediction mode”, “in-picture mode” and “in-picture prediction mode” may be used interchangeably and interchangeably.

Hereinafter, the terms “inter mode”, “inter prediction mode”, “inter-screen mode” and “inter-prediction mode” may be used in the same meaning and may be used interchangeably.

Hereinafter, the term “video” may refer to only a part of an image or may refer to a block. Also, processing of “image” may indicate sequential processing of a plurality of blocks.

Also, the encoding device 100 may generate a bitstream including encoded information through encoding of a target image, and output and store the generated bitstream. The generated bitstream may be stored in a computer readable recording medium and may be streamed through a wired and/or wireless transmission medium.

As the prediction mode, if the intra mode is used, the switch 115 can be switched to intra. As the prediction mode, when the inter mode is used, the switch 115 can be switched to inter.

The encoding apparatus 100 may generate a prediction block for the target block. Also, after the prediction block is generated, the encoding apparatus 100 may encode a residual block for the target block by using the target block and the residual of the prediction block.

When the prediction mode is the intra mode, the intra predictor 120 may use a pixel of a block already encoded and/or decoded, which is adjacent to the target block, as a reference sample. The intra predictor 120 may perform spatial prediction on the target block using the reference sample, and generate prediction samples for the target block through the spatial prediction. A prediction sample may mean a sample within a prediction block.

The inter prediction unit 110 may include a motion estimation unit and a motion compensation unit.

When the prediction mode is the inter mode, the motion prediction unit may search for a region that best matches the target block from the reference image in the motion prediction process, and derive motion vectors for the target block and the searched region using the searched region. can do. At this time, the motion prediction unit may use a search area as a search target area.

The reference picture may be stored in the reference picture buffer 190, and when encoding and/or decoding of the reference picture is processed, the encoded and/or decoded reference picture may be stored in the reference picture buffer 190.

As a decoded picture is stored, the reference picture buffer 190 may be a decoded picture buffer (DPB).

The motion compensator may generate a prediction block for the target block by performing motion compensation using a motion vector. Here, the motion vector may be a 2D vector used for inter prediction. Also, the motion vector may indicate an offset between the target image and the reference image.

The motion estimation unit and the motion compensation unit may generate a prediction block by applying an interpolation filter to a partial region in a reference image when a motion vector has a non-integer value. In order to perform inter prediction or motion compensation, methods of motion prediction and motion compensation of PUs included in the CU based on the CU include skip mode, merge mode, and advanced motion vector prediction (Advanced Motion Vector Prediction (AMVP) mode and current picture reference mode may be determined, and inter prediction or motion compensation may be performed according to each mode.

The subtractor 125 may generate a residual block that is a difference between the target block and the prediction block. A residual block may also be referred to as a residual signal.

The residual signal may mean a difference between the original signal and the predicted signal. Alternatively, the residual signal may be a signal generated by transforming, quantizing, or transforming and quantizing the difference between the original signal and the prediction signal. A residual block may be a residual signal for a block unit.

The transform unit 130 may generate transform coefficients by performing transform on the residual block, and may output the generated transform coefficients. Here, the transform coefficient may be a coefficient value generated by performing transform on the residual block.

The conversion unit 130 may use one of a plurality of predefined conversion methods in performing the conversion.

A plurality of predefined transform methods may include discrete cosine transform (DCT), discrete sine transform (DST), and Karhunen-Loeve transform (KLT) based transform. there is.

A transform method used for transforming the residual block may be determined according to at least one of coding parameters for the target block and/or neighboring blocks. For example, the transform method may be determined based on at least one of an inter prediction mode for a PU, an intra prediction mode for a PU, a TU size, and a TU shape. Alternatively, transformation information indicating a transformation method may be signaled from the encoding apparatus 100 to the decoding apparatus 200.

When a transform skip mode is applied, the transform unit 130 may skip transforming the residual block.

A quantized transform coefficient level or a quantized level may be generated by applying quantization to the transform coefficients. Hereinafter, in the embodiments, a quantized transform coefficient level and a quantized level may also be referred to as a transform coefficient.

The quantization unit 140 may generate a quantized transform coefficient level (ie, a quantized level or a quantized coefficient) by quantizing a transform coefficient according to a quantization parameter. The quantization unit 140 may output the generated quantized transform coefficient level. At this time, the quantization unit 140 may quantize the transform coefficient using a quantization matrix.

The entropy encoding unit 150 may generate a bitstream by performing entropy encoding according to a probability distribution based on values calculated by the quantization unit 140 and/or coding parameter values calculated in the encoding process. . The entropy encoding unit 150 may output the generated bitstream.

The entropy encoding unit 150 may perform entropy encoding on information about pixels of an image and information for decoding an image. For example, information for decoding an image may include a syntax element and the like.

When entropy encoding is applied, a small number of bits may be allocated to a symbol having a high probability of occurrence, and a large number of bits may be allocated to a symbol having a low probability of occurrence. As symbols are represented through such allocation, the size of bitstrings for symbols that are encoding targets can be reduced. Therefore, compression performance of image encoding can be improved through entropy encoding.

In addition, the entropy encoding unit 150 uses exponential golomb, context-adaptive variable length coding (CAVLC), and context-adaptive binary arithmetic coding (Context-Adaptive Binary Coding) for entropy encoding. A coding method such as Arithmetic Coding (CABAC) may be used. For example, the entropy encoding unit 150 may perform entropy encoding using a variable length coding/code (VLC) table. For example, the entropy encoding unit 150 may derive a binarization method for a target symbol. Also, the entropy encoding unit 150 may derive a probability model of the target symbol/bin. The entropy encoding unit 150 may perform arithmetic encoding using the derived binarization method, probability model, and context model.

The entropy encoding unit 150 may change coefficients in the form of a 2-dimensional block into a form of a 1-dimensional vector through a transform coefficient scanning method in order to encode the quantized transform coefficient level.

A coding parameter may be information required for encoding and/or decoding. The coding parameter may include information that is encoded in the encoding device 100 and transmitted from the encoding device 100 to the decoding device, and may include information that can be derived in an encoding or decoding process. For example, as information transmitted to the decoding device, there is a syntax element.

A coding parameter may include information derived from an encoding process or a decoding process, as well as information (or flags and indexes, etc.) encoded in an encoding device and signaled from an encoding device to a decoding device, such as a syntax element. there is. Also, the coding parameter may include information required for encoding or decoding an image. For example, the size of the unit/block, the shape of the unit/block, the depth of the unit/block, division information of the unit/block, division structure of the unit/block, information indicating whether the unit/block is divided in the form of a quad tree, Information indicating whether the unit/block is split into a binary tree, the split direction of the binary tree (horizontal or vertical), the split type of the binary tree (symmetric or asymmetric), the unit/block is a ternary tree information indicating whether it is split into ternary tree type split direction (horizontal direction or vertical direction), ternary tree type split type (symmetric split or asymmetric split, etc.), unit/block multi-type tree information indicating whether or not to split into a multi-type tree, combination and direction (horizontal or vertical, etc.) of multi-type tree split, split type of multi-type tree split (symmetric split or asymmetric split), multi-type Split tree in tree form (binary tree or ternary tree), type of prediction mode (intra prediction or inter prediction), intra prediction mode/direction, intra luma prediction mode/direction, intra chroma prediction mode/direction, intra partitioning information, inter prediction Split information, coding block split flag, prediction block split flag, transform block split flag, reference sample filtering method, reference sample filter tap (tap), reference sample filter coefficient, prediction block filtering method, prediction block filter tap, prediction block filter coefficient , prediction block boundary filtering method, prediction block boundary filter tap, prediction block boundary filter coefficient, inter prediction mode, motion information, motion vector, motion vector difference, reference picture index, inter prediction direction, inter prediction indicator, prediction list utilization ) flag, reference picture list, reference picture, 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, interpolation filter type, interpolation filter filter tap, interpolation filter filter coefficient, motion vector magnitude, motion vector representation accuracy, transform type, transform size, and primary transform. Information indicating whether or not to use, information indicating whether to use additional (secondary) transform, primary transform selection information (or primary transform index), secondary transform selection information (or secondary transform index), residual Information indicating the presence or absence of a signal, a coded block pattern, a coded block flag, a quantization parameter, a residual quantization parameter, a quantization matrix, information about an intra-loop filter, and an intra-loop filter Information indicating whether to apply, coefficients of intra-loop filter, filter tap of intra-loop, shape/form of intra-loop filter, information indicating whether to apply deblocking filter, information indicating whether to apply deblocking filter, coefficient, filter tap of deblocking filter, strength of deblocking filter, shape/shape of deblocking filter, information indicating whether to apply adaptive sample offset, adaptive sample offset value, adaptive sample offset category, adaptive sample Offset type, information indicating whether the adaptive in-loop filter is applied, coefficients of the adaptive in-loop filter, filter tap of the adaptive in-loop filter, shape/shape of the adaptive in-loop filter Shape, binarization/inverse binarization method, context model, context model determination method, context model update method, information indicating whether regular mode is performed, information indicating whether bypass mode is performed, significant coefficient flags, last Significant coefficient flag, coefficient group unit coding flag, last significant coefficient position, flag indicating whether coefficient value is greater than 1, flag indicating whether coefficient value is greater than 2, flag indicating whether coefficient value is greater than 3 , remaining coefficient value information, sign information, reconstructed luma sample, reconstructed chroma sample, context bin, bypass bin, residual luma sample, residual chroma sample, transform coefficient, luma transform coefficient, chroma transform coefficient, Quantized level, luma quantized level, chroma quantized level, transform coefficient level, luma transform coefficient level, chroma transform coefficient level, transform coefficient level scanning method, size of motion vector search region from the side of a decoding device, decoding device The shape of the motion vector search area on the side, the number of motion vector searches on the side of the decoding device, CTU size, minimum block size, maximum block size, maximum block depth, minimum block depth, image display/output order, slice identification information , slice type, slice division information, tile group identification information, tile group type, tile group division information, tile identification information, tile type, tile division information, picture type, bit depth, input sample bit depth, reconstructed sample bit depth , residual sample bit depth, transform coefficient bit depth, quantized level bit depth, luma signal information, chroma signal information, a color space of a target block, and a value of at least one of a color space of a residual block, Combined forms or statistics can be included in the coding parameters. In addition, information related to the aforementioned coding parameters may also be included in the coding parameters. Information used to calculate and/or derive the aforementioned coding parameters may also be included in the coding parameters. Information calculated or derived using the aforementioned coding parameters may also be included in the coding parameters.

The primary transform selection information may indicate a primary transform applied to the target block.

The secondary transform selection information may indicate a secondary transform applied to the target block.

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

Here, signaling information may mean that the encoding device 100 includes entropy-encoded information generated by performing entropy encoding on a flag or index in a bitstream, , in the decoding apparatus 200, may mean obtaining information by performing entropy decoding on entropy-encoded information extracted from a bitstream. Here, the information may include flags and indexes.

A signal may refer to signaled information. Hereinafter, information on images and blocks may be referred to as signals. Also, in the following, the terms “information” and “signal” may be used interchangeably and may be used interchangeably. For example, the specific signal may be a signal representing a specific block. An original signal may be a signal representing a target block. A prediction signal may be a signal representing a prediction block. A residual signal may be a signal representing a residual block.

A bitstream may include information according to a specified syntax. The encoding device 100 may generate a bitstream including information according to a specified syntax. The encoding device 200 may obtain information from a bitstream according to a specified syntax.

Since encoding through inter prediction is performed by the encoding apparatus 100, the encoded target image may be used as a reference image for other image(s) to be processed later. Accordingly, the encoding apparatus 100 may reconstruct or decode the encoded target image and store the reconstructed or decoded image in the reference picture buffer 190 as a reference image. Inverse quantization and inverse transformation may be performed on the encoded target image for decoding.

The quantized level may be inversely quantized in the inverse quantization unit 160 and inversely transformed in the inverse transformation unit 170 . The inverse quantization unit 160 may generate inverse quantized coefficients by performing inverse quantization on the quantized level. The inverse transform unit 170 may generate inverse quantized and inverse transformed coefficients by performing an inverse transform on the inverse quantized coefficients.

The inverse quantized and inverse transformed coefficients may be combined with the prediction block through the adder 175. A reconstructed block may be generated by adding the inverse quantized and inverse transformed coefficients and the prediction block. Here, the inverse quantized and/or inverse transformed coefficient may mean a coefficient on which at least one of dequantization and inverse-transformation has been performed, and may mean a reconstructed residual block. Here, the reconstructed block may mean a recovered block or a decoded block.

The reconstructed block may pass through the filter unit 180 . The filter unit 180 may include at least one of a deblocking filter, a sample adaptive offset (SAO), an adaptive loop filter (ALF), and a non-local filter (NLF). One or more may be applied to a reconstructed sample, reconstructed block or reconstructed picture. The filter unit 180 may also be referred to as an in-loop filter.

The deblocking filter may remove block distortion generated at a boundary between blocks in a reconstructed picture. In order to determine whether to apply the deblocking filter, it may be determined whether to apply the deblocking filter to the target block based on pixel(s) included in several columns or rows included in the block.

When a deblocking filter is applied to a target block, the applied filter may vary according to the required strength of deblocking filtering. In other words, among different filters, a filter determined according to the strength of deblocking filtering may be applied to the target block. When a deblocking filter is applied to the target block, a long-tap filter, a strong filter, a weak filter, and a Gaussian filter are selected according to the strength of the deblocking filtering required. ) may be applied to the target block.

Also, when vertical filtering and horizontal filtering are performed on a target block, horizontal filtering and vertical filtering may be processed in parallel.

SAO may add an appropriate offset to a pixel value of a pixel to compensate for a coding error. The SAO may perform correction using an offset on a difference between an original image and an image to which deblocking is applied in units of pixels for an image to which deblocking is applied. In order to perform offset correction on an image, a method of dividing pixels included in the image into a certain number of areas, determining an area to be offset from among the divided areas, and applying the offset to the determined area will be used. Alternatively, a method of applying an offset in consideration of edge information of each pixel of the image may be used.

ALF may perform filtering based on a value obtained by comparing the reconstructed image and the original image. After dividing the pixels included in the image into predetermined groups, a filter to be applied to each divided group may be determined, and filtering may be performed differentially for each group. Information related to whether to apply the adaptive loop filter may be signaled for each CU. This information may be signaled for the luma signal. The shape and filter coefficients of ALF to be applied to each block may be different for each block. Alternatively, a fixed type of ALF may be applied to the block regardless of the characteristics of the block.

The non-local filter may perform filtering based on reconstructed blocks similar to the target block. A region similar to the target block may be selected from the reconstructed image, and filtering of the target block may be performed using statistical properties of the selected similar region. Information related to whether to apply the non-local filter may be signaled to the CU. Also, shapes and filter coefficients of non-local filters to be applied to blocks may be different according to blocks.

A reconstructed block or a reconstructed image that has passed through the filter unit 180 may be stored in the reference picture buffer 190 as a reference picture. A reconstructed block that has passed through the filter unit 180 may be part of a reference picture. In other words, the reference picture may be a reconstructed picture composed of reconstructed blocks that have passed through the filter unit 180. The stored reference picture can then be used for inter prediction or motion compensation.

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

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

Referring to FIG. 2 , the decoding apparatus 200 includes 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, and a switch 245. , an adder 255, a filter unit 260, and a reference picture buffer 270.

The decoding device 200 may receive the bitstream output from the encoding device 100. The decoding apparatus 200 may receive a bitstream stored in a computer readable recording medium or may receive a bitstream streamed through a wired/wireless transmission medium.

The decoding apparatus 200 may perform intra mode and/or inter mode decoding on a bitstream. Also, the decoding apparatus 200 may generate a reconstructed image or a decoded image through decoding, and output the generated reconstructed image or decoded image.

For example, conversion to an intra mode or an inter mode according to a prediction mode used for decoding may be performed by the switch 245 . When the prediction mode used for decoding is an intra mode, the switch 245 may be switched to intra mode. When the prediction mode used for decoding is the inter mode, the switch 245 may be switched to inter mode.

The decoding apparatus 200 may obtain a reconstructed residual block by decoding the input bitstream and generate a prediction block. When the reconstructed residual block and the prediction block are obtained, the decoding apparatus 200 may generate a reconstructed block to be decoded by summing the reconstructed residual block and the prediction block.

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 (ie, quantized levels or quantized coefficients). Here, the entropy decoding method may be similar to the above-described entropy encoding method. For example, the entropy decoding method may be a reverse process of the above-described entropy encoding method.

The entropy decoding unit 210 may change a coefficient in the form of a 1-dimensional vector into a form of a 2-dimensional block through a transform coefficient scanning method in order to decode the quantized transform coefficient level.

For example, the coefficients may be changed into a 2D block form by scanning the coefficients of a block using an upper-right diagonal scan. Alternatively, which scan to be used among the upper-right diagonal scan, vertical scan, and horizontal scan may be determined according to the size of the block and/or the intra prediction mode.

The quantized coefficient may be inversely quantized in the inverse quantization unit 220 . The inverse quantization unit 220 may generate inverse quantized coefficients by performing inverse quantization on the quantized coefficients. In addition, the inverse quantized coefficient may be inversely transformed in the inverse transform unit 230 . The inverse transform unit 230 may generate a reconstructed residual block by performing an inverse transform on the inverse quantized coefficient. As a result of performing inverse quantization and inverse transformation on the quantized coefficients, a reconstructed residual block may be generated. In this case, the inverse quantization unit 220 may apply a quantization matrix to the quantized coefficients in generating the reconstructed residual block.

When the intra mode is used, the intra predictor 240 may generate a prediction block by performing spatial prediction on the target block using pixel values of previously decoded blocks adjacent to the target block.

The inter prediction unit 250 may include a motion compensation unit. Alternatively, the inter prediction unit 250 may be referred to as a motion compensation unit.

When the inter mode is used, the motion compensator 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 .

When the motion vector has a non-integer value, the motion compensator may apply an interpolation filter to a partial region in the reference image and generate a prediction block using the reference image to which the interpolation filter is applied. The motion compensation unit may determine which mode among skip mode, merge mode, AMVP mode, and current picture reference mode is a motion compensation method used for a PU included in a CU based on the CU to perform motion compensation, and the determined mode Accordingly, motion compensation may be performed.

The reconstructed residual block and the prediction block may be added through an adder 255. Adder 255 may produce a reconstructed block by adding the reconstructed residual block and the prediction block.

The reconstructed block may pass through the filter unit 260 . The filter unit 260 may apply at least one of a deblocking filter, an SAO filter, an ALF filter, and a non-local filter to a reconstructed block or a reconstructed image. A reconstructed image may be a picture including a reconstructed block.

The filter unit 260 may output a reconstructed image.

The reconstructed block and/or the reconstructed image that has passed through the filter unit 260 may be stored as a reference picture in the reference picture buffer 270 . A reconstructed block that has passed through the filter unit 260 may be part of a reference picture. In other words, the reference picture may be a reconstructed image composed of reconstructed blocks that have passed through the filter unit 260 . The stored reference picture can then be used for inter prediction and/or motion compensation.

3 is a diagram schematically illustrating a division structure of an image when encoding and decoding an image.

3 may schematically show an example in which one unit is divided into a plurality of sub-units.

In order to efficiently segment an image, a coding unit (CU) may be used in encoding and decoding. A unit may be a term that collectively refers to 1) a block including image samples and 2) a syntax element. For example, “division of a unit” may mean “division of a block corresponding to a unit”.

A CU may be used as a base unit for image encoding and/or decoding. In addition, the CU may be used as a unit to which one selected mode of intra mode and inter mode is applied in video encoding and/or decoding. In other words, in video encoding and/or decoding, it may be determined which mode among the intra mode and the inter mode is to be applied to each CU.

Also, a CU may be a base unit in encoding and/or decoding of prediction, transform, quantization, inverse transform, inverse quantization, and transform coefficients.

Referring to FIG. 3 , an image 300 may be sequentially divided into units of largest coding units (LCUs). For each LCU, a partition structure may be determined. Here, LCU may be used as the same meaning as Coding Tree Unit (CTU).

Division of a unit may mean division of a block corresponding to the unit. The block division information may include depth information about the depth of a unit. Depth information may indicate the number and/or degree of division of a unit. One unit may be hierarchically divided into a plurality of sub-units with depth information based on a tree structure.

Each divided sub-unit may have depth information. Depth information may be information indicating the size of a CU. Depth information may be stored for each CU.

Each CU may have depth information. When a CU is split, CUs generated by splitting may have a depth increased by 1 from the depth of the split CU.

The division structure may refer to a distribution of CUs in the LCU 310 to efficiently encode an image. This distribution may be determined according to whether one CU is to be divided into a plurality of CUs. The number of divided CUs may be a positive integer greater than or equal to 2, including 2, 4, 8 and 16, and the like.

The horizontal and vertical sizes of the CU generated by division may be smaller than the horizontal and vertical sizes of the CU before division, depending on the number of CUs generated by division. For example, the horizontal size and vertical size of the CU generated by division may be half of the horizontal size and half of the vertical size of the CU before division.

A divided CU may be recursively divided into a plurality of CUs in the same way. By recursive division, at least one of the horizontal size and the vertical size of the divided CU may be reduced compared to at least one of the horizontal and vertical sizes of the CU before division.

The division of the CU may be made recursively to a predefined depth or to a predefined size.

For example, the depth of CU may have a value of 0 to 3. The size of the CU may range from 64x64 to 8x8 depending on the depth of the CU.

For example, the depth of the LCU 310 may be 0, and the depth of the smallest coding unit (SCU) may be a predefined maximum depth. Here, the LCU may be a CU having the largest coding unit size as described above, and the SCU may be a CU having the smallest coding unit size.

The division may start from the LCU 310, and the depth of the CU may increase by 1 whenever the horizontal size and/or the vertical size of the CU are reduced by the division.

For example, for each depth, a CU that is not split may have a size of 2Nx2N. In addition, in the case of a CU to be divided, a CU of 2Nx2N size may be divided into 4 CUs having a size of NxN. The size of N can be halved every time the depth increases by 1.

Referring to FIG. 3 , an LCU having a depth of 0 may be 64x64 pixels or a 64x64 block. 0 may be the minimum depth. A SCU with a depth of 3 can be 8x8 pixels or an 8x8 block. 3 may be the maximum depth. In this case, the CU of the 64x64 block, which is the LCU, may be expressed as a depth of 0. A CU of a 32x32 block can be represented with a depth of 1. A CU of a 16x16 block can be represented with a depth of 2. A CU of an 8x8 block, which is an SCU, can be expressed as a depth of 3.

Information on whether the CU is split may be expressed through split information of the CU. The division information may be 1 bit of information. All CUs except for the SCU may include partition information. For example, a value of partition information of a CU that is not split may be a first value, and a value of partition information of a CU that is split may be a second value. When the split information indicates whether the CU splits, the first value may be 0 and the second value may be 1.

For example, when one CU is divided into 4 CUs, the horizontal and vertical sizes of each CU of the 4 CUs generated by the division are half of the horizontal size and half of the vertical size of the CU before the division, respectively. can When a 32x32 CU is divided into 4 CUs, the sizes of the 4 CUs may be 16x16. When one CU is divided into 4 CUs, it can be said that the CU is divided in a quad-tree form. In other words, it can be seen that quad-tree partitioning is applied to the CU.

For example, when one CU is split into two CUs, the horizontal size or vertical size of each CU of the two CUs generated by the split is half of the horizontal size or half of the vertical size of the CU before the split, respectively. can When a CU of 32x32 size is vertically divided into two CUs, the sizes of the two divided CUs may be 16x32. When a CU of 32x32 size is horizontally divided into two CUs, the sizes of the two divided CUs may be 32x16. When one CU is divided into two CUs, it can be said that the CU is divided in a binary-tree form. In other words, it can be seen that binary-tree partitioning has been applied to the CU.

For example, when one CU is divided into three CUs, the three divided CUs may be generated by dividing the horizontal or vertical size of the CU before being divided at a ratio of 1:2:1. For example, when a CU of 16x32 size is divided into 3 CUs in the horizontal direction, the 3 divided CUs may have sizes of 16x8, 16x16 and 16x8, respectively, from the top. For example, when a CU of size 32x32 is divided into three CUs in the vertical direction, the three divided CUs may have sizes of 8x32, 16x32, and 8x32 from the left, respectively. When one CU is divided into three CUs, it can be said that the CU is divided in a ternary-tree form. In other words, it can be seen that ternary-tree partitioning is applied to the CU.

Both quad-tree partitioning and binary-tree partitioning are applied to the LCU 310 of FIG. 3 .

In the encoding apparatus 100, a 64x64 Coding Tree Unit (CTU) may be divided into a plurality of smaller CUs by a recursive quad-tree structure. One CU can be divided into 4 CUs with equal sizes. CUs can be partitioned recursively, and each CU can have a structure of a quad tree.

Through recursive segmentation on the CU, an optimal segmentation method that results in the smallest rate-distortion ratio can be selected.

The CTU 320 of FIG. 3 is an example of a CTU to which quad tree partitioning, binary tree partitioning, and ternary tree partitioning are all applied.

As described above, in order to divide the CTU, at least one of quad tree partitioning, binary tree partitioning, and ternary tree partitioning may be applied to the CTU. Partitions can be applied based on a specified priority order.

For example, quad tree partitioning may be preferentially applied to CTUs. A CU that cannot be further divided into a quad tree may correspond to a leaf node of a quad tree. A CU corresponding to a leaf node of a quad tree may be a root node of a binary tree and/or a ternary tree. That is, a CU corresponding to a leaf node of a quad tree may be split in the form of a binary tree or a ternary tree, or may not be split any further. At this time, quad tree splitting is not applied again to a CU generated by applying binary tree splitting or ternary tree splitting to a CU corresponding to a leaf node of a quad tree, so that block splitting and/or signaling of block splitting information is performed. can be done effectively.

Splitting of a CU corresponding to each node of the quad tree may be signaled using quad splitting information. Quad splitting information having a first value (eg, “1”) may indicate that the CU is split in the form of a quad tree. Quad splitting information having a second value (eg, “0”) may indicate that the CU is not split in the form of a quad tree. The quad division information may be a flag having a specified length (eg, 1 bit).

There may be no priority between binary tree splitting and ternary tree splitting. That is, a CU corresponding to a leaf node of a quad tree may be partitioned in a binary tree form or a ternary tree form. In addition, a CU generated by binary tree splitting or ternary tree splitting may be split again into a binary tree shape or a ternary tree shape, or may not be split any further.

Partitioning in the case where there is no priority between binary tree partitioning and ternary tree partitioning may be referred to as multi-type tree partitioning. That is, a CU corresponding to a leaf node of a quad tree may become a root node of a multi-type tree. Splitting of a CU corresponding to each node of the multi-type tree may be signaled using at least one of information indicating whether or not the multi-type tree is split, splitting direction information, and splitting tree information. For splitting of a CU corresponding to each node of the multi-type tree, information indicating whether to split or not, splitting direction information, and splitting tree information may be signaled sequentially.

For example, information indicating whether a multi-type tree having a first value (eg, “1”) is split may indicate that the corresponding CU is split into a multi-type tree. Information indicating whether the multi-type tree having a second value (eg, “0”) is split may indicate that the corresponding CU is not split into a multi-type tree.

When a CU corresponding to each node of a multi-type tree is split into a multi-type tree, the corresponding CU may further include splitting direction information.

The splitting direction information may indicate a splitting direction of multi-type tree splitting. The division direction information having a first value (eg, “1”) may indicate that the corresponding CU is divided in the vertical direction. The division direction information having a second value (eg, “0”) may indicate that the corresponding CU is divided in the horizontal direction.

When a CU corresponding to each node of a multi-type tree is split into a multi-type tree, the corresponding CU may further include split tree information. Split tree information may indicate a tree used for multi-type tree split.

For example, split tree information having a first value (eg, “1”) may indicate that the corresponding CU is split in the form of a binary tree. Split tree information having a second value (eg, “0”) may indicate that the corresponding CU is split in the form of a ternary tree.

Here, each of the aforementioned information indicating whether to split or not, split tree information, and split direction information may be a flag having a specified length (eg, 1 bit).

At least one of the above-described quad splitting information, information indicating whether to split a multi-type tree, splitting direction information, and splitting tree information may be entropy-encoded and/or entropy-decoded. For entropy encoding/decoding of such information, information of neighboring CUs adjacent to the target CU may be used.

For example, it may be considered that there is a high probability that the division shape of the left CU and/or the upper CU (that is, whether or not to split, the split tree, and/or the split direction) and the split shape of the target CU are similar to each other. Accordingly, context information for entropy encoding and/or entropy decoding of information of a target CU may be derived based on information of a neighboring CU. In this case, the information of the neighboring CU may include at least one of 1) quad split information, 2) information indicating whether the multi-type tree is split, 3) split direction information, and 4) split tree information of the neighboring CU.

As another embodiment, binary tree splitting may be preferentially performed among binary tree splitting and ternary tree splitting. That is, binary tree splitting is applied first, and a CU corresponding to a leaf node of the binary tree may be set as a root node of the ternary tree. In this case, quad tree splitting and binary tree splitting may not be performed on a CU corresponding to a node of a ternary tree.

A CU that is not further split by quad tree splitting, binary tree splitting, and/or ternary tree splitting may become a unit of encoding, prediction, and/or transformation. That is, for prediction and/or transformation, the CU may not be further split. Accordingly, a partitioning structure and partitioning information for dividing a CU into prediction units and/or transform units may not exist in the bitstream.

However, when the size of a CU serving as a unit of division is larger than the size of the maximum transform block, this CU may be recursively split until the size of the CU is less than or equal to the size of the maximum transform block. For example, when the size of a CU is 64x64 and the size of a maximum transform block is 32x32, the CU may be divided into four 32x32 blocks for transform. For example, when the size of a CU is 32x64 and the size of a maximum transform block is 32x32, the CU may be divided into two 32x32 blocks for transform.

In this case, information on whether the CU is split for conversion may not be separately signaled. Without signaling, whether to split the CU may be determined by comparing the horizontal size (and/or vertical size) of the CU and the horizontal size (and/or vertical size) of the largest transform block. For example, when the horizontal size of the CU is greater than the horizontal size of the largest transform block, the CU may be vertically divided into two parts. In addition, when the vertical size of the CU is greater than the vertical size of the largest transform block, the CU may be divided into two horizontally.

Information on the maximum size and/or minimum size of a CU and information on the maximum size and/or minimum size of a transform block may be signaled or determined at a higher level for a CU. For example, the upper level may be a sequence level, a picture level, a tile level, a tile group level, and a slice level. For example, the minimum size of a CU may be determined to be 4x4. For example, the maximum size of a transform block may be determined to be 64x64. For example, the minimum size of a transform block may be determined to be 4x4.

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

Difference information about the size of the CTU and the maximum size of the transform block may be signaled or determined at a higher level for the CU. For example, the higher level may be a sequence level, a picture level, a slice level, a tile group level, and a tile level. Information about the maximum size of the CU corresponding to each node of the binary tree (ie, the maximum size of the binary tree) may be determined based on the size and difference information of the CTU. The maximum size of the CU corresponding to each node of the ternary tree (ie, the maximum size of the ternary tree) may have a different value depending on the slice type. For example, within an intra slice, the maximum size of the ternary tree may be 32x32. Also, for example, within an inter slice, the maximum size of a ternary tree may be 128x128. For example, the minimum size of a CU corresponding to each node of a binary tree (say, the minimum size of a binary tree) and/or the minimum size of a CU corresponding to each node of a ternary tree (say, the minimum size of a ternary tree) is Can be set to a minimum size.

As another example, the maximum size of the binary tree and/or the maximum size of the ternary tree may be signaled or determined at the slice level. Also, 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.

Based on the aforementioned various block sizes and various depths, quad splitting information, information indicating whether to split a multi-type tree, splitting tree information, and/or splitting direction information may or may not exist in the bitstream.

For example, if the size of the CU is not larger than the quad tree minimum size, the CU may not include quad splitting information, and the quad splitting information for the CU may be inferred as a second value.

For example, if the size (horizontal size and vertical size) of a CU corresponding to a node of a multi-type tree is larger than the binary tree maximum size (horizontal size and vertical size) and/or the ternary tree maximum size (horizontal size and vertical size) If larger, the CU may not be partitioned into binary tree form and/or ternary tree form. According to this determination method, information indicating whether to split the multi-type tree may not be signaled and may be inferred as a second value.

Alternatively, the size (horizontal size and vertical size) of a CU corresponding to a node of a multi-type tree is equal to the minimum size (horizontal size and vertical size) of a binary tree, or the size of a CU (horizontal size and vertical size) is equal to the size of a ternary tree. When equal to twice the minimum size (horizontal size and vertical size), the CU may not be split into a binary tree shape and/or a ternary tree shape. According to this determination method, information indicating whether to split the multi-type tree may not be signaled and may be inferred as a second value. This is because when the CU is divided into a binary tree shape and/or a ternary tree shape, a CU smaller than the minimum size of the binary tree and/or the minimum size of the ternary tree is generated.

Alternatively, binary tree splitting or ternary tree splitting may be limited based on the size of the virtual pipeline data unit (ie, pipeline buffer size). For example, binary tree splitting or ternary tree splitting may be limited when a CU is split into sub-CUs not suitable for a pipeline buffer size by binary tree splitting or ternary tree splitting. The pipeline buffer size may be equal to the size of the largest transform block (eg, 64X64).

For example, when the pipeline buffer size is 64X64, the following partitions can be limited.

- ternary tree split over NxM (N and/or M is 128) CUs

- horizontal binary tree partitioning over 128xN (N <= 64) CUs

- Vertical binary tree partitioning for Nx128 (N <= 64) CUs

Alternatively, when the depth in the multi-type tree of a CU corresponding to a node of the multi-type tree is equal to the maximum depth of the multi-type tree, the CU may not be split into a binary tree form and/or a ternary tree form. According to this determination method, information indicating whether to split the multi-type tree may not be signaled and may be inferred as a second value.

Alternatively, only when at least one of vertical binary tree splitting, horizontal binary tree splitting, vertical ternary tree splitting, and horizontal ternary tree splitting is possible for a CU corresponding to a node of a multi-type tree, a multi-type tree Information indicating whether to divide may be signaled. Otherwise, the CU may not be partitioned in binary tree form and/or ternary tree form. According to this determination method, information indicating whether to split the multi-type tree may not be signaled and may be inferred as a second value.

Alternatively, only when both vertical binary tree splitting and horizontal binary tree splitting are possible for a CU corresponding to a node of a multi-type tree, or both vertical ternary tree splitting and horizontal ternary tree splitting are possible, splitting direction information may be signaled. Otherwise, the division direction information may not be signaled and may be inferred as a value indicating a direction in which the CU may be divided.

Alternatively, only when both vertical binary tree splitting and vertical ternary tree splitting are possible for a CU corresponding to a node of a multi-type tree, or both horizontal binary tree splitting and horizontal ternary tree splitting are possible, split tree information may be signaled. Otherwise, the split tree information may not be signaled and may be inferred as a value indicating a tree applicable to split of a CU.

4 is a diagram illustrating a form of a prediction unit that a coding unit may include.

Among the CUs divided from the LCU, a CU that is not further divided may be divided into one or more prediction units (PUs).

A PU may be a basic unit for prediction. A PU may be coded and decoded in any one of skip mode, inter mode, and intra mode. A PU may be divided into various types according to each mode. For example, the target block described above with reference to FIG. 1 and the target block described above with reference to FIG. 2 may be PUs.

A CU may not be divided into PUs. When a CU is not divided into PUs, the size of a CU and the size of a PU may be the same.

In skip mode, there may not be a split within a CU. In the skip mode, a 2Nx2N mode 410 in which sizes of PU and CU are the same may be supported without division.

In inter mode, 8 divided types can be supported within the CU. For example, in inter mode, 2Nx2N mode 410, 2NxN mode 415, Nx2N mode 420, NxN mode 425, 2NxnU mode 430, 2NxnD mode 435, nLx2N mode 440 and nRx2N Mode 445 may be supported.

In intra mode, 2Nx2N mode 410 and NxN mode 425 may be supported.

In the 2Nx2N mode 410, a PU having a size of 2Nx2N may be encoded. A PU having a size of 2Nx2N may mean a PU having the same size as that of a CU. For example, a 2Nx2N PU may have a size of 64x64, 32x32, 16x16 or 8x8.

In the NxN mode 425, a PU having a size of NxN may be encoded.

For example, in intra prediction, when the size of a PU is 8x8, 4 divided PUs can be coded. The size of the divided PU may be 4x4.

When a PU is coded using an intra mode, the PU may be coded using one intra prediction mode among a plurality of intra prediction modes. For example, High Efficiency Video Coding (HEVC) technology may provide 35 intra prediction modes, and a PU may be coded in one intra prediction mode among the 35 intra prediction modes.

Which mode among the 2Nx2N mode 410 and the NxN mode 425 the PU will be encoded may be determined by a rate-distortion cost.

The encoding apparatus 100 may perform an encoding operation on a PU having a size of 2Nx2N. Here, the encoding operation may be encoding a PU in each of a plurality of intra prediction modes usable by the encoding apparatus 100 . An optimal intra prediction mode for a PU having a size of 2Nx2N may be derived through an encoding operation. An optimal intra prediction mode may be an intra prediction mode that generates a minimum rate-distortion cost for encoding a 2Nx2N PU among a plurality of intra prediction modes usable by the encoding apparatus 100 .

Also, the encoding apparatus 100 may sequentially perform an encoding operation on each PU of NxN-divided PUs. Here, the encoding operation may be encoding a PU in each of a plurality of intra prediction modes usable by the encoding apparatus 100 . An optimal intra prediction mode for an NxN-sized PU may be derived through an encoding operation. An optimal intra-prediction mode may be an intra-prediction mode that generates the minimum rate-distortion cost for encoding an NxN-sized PU among a plurality of intra-prediction modes usable by the encoding apparatus 100.

The encoding apparatus 100 may determine which of the 2Nx2N-sized PU and the NxN-sized PUs to encode based on the comparison of the rate-distortion cost of the 2Nx2N-sized PU and the rate-distortion costs of the NxN-sized PUs.

One CU may be divided into one or more PUs, and a PU may also be divided into a plurality of PUs.

For example, when one PU is divided into 4 PUs, the horizontal size and vertical size of each of the 4 PUs generated by the division are half of the horizontal size and half of the vertical size of the PU before the division, respectively. can When a 32x32 PU is divided into 4 PUs, the sizes of the 4 PUs may be 16x16. When one PU is divided into 4 PUs, it can be said that the PU is divided in a quad-tree shape.

For example, when one PU is split into two PUs, the horizontal size or vertical size of each PU of the two PUs generated by the split is half of the horizontal size or half of the vertical size of the PU before splitting, respectively. can When a 32x32 PU is vertically divided into 2 PUs, the sizes of the 2 PUs may be 16x32. When a 32x32-sized PU is horizontally divided into two PUs, the sizes of the two divided PUs may be 32x16. When one PU is divided into two PUs, it can be said that the PU is divided in the form of a binary-tree.

5 is a diagram illustrating a form of a transform unit that may be included in a coding unit.

A transform unit (TU) may be a basic unit used for transformation, quantization, inverse transformation, inverse quantization, entropy encoding, and entropy decoding processes within the CU.

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

Among the CUs split from the LCU, a CU that is no longer split into CUs may be split into one or more TUs. In this case, the division structure of the TU may be a quad-tree structure. For example, as shown in FIG. 5 , one CU 510 may be divided one or more times according to a quad-tree structure. Through division, one CU 510 may be composed of TUs of various sizes.

When one CU is split two or more times, the CU can be considered to be split recursively. Through splitting, one CU may be composed of TUs having various sizes.

Alternatively, one CU may be divided into one or more TUs based on the number of vertical and/or horizontal lines dividing the CU.

A CU may be divided into symmetric TUs or asymmetric TUs. For division into asymmetric TUs, information on the size and/or shape of a TU may be signaled from the encoding device 100 to the decoding device 200. Alternatively, the size and/or shape of the TU may be derived from information on the size and/or shape of the CU.

A CU may not be divided into TUs. When a CU is not divided into TUs, the size of a CU and the size of a TU may be the same.

One CU may be divided into one or more TUs, and a TU may also be divided into a plurality of TUs.

For example, when one TU is divided into 4 TUs, the horizontal and vertical sizes of each TU of the 4 TUs generated by the division are half of the horizontal size and half of the vertical size of the TU before the division, respectively. can When a TU of 32x32 size is divided into 4 TUs, the sizes of the 4 divided TUs may be 16x16. When one TU is divided into 4 TUs, it can be said that the TU is divided in the form of a quad-tree.

For example, when one TU is split into two TUs, the horizontal size or vertical size of each TU of the two TUs generated by the split is half of the horizontal size or half of the vertical size of the TU before the split, respectively. can When a 32x32 size TU is vertically split into two TUs, the sizes of the two split TUs may be 16x32. When a TU of size 32x32 is horizontally divided into two TUs, the sizes of the two divided TUs may be 32x16. When one TU is split into two TUs, it can be said that the TU is split in a binary-tree format.

CUs may be divided in other ways than shown in FIG. 5 .

For example, one CU may be split into 3 CUs. The horizontal size or vertical size of the three divided CUs may be 1/4, 1/2, and 1/4 of the horizontal or vertical size of the CU before division, respectively.

For example, when a CU of 32x32 size is vertically divided into 3 CUs, the sizes of the 3 divided CUs may be 8x32, 16x32 and 8x32, respectively. In this way, when one CU is split into three CUs, it can be considered that the CU is split in the form of a ternary tree.

One of the exemplified quad tree-type partitioning, binary tree-type partitioning, and ternary tree-type partitioning may be applied for partitioning of the CU, and a plurality of partitioning schemes may be combined together and used for partitioning of the CU. . In this case, a case in which a plurality of partitioning methods are combined and used may be referred to as a composite tree type of partitioning.

6 shows division of a block according to an example.

In the process of encoding and/or decoding an image, a target block may be divided as shown in FIG. 6 . For example, the target block may be a CU.

For division of the target block, an indicator indicating division information may be signaled from the encoding apparatus 100 to the decoding apparatus 200. The division information may be information indicating how the target block is divided.

Splitting information includes a split flag (hereinafter referred to as “split_flag”), a quad-binary flag (hereinafter referred to as “QB_flag”), a quad-tree flag (hereinafter referred to as “quadtree_flag”), and a binary tree flag (hereinafter referred to as “binarytree_flag”). ") and a binary type flag (hereinafter referred to as "Btype_flag").

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

QB_flag may be a flag indicating in which form a block is divided into a quad tree form and a binary tree form. For example, a value of 0 of QB_flag may indicate that a block is divided in a quad tree form. A value of 1 of QB_flag may indicate that a block is divided in the form of a binary tree. Alternatively, a value of 0 of QB_flag may indicate that the block is divided in the form of a binary tree. A value of 1 for QB_flag may indicate that a block is divided in a quad tree form.

quadtree_flag may be a flag indicating whether a block is split into a quad tree. For example, a value of 1 of quadtree_flag may indicate that a block is split into a quad tree. A value of 0 of quadtree_flag may indicate that a block is not split in the form of a quad tree.

binarytree_flag may be a flag indicating whether a block is split in the form of a binary tree. For example, a value of 1 in binarytree_flag may indicate that a block is split into a binary tree. A value of 0 in binarytree_flag may indicate that a block is not split in the form of a binary tree.

Btype_flag may be a flag indicating whether the block is split vertically or horizontally when the block is split in the form of a binary tree. For example, a value of 0 in Btype_flag may indicate that a block is divided in a horizontal direction. A value of 1 of Btype_flag may indicate that a block is divided in a vertical direction. Alternatively, a value of 0 of Btype_flag may indicate that the block is divided in the vertical direction. A value of 1 of Btype_flag may indicate that the block is divided in the horizontal direction.

For example, partition information for the block of FIG. 6 can be derived by signaling at least one of quadtree_flag, binarytree_flag, and Btype_flag as shown in Table 1 below.

[Table 1]

For example, the split information for the block of FIG. 6 can be derived by signaling at least one of split_flag, QB_flag, and Btype_flag as shown in Table 2 below.

[Table 2]

The partitioning method may be limited to a quad tree or only 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 to split into a quad tree type or a flag indicating whether to split into a binary tree type. The size and shape of the block may be derived according to the depth information of the block, and the depth information may be signaled from the encoding apparatus 100 to the decoding apparatus 200.

When the size of a block falls within a specified range, only quad-tree partitioning may be possible. For example, the specified range may be defined by at least one of a maximum block size and a minimum block size in which quad-tree partitioning is possible.

Information indicating the maximum block size and/or minimum block size in which only quad tree-type division is possible may be signaled from the encoding apparatus 100 to the decoding apparatus 200 through a bitstream. Also, this information may be signaled for at least one unit of video, sequence, picture, parameter, tile group, and slice (or segment).

Alternatively, the maximum block size and/or the minimum block size may be fixed sizes predefined in the encoding apparatus 100 and the decoding apparatus 200. For example, when the size of a block is larger than 64x64 and smaller than 256x256, only quad-tree partitioning may be possible. In this case, split_flag may be a flag indicating whether to split into a quad tree.

If the size of the block is larger than the maximum transform block size, only a quad tree type of splitting may be possible. In this case, the divided block may be at least one of CU and TU.

In this case, split_flag may be a flag indicating whether to split into a quad tree.

When the size of a block falls within a specified range, only binary tree or ternary tree partitioning may be possible. Here, for example, the specified range may be defined by at least one of a maximum block size and a minimum block size in which binary tree or ternary tree partitioning is possible.

Information representing a maximum block size and/or a minimum block size in which only binary tree-type division or ternary tree-type division is possible 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 unit of sequence, picture, and slice (or segment).

Alternatively, the maximum block size and/or the minimum block size may be fixed sizes predefined in the encoding apparatus 100 and the decoding apparatus 200. For example, when the size of a block is greater than 8x8 and less than 16x16, only binary tree partitioning may be possible. In this case, split_flag may be a flag indicating whether to split into a binary tree type or a ternary tree type.

The above-described quad tree partitioning may be equally applied to binary tree and/or ternary tree partitioning.

The division of a block may be limited by the previous division. For example, when a block is divided into a specified binary tree shape to generate a plurality of divided blocks, each divided block may be additionally divided only into the specified tree shape. Here, the specified tree type may be at least one of a binary tree type, a ternary tree type, and a quad tree type.

When the horizontal size or vertical size of a divided block corresponds to a size that cannot be further divided, the aforementioned indicator may not be signaled.

7 is a diagram for explaining an embodiment of an intra prediction process.

Arrows extending from the center to the periphery of the graph of FIG. 7 may represent prediction directions of directional intra prediction modes. Also, numbers displayed adjacent to arrows may represent examples of mode values assigned to an intra prediction mode or a prediction direction of the intra prediction mode.

In FIG. 7 , the number 0 may indicate a planar mode, which is a non-directional intra prediction mode. Number 1 may indicate a DC mode, which is a non-directional intra prediction mode.

Intra coding and/or decoding may be performed using reference samples of neighboring blocks of the target block. A neighboring block may be a reconstructed neighboring block. A reference sample may mean a neighboring sample.

For example, intra encoding and/or decoding may be performed using a value of a reference sample included in a reconstructed neighboring block or a coding parameter.

The encoding apparatus 100 and/or the decoding apparatus 200 may generate a prediction block by performing intra prediction on the target block based on sample information in the target image. When performing intra prediction, the encoding apparatus 100 and/or the decoding apparatus 200 may generate a prediction block for a target block by performing intra prediction based on sample information in a target image. When intra 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.

A prediction block may refer to a block generated as a result of performing intra prediction. A prediction block may correspond to at least one of CU, PU, and TU.

A unit of a prediction block may be the size of at least one of CU, PU, and TU. The prediction block may have a square shape with a size of 2Nx2N or NxN. The size of NxN may include 4x4, 8x8, 16x16, 32x32, and 64x64.

Alternatively, the prediction block may be a square block having a size of 2x2, 4x4, 8x8, 16x16, 32x32, or 64x64, or may be a rectangular block having a size of 2x8, 4x8, 2x16, 4x16, or 8x16. there is.

Intra prediction may be performed according to an intra prediction mode for a target block. The number of intra prediction modes that the target block may have may be a predefined fixed value or may be a value determined differently according to properties of the prediction block. For example, properties of the prediction block may include a size of the prediction block and a type of the prediction block. Also, the properties of the prediction block may indicate coding parameters for the prediction block.

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

The intra prediction mode may be a non-directional mode or a directional mode.

For example, the intra prediction mode may include two non-directional modes and 65 directional modes, corresponding to numbers 0 to 66 shown in FIG. 7 .

For example, when a specified intra prediction method is used, the intra prediction mode may include two non-directional modes and 93 directional modes, corresponding to numbers -14 to 80 shown in FIG. 7 .

The two non-directional modes may include DC mode and Planar mode.

The directional mode may be a prediction mode having a specific direction or a specific angle. The directional mode may also be referred to as an argular mode.

The intra prediction mode may be expressed as at least one of a mode number, a mode value, a mode angle, and a mode direction. That is to say, the terms “(mode) number of an intra prediction mode”, “(mode) value of an intra prediction mode”, “(mode) angle of an intra prediction mode” and “(mode) direction of an intra prediction mode) have the same meaning. and can be used interchangeably.

The number of intra prediction modes may be M. M may be 1 or more. In other words, the number of intra prediction modes may be M including the number of non-directional modes and the number of directional modes.

The number of intra prediction modes may be fixed to M regardless of block sizes and/or color components. For example, the number of intra prediction modes may be fixed to either 35 or 67 regardless of the size of a block.

Alternatively, the number of intra prediction modes may be different according to the shape, size, and/or color component type of a block.

For example, in FIG. 7 , directional prediction modes indicated by dotted lines may be applied only to prediction for non-square blocks.

For example, as the block size increases, the number of intra prediction modes may increase. Alternatively, as the block size increases, the number of intra prediction modes may decrease. When the size of a block is 4x4 or 8x8, the number of intra prediction modes may be 67. When the size of a block is 16x16, the number of intra prediction modes may be 35. When the size of a block is 32x32, the number of intra prediction modes may be 19. When the size of a block is 64x64, the number of intra prediction modes may be 7.

For example, the number of intra prediction modes may be different depending on whether the color component is a luma signal or a chroma signal. Alternatively, the number of intra prediction modes of the luma component block may be greater than the number of intra prediction modes of the chroma component block.

For example, in the case of a vertical mode having a mode value of 50, prediction may be performed in a vertical direction based on a pixel value of a reference sample. For example, in the case of a horizontal mode having a mode value of 18, prediction may be performed in a horizontal direction based on a pixel value of a reference sample.

Even in a directional mode other than the aforementioned mode, the encoding apparatus 100 and the decoding apparatus 200 may perform intra prediction on a target unit using a reference sample according to an angle corresponding to the directional mode.

An intra prediction mode positioned to the right of the vertical mode may be named a vertical-right mode. An intra-prediction mode positioned below the horizontal mode may be named a horizontal-below mode. For example, in FIG. 7 , intra prediction modes having a mode value of one of 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, and 66 are vertical It can be the right modes. Intra prediction modes having a mode value of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, and 17 may be horizontal bottom modes.

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

The directional mode may include an angular mode. Modes other than the DC mode and the planner mode among the plurality of intra prediction modes may be directional modes.

When the intra prediction mode is the DC mode, a prediction block may be generated based on an average of pixel values of a plurality of reference samples. For example, a pixel value of a prediction block may be determined based on an average of pixel values of a plurality of reference samples.

The number of intra prediction modes and the mode value of each intra prediction mode described above may be merely illustrative. The number of the aforementioned intra prediction modes and the mode value of each intra prediction mode may be differently defined according to embodiments, implementations, and/or needs.

In order to perform intra prediction on the target block, a step of checking whether samples included in the reconstructed neighboring block can be used as reference samples of the target block may be performed. If there is a sample that cannot be used as a reference sample of the target block among the samples of the neighboring block, a value generated by copying and/or interpolation using at least one sample value among the samples included in the reconstructed neighboring block This can be replaced with the sample value of a sample that cannot be used as a reference sample. If a value generated by copying and/or interpolation is replaced with a sample value of a sample, the sample may be used as a reference sample of the target block.

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

The type of filter applied to at least one of the reference sample and the prediction sample may differ according to at least one of an intra prediction mode of the target block, a size of the target block, and a shape of the target block. The type of filter may be classified according to one or more of a length of a filter tap, a value of a filter coefficient, and a filter strength. The length of the filter tap may mean the number of filter taps. Also, the number of filter taps may mean the length of a filter.

When the intra prediction mode is the planner mode, in generating the prediction block of the target block, the upper reference sample of the target sample, the left reference sample of the target sample, and the upper right reference sample of the target block according to the location of the prediction target sample in the prediction block. A sample value of the prediction target sample may be generated using a weight-sum of the lower left reference sample of the target block.

When the intra prediction mode is the DC mode, an average value of upper reference samples and left reference samples of the target block may be used to generate a prediction block of the target block. In addition, filtering using values of reference samples may be performed on specified rows or specified columns within the target block. The specified rows may be one or more top rows adjacent to the reference sample. The specified columns may be one or more left columns adjacent to the reference sample.

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

Interpolation in units of real numbers may be performed to generate the prediction samples described above.

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

For example, if the intra prediction modes of the target block and the neighboring block are the same, it may be signaled that the intra prediction modes of the target block and the neighboring block are the same using a predefined flag.

For example, an indicator indicating an intra prediction mode identical to that of a target block among intra prediction modes of a plurality of neighboring blocks may be signaled.

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

8 is a diagram for explaining reference samples used in an intra prediction process.

The reconstructed reference samples used for intra prediction of the target block include below-left reference samples, left reference samples, above-left corner reference samples, and above reference samples. s and above-right reference samples, etc.

For example, the left reference samples may refer to reconstructed reference pixels adjacent to the left side of the target block. Top reference samples may refer to reconstructed reference pixels adjacent to the top of the target block. The top left corner reference sample may refer to a reconstructed reference pixel located at the top left corner of the target block. In addition, the lower left reference samples may refer to a reference sample located at the lower end of the left sample line among samples located on the same line as the left sample line composed of the left reference samples. The upper right reference samples may refer to reference samples located to the right of the upper pixel line among samples located on the same line as the upper sample line composed of the upper reference samples.

When the size of the target block is NxN, each of the lower left reference samples, the left reference samples, the upper reference samples, and the upper right reference samples may be N.

A prediction block may be generated through intra prediction of the target block. Generation of the prediction block may include determining values of pixels of the prediction block. The size of the target block and the prediction block may be the same.

Reference samples used for intra prediction of the target block may vary according to the intra prediction mode of the target block. A direction of an intra prediction mode may represent a dependency relationship between reference samples and pixels of a prediction block. For example, a value of a specified reference sample may be used as a value of one or more specified pixels of a prediction block. In this case, the specified reference sample and the specified one or more pixels of the prediction block may be samples and pixels specified as a straight line in the direction of the intra prediction mode. In other words, the value of the specified reference sample may be copied to a value of a pixel located in a direction opposite to the direction of the intra prediction mode. Alternatively, the value of a pixel of the prediction block may be a value of a reference sample located in the direction of the intra prediction mode based on the position of the pixel.

For example, when the intra prediction mode of the target block is a vertical mode, upper reference samples may be used for intra prediction. When the intra prediction mode is a vertical mode, a value of a pixel of a prediction block may be a value of a reference sample positioned vertically above the position of the pixel. Accordingly, the top reference samples that are top adjacent to the target block may be used for intra prediction. Also, values of pixels in one row of the prediction block may be the same as values of upper reference samples.

For example, when the intra prediction mode of the target block is a horizontal mode, left reference samples may be used for intra prediction. When the intra prediction mode is a horizontal mode, a value of a pixel of a prediction block may be a value of a reference sample located horizontally to the left of the pixel. Accordingly, left reference samples adjacent to the left of the target block may be used for intra prediction. Also, values of pixels of one column of the prediction block may be the same as values of left reference samples.

For example, when the mode value of the intra prediction mode of the target block is 34, at least some of the left reference samples, the upper left corner reference sample, and the upper reference samples may be used for intra prediction. When the mode value of the intra prediction mode is 34, the value of the pixel of the prediction block may be the value of the reference sample located diagonally at the top left of the pixel.

Also, when an intra prediction mode having a mode value of one of 52 to 66 is used, at least some of the upper right reference samples may be used for intra prediction.

Also, when an intra prediction mode having a mode value of one of 2 to 17 is used, at least some of lower left reference samples may be used for intra prediction.

Also, when an intra prediction mode having a mode value of one of 19 to 49 is used, the upper left corner reference sample may be used for intra prediction.

The number of reference samples used to determine the pixel value of one pixel of the prediction block may be one or two or more.

As described above, a pixel value of a pixel of a prediction block may be determined according to 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 reference sample indicated by the position of the pixel and the direction of the intra prediction mode is an integer position, a value of one reference sample indicated by the integer position may be used to determine a pixel value of a pixel of the prediction block.

When the position of the reference sample indicated by the position of the pixel and the direction of the intra prediction mode is not an integer position, an interpolated reference sample may be generated based on two reference samples closest to the position of the reference sample. there is. The value of the interpolated reference sample may be used to determine a pixel value of a pixel of the predictive block. In other words, when the location of a pixel of the prediction block and the location of a reference sample indicated by the direction of the intra prediction mode indicate a gap between two reference samples, an interpolated value will be generated based on the values of the two samples. can

A prediction block generated by prediction may not be the same as the original target block. In other words, a prediction error, which is a difference between a target block and a prediction block, may exist, and a prediction error may also exist between pixels of the target block and pixels of the prediction block.

Hereinafter, the terms “difference”, “error” and “residual” may be used interchangeably and interchangeably.

For example, in the case of directional intra prediction, a larger prediction error may occur as the distance between a pixel of a prediction block and a reference sample increases. A discontinuity may occur between a prediction block generated by such a prediction error and a neighboring block.

Filtering of prediction blocks may be used to reduce prediction errors. Filtering may be adaptively applying a filter to a region considered to have a large prediction error among prediction blocks. For example, a region considered to have a large prediction error may be a boundary of a prediction block. Also, depending on the intra-prediction mode, a region considered to have a large prediction error among prediction blocks may be different, and filter characteristics may be different.

As shown in FIG. 8 , at least one of reference lines 0 to 3 may be used for intra prediction of a target block.

Each reference line in FIG. 8 may represent a reference sample line including one or more reference samples. The smaller the reference line number, the closer the reference sample line may be to the target block.

Samples of segment A and segment F may be obtained through padding using the nearest samples of segment B and segment E, respectively, instead of being obtained from reconstructed neighboring blocks.

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 used for intra prediction of a target block among a plurality of reference sample lines. For example, index information may have a value of 0 to 3.

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 a reference sample line other than reference sample line 0 is used, filtering of a prediction block described later may not be performed.

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

For example, the first color component may be a luma component, and the second color component may be a chroma component.

For intra prediction between color components, parameters of a linear model between a first color component and a second color component may be derived based on a template.

The template may include the top reference sample and/or the left reference sample of the target block, and the top reference sample and/or left reference sample of the reconstructed block of the first color component corresponding to these reference samples. there is.

For example, the parameters of the linear model are: 1) the value of the sample of the first color component having the largest value among the samples in the template, 2) the value of the sample of the second color component corresponding to the sample of this 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.

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

Depending on the image format, sub-sampling may be performed on the neighboring samples of the reconstructed block of the first color component and the corresponding reconstructed block. For example, when one sample of the second color component corresponds to four samples of the first color component, one corresponding sample may be calculated by subsampling the four samples of the first color component. there is. When subsampling is performed, derivation of parameters of the linear model and intra prediction between color components may be performed based on the subsampled corresponding samples.

Whether intra prediction between color components is performed and/or a template range may be signaled as an intra prediction mode.

A target block may be divided into 2 or 4 sub-blocks in a horizontal direction and/or a vertical direction.

Divided sub-blocks may be sequentially reconstructed. That is, as intra prediction is performed on a sub-block, a sub-prediction block for the sub-block may be generated. In addition, as inverse quantization and/or inverse transformation are performed on the sub-block, a sub-residual block for the sub-block may be generated. A reconstructed sub-block may be generated by adding the sub-prediction block to the sub-residual block. The reconstructed subblock may be used as a reference sample for intra prediction of a subblock of a later order.

A sub-block may be a block including a specified number (eg, 16) or more samples. Thus, for example, when the target block is an 8x4 block or a 4x8 block, the target block may be divided into two sub-blocks. Also, if the target block is a 4x4 block, the target block cannot be divided into sub blocks. If the target block has other sizes, the target block may be divided into 4 sub-blocks.

Information about whether intra prediction based on these sub-blocks is performed and/or a division direction (horizontal direction or vertical direction) may be signaled.

Such sub-block-based intra prediction may be limited to being performed only when the reference sample line 0 is used. When sub-block-based intra prediction is performed, filtering of a prediction block described later may not be performed.

A final prediction block may be generated by performing filtering on a prediction block generated by intra prediction.

Filtering may be performed by applying specific weights to the filtering target sample, the left reference sample, the top reference sample, and/or the top left reference sample.

A weight and/or a reference sample (or a range of reference samples or a location of the reference sample) used for filtering may be determined based on at least one of a block size, an intra prediction mode, and a position of a sample to be filtered within a prediction block. there is.

For example, filtering may be performed only for a specified intra prediction mode (eg, DC mode, planar mode, vertical mode, horizontal mode, diagonal mode, and/or adjacent diagonal mode).

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

The intra prediction mode of the target block may be derived using intra prediction modes of neighboring blocks existing around the target block, and the derived intra prediction mode may be entropy-encoded and/or entropy-decoded.

For example, if the intra-prediction mode of the target block and the intra-prediction mode of the neighboring block are the same, information that the intra-prediction mode of the target block and the intra-prediction mode of the neighboring block are the same may be signaled using specified flag information. .

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

For example, if the intra prediction mode of the target block and the intra prediction mode of the neighboring block are different, entropy encoding and/or entropy decoding based on the intra prediction mode of the neighboring block is performed to obtain information about the intra prediction mode of the target block. Entropy encoding and/or entropy decoding may be performed.

9 is a diagram for explaining an embodiment of an inter prediction process.

The rectangle shown in FIG. 9 may represent an image (or picture). Also, arrows in FIG. 9 may indicate prediction directions. An arrow pointing from a first picture to a second picture may indicate that the second picture refers to the first picture. That is, an image may be encoded and/or decoded according to a prediction direction.

Each image may be classified into an intra picture (I picture), a uni-prediction picture (P picture), and a bi-prediction picture (B picture) according to an encoding type. Each picture may be coded and/or decoded according to the coding type of each picture.

When a target image to be encoded is an I-picture, the target image may be encoded using data within the image itself without inter prediction referring to another image. For example, an I picture can be coded only with intra prediction.

When the target image is a P picture, the target image may be coded through inter prediction using only reference pictures existing in one direction. Here, the unidirectional direction may be a forward direction or a reverse direction.

When the target video is a B picture, the target video may be coded through inter prediction using reference pictures existing in both directions or inter prediction using reference pictures existing in one of forward and backward directions. Here, both directions may be forward and reverse directions.

P-pictures and B-pictures that are coded and/or decoded using reference pictures may be regarded as pictures in which inter prediction is used.

Below, inter prediction in inter mode according to an embodiment is described in detail.

Inter prediction or motion compensation may be performed using a reference image and motion information.

In the inter mode, the encoding apparatus 100 may perform inter prediction and/or motion compensation on a target block. The decoding apparatus 200 may perform inter prediction and/or motion compensation corresponding to inter prediction and/or motion compensation performed in the encoding apparatus 100 on a target block.

Motion information on the target block may be derived during inter prediction by each of the encoding apparatus 100 and the decoding apparatus 200 . Motion information may be derived using motion information of a reconstructed neighboring block, motion information of a collocated block, and/or motion information of a block adjacent to the collocated block.

For example, the encoding apparatus 100 or the decoding apparatus 200 performs prediction and/or motion compensation by using motion information of a spatial candidate and/or a temporal candidate as motion information of a target block. can be done A target block may mean a PU and/or a PU partition.

A spatial candidate may be a reconstructed block that is spatially adjacent to the target block.

A temporal candidate may be a reconstructed block corresponding to a target block in an already reconstructed collocated picture (col picture).

In inter prediction, the encoding apparatus 100 and the decoding apparatus 200 may improve encoding efficiency and decoding efficiency by using motion information of spatial candidates and/or temporal candidates. Motion information of spatial candidates may be referred to as spatial motion information. Motion information of temporal candidates may be referred to as temporal motion information.

Hereinafter, motion information of a spatial candidate may be motion information of a PU including the spatial candidate. Motion information of the temporal candidate may be motion information of a PU including the temporal candidate. Motion information of the candidate block may be motion information of a PU including the candidate block.

Inter prediction may be performed using a reference picture.

A reference picture may be at least one of a previous picture of the target picture or a subsequent picture of the target picture. A reference picture may refer to an image used for prediction of a target block.

In inter prediction, a region within a reference picture may be specified by using a reference picture index (or refIdx) indicating a reference picture and a motion vector to be described later. Here, a specified region within a reference picture may represent a reference block.

Inter prediction may select a reference picture and may select a reference block corresponding to a target block within the reference picture. In addition, inter prediction may generate a prediction block for a target block using the selected reference block.

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

A spatial candidate may be a block that 1) exists in the target picture, 2) has already been reconstructed through encoding and/or decoding, and 3) is located adjacent to or at a corner of the target block. Here, the block located at the 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. "A block located at a corner of a target block" may have the same meaning as "a block adjacent to a corner of a target block". A "block located at a corner of a target block" may be included in a "block adjacent to the target block".

For example, spatial candidates include a reconstructed block located to the left of the target block, a reconstructed block located on top of the target block, a reconstructed block located at the lower left corner of the target block, and a reconstructed block located at the upper right corner of the target block. It can be a reconstructed block or a reconstructed block located in the upper left corner of the target block.

Each of the encoding apparatus 100 and the decoding apparatus 200 may identify a block existing at a position spatially corresponding to a target block within a col picture. A position of a target block in a target picture and a position of an identified block in a collocated picture may correspond to each other.

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 inside and/or outside the identified block.

For example, the collocated block may include a first collocated block and a second collocated block. When the coordinates of the identified block are (xP, yP) and the size of the identified block is (nPSW, nPSH), the first collocated block may be a block located at the coordinates (xP + nPSW, yP + nPSH). The second collocated block may be a block located at coordinates (xP + (nPSW >> 1), yP + (nPSH >> 1). The second call block may be selectively used when the first call block is unavailable.

The motion vector of the target block may be determined based on the motion vector of the collocated block. Each of the encoding apparatus 100 and the decoding apparatus 200 may scale a motion vector of a collocated block. A scaled motion vector of the collocated block may be used as a motion vector of the target block. Also, a motion vector of motion information of a temporal candidate stored in the list may be a scaled motion vector.

A ratio of the motion vector of the target block and the motion vector of the collocated block may be equal to the ratio of the first temporal distance to the second temporal distance. The first temporal distance may be the distance between the reference picture of the target block and the target picture. The second temporal distance may be a distance between a reference picture of a collocated block and a collocated picture.

A method of deriving motion information may change according to an inter prediction mode of a target block. For example, as inter prediction modes applied for inter prediction, an Advanced Motion Vector Predictor (AMVP) mode, a merge mode and a skip mode, a merge mode with motion vector difference, There may be sub-block merge mode, triangulation mode, inter-intra combined prediction mode, affine inter mode, and current picture reference mode. Merge mode may also be referred to as motion merge mode. Below, each of the modes is described in detail.

1) AMVP mode

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

1-1) Creating a predicted motion vector candidate list

When the AMVP mode is used as the prediction mode, each of the encoding apparatus 100 and the decoding apparatus 200 may generate a predictive motion vector candidate list using a motion vector of a spatial candidate, a motion vector of a temporal candidate, and a zero vector. there is. The predicted motion vector candidate list may include one or more predicted 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 predicted motion vector candidate.

Hereinafter, the terms “predicted motion vector (candidate)” and “motion vector (candidate)” may be used interchangeably and may be used interchangeably.

Hereinafter, the terms “predicted motion vector candidate” and “AMVP candidate” may be used interchangeably and may be used interchangeably.

Hereinafter, the terms “prediction motion vector candidate list” and “AMVP candidate list” may be used in the same meaning and may be used interchangeably.

Spatial candidates may include reconstructed spatial neighboring blocks. In other words, the motion vector of the reconstructed neighboring block may be referred to as a spatial prediction motion vector candidate.

Temporal candidates may include a collocated block and a block adjacent to the collocated block. In other words, a motion vector of a collocated block or a motion vector of a block adjacent to the collocated block may be referred to as a temporal prediction motion vector candidate.

The zero vector may be a (0, 0) motion vector.

The predictive motion vector candidate may be a motion vector predictor for motion vector prediction. Also, in the encoding apparatus 100, the predicted motion vector candidate may be an initial motion vector search position.

1-2) Motion vector search using a predictive motion vector candidate list

The encoding apparatus 100 may determine a motion vector to be used for encoding a target block within a search range by using the predictive motion vector candidate list. Also, the encoding apparatus 100 may determine a predictive motion vector candidate to be used as a predictive motion vector of the target block from among predictive motion vector candidates in the predictive motion vector candidate list.

A motion vector to be used for encoding a target block may be a motion vector that can be encoded with minimal cost.

Also, the encoding apparatus 100 may determine whether to use the AMVP mode in encoding the target block.

1-3) Transmission of inter prediction information

The encoding apparatus 100 may generate a bitstream including inter prediction information required for inter prediction. The decoding apparatus 200 may perform inter prediction on a target block using inter prediction information of a bitstream.

The inter-prediction information includes 1) mode information indicating whether the AMVP mode is used, 2) predictive motion vector index, 3) motion vector difference (MVD), 4) reference direction, and 5) reference picture index. can do.

Hereinafter, the terms “predicted motion vector index” and “AMVP index” may be used interchangeably and may be used interchangeably.

Also, the inter prediction information may include a residual signal.

When the mode information indicates that the AMVP mode is used, the decoding apparatus 200 may obtain a predicted motion vector index, a motion vector difference, a reference direction, and a reference picture index from a bitstream through entropy decoding.

The predictive motion vector index may indicate a predictive motion vector candidate used for prediction of a target block among predictive motion vector candidates included in the predictive motion vector candidate list.

1-4) Inter prediction of AMVP mode using inter prediction information

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

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

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

A motion vector actually used for inter prediction of a target block may not coincide with a predicted motion vector. MVD may be used to represent a difference between a motion vector actually used for inter prediction of a target block and a predicted motion vector. The encoding apparatus 100 may derive a predicted motion vector similar to a motion vector actually used for inter prediction of a target block in order to use an MVD having a size as small as possible.

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

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

The decoding apparatus 200 may derive the motion vector of the target block by adding 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.

Also, the encoding apparatus 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 apparatus 200 may extract entropy-encoded MVD resolution information from a bitstream, and obtain MVD resolution information by applying entropy decoding to the entropy-encoded MVD resolution information. The decoding apparatus 200 may adjust the resolution of the MVD using the MVD resolution information.

Meanwhile, the encoding device 100 may calculate the MVD based on the affine model. The decoding apparatus 200 may derive an affine control motion vector of the target block through the sum of the MVD and the affine control motion vector candidate, and may derive a motion vector for a subblock using the affine control motion vector. there is.

A reference direction may indicate a reference picture list used for prediction of a target block. For example, the reference direction may indicate one of a reference picture list L0 and a reference picture list L1.

The reference direction indicates only a reference picture list used for prediction of the target block, and may not indicate that directions of reference pictures are limited to a forward direction or a backward direction. In other words, each of the reference picture list L0 and the reference picture list L1 may include forward and/or backward pictures.

That the reference direction is uni-direction may mean that one reference picture list is used. Bi-direction of the reference direction may mean that two reference picture lists are used. In other words, the reference direction can indicate one of: that only the reference picture list L0 is used, that only the reference picture list L1 is used, and two reference picture lists.

The reference picture index may indicate a reference picture used for prediction of a target block among reference pictures of a reference picture list. The encoding apparatus 100 may generate an entropy-coded reference picture index by applying entropy encoding to the reference picture index, and may generate a bitstream including the entropy-coded reference picture index. The entropy-encoded reference picture index may be signaled from the encoding device 100 to the decoding device 200 through a bitstream. The decoding apparatus 200 may extract an entropy-encoded reference picture index from a bitstream and obtain the reference picture index by applying entropy decoding to the entropy-encoded reference picture index.

When two reference picture lists are used for prediction of a target block. One reference picture index and one motion vector may be used for each reference picture list. Also, when two reference picture lists are used for prediction of a target block, two prediction blocks may be specified for the target block. For example, a (final) prediction block of the target block may be generated through an average or a weighted-sum of two prediction blocks of the target block.

The motion vector of the target block may be derived by the predicted motion vector index, MVD, reference direction, and reference picture index.

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 a derived motion vector in a reference picture indicated by a reference picture index.

By encoding the predicted motion vector index and the MVD without encoding the motion vector of the target block, the amount of bits transmitted from the encoding apparatus 100 to the decoding apparatus 200 can be reduced and encoding efficiency can be improved.

Motion information of a neighboring block reconstructed with respect to the target block may be used. In a specific inter-prediction mode, the encoding apparatus 100 may not separately encode the motion information of the target block itself. Motion information of the target block is not encoded, and other information capable of inducing motion information of the target block through motion information of the reconstructed neighboring block may be encoded instead. As other information is encoded instead, the amount of bits transmitted to the decoding apparatus 200 can be reduced and encoding efficiency can be improved.

For example, as an inter prediction mode in which motion information of the target block is not directly encoded, there may be a skip mode and/or a merge mode. In this case, the encoding apparatus 100 and the decoding apparatus 200 may use an identifier and/or an index indicating which unit's motion information among reconstructed neighboring units is used as the motion information of the target unit.

2) Merge Mode

As a method of deriving motion information of a target block, there is a merge. Merge may refer to merging of motions of a plurality of blocks. Merge may mean applying motion information of one block to another block as well. In other words, merge mode may refer to a mode in which motion information of a target block is derived from motion information of neighboring blocks.

When merge mode is used, the encoding apparatus 100 may perform prediction of motion information of a target block by using motion information of a spatial candidate and/or motion information of a temporal candidate. The spatial candidate may include a reconstructed spatial neighboring block spatially adjacent to the target block. A spatial neighboring block may include a left neighboring block and an upper neighboring block. Temporal candidates may include call blocks. The terms "spatial candidate" and "spatial merge candidate" may be used interchangeably and may be used interchangeably. The terms "temporal candidate" and "temporal merge candidate" may be used interchangeably and may be used interchangeably.

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 a target block and a prediction block.

2-1) Creating a merge candidate list

When the merge mode is used, each of the encoding apparatus 100 and the decoding apparatus 200 may generate a merge candidate list using motion information of spatial candidates and/or motion information of temporal candidates. 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 mean an inter prediction indicator.

The merge candidate list may include merge candidates. A merge candidate may be motion information. In other words, the merge candidate list may be a list in which motion information is stored.

Merge candidates may be motion information such as temporal candidates and/or spatial candidates. In other words, the merge candidate list may include motion information such as a temporal candidate and/or a spatial candidate.

Also, the merge candidate list may include a new merge candidate generated by a combination of merge candidates already existing in the merge candidate list. In other words, the merge candidate list may include new motion information generated by combining motion information already existing in the merge candidate list.

Also, the merge candidate list may include a history-based merge candidate. A history-based merge candidate may be motion information of a block encoded and/or decoded prior to a target block.

Also, the merge candidate list may include a merge candidate based on an average of two merge candidates.

Merge candidates may be specified modes for deriving inter prediction information. A merge candidate may be information indicating a specific mode for deriving inter prediction information. Inter prediction information of a target block may be derived according to a specified mode indicated by a merge candidate. In this case, the specified mode may include a process of deriving a series of inter prediction information. This specified mode may be an inter prediction information derivation mode or a motion information derivation mode.

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

For example, the motion information derivation modes in the merge candidate list may be at least one of 1) a motion information derivation mode in units of sub-blocks and 2) an affine motion information derivation mode.

Also, the merge candidate list may include motion information of a zero vector. A zero vector may be referred to as a zero merge candidate.

In other words, the motion information in the merge candidate list is: 1) motion information of a spatial candidate, 2) motion information of a temporal candidate, 3) motion information generated by a combination of motion information already existing in the merge candidate list, 4) zero vector may be at least one of

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

The merge candidate list may be generated before prediction by merge mode is performed.

The number of merge candidates in the merge candidate list may be predefined. The encoding device 100 and the decoding device 200 may add merge candidates to the merge candidate list according to a predefined method and a predefined order so that the merge candidate list includes 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 the same through a predefined method and a predefined ranking.

Merge may be applied in units of CUs or units of PUs. When merging is performed in units of CUs or PUs, the encoding apparatus 100 may transmit a bitstream including predefined information to the decoding apparatus 200. For example, the predefined information includes 1) information indicating whether merging is to be performed for each block partition, 2) which block to merge with among blocks that are spatial candidates and/or temporal candidates for the target block. It may contain information on whether

2-2) Motion vector search using merge candidate list

The encoding apparatus 100 may determine a merge candidate to be used for encoding a target block. For example, the encoding apparatus 100 may perform predictions on a target block using merge candidates of the merge candidate list and generate residual blocks for the merge candidates. The encoding apparatus 100 may use a merge candidate that requires the least cost in encoding the prediction and residual blocks for encoding the target block.

Also, the encoding apparatus 100 may determine whether to use merge mode in encoding a target block.

2-3) Transmission of inter prediction information

The encoding apparatus 100 may generate a bitstream including inter prediction information required for inter prediction. The encoding apparatus 100 may generate entropy-encoded inter-prediction information by performing entropy encoding on the inter-prediction information, and may transmit a bitstream including the entropy-encoded inter-prediction information to the decoding apparatus 200. Entropy-encoded inter prediction information may be signaled from the encoding apparatus 100 to the decoding apparatus 200 through a bitstream. The decoding apparatus 200 may extract entropy-encoded inter prediction information from a bitstream and obtain inter prediction information by performing entropy decoding on the entropy-encoded inter prediction information.

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

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

Also, the inter prediction information may include a residual signal.

The decoding apparatus 200 may obtain the merge index from the bitstream only when the mode information indicates that the merge mode is used.

Mode information may be a merge flag. A unit of mode information may be a block. Information about a block may include mode information, and the mode information may indicate whether merge mode is applied to the block.

The merge index may indicate a merge candidate used for prediction of a target block among merge candidates included in the merge candidate list. Alternatively, the merge index may indicate which block among neighboring blocks that are spatially or temporally adjacent to the target block is merged with.

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

The correction information may be information used for motion vector correction. The encoding device 100 may generate correction information. The decoding apparatus 200 may correct the motion vector of the merge candidate selected by the merge index based on the correction information.

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

2-4) Inter prediction of merge mode using inter prediction information

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

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.

3) Skip Mode

The skip mode may be a mode in which motion information of a spatial candidate or motion information of a temporal candidate is applied to a target block as it is. Also, the skip mode may be a mode not using a residual signal. That is to say, when skip mode is used, the reconstructed block may be the same as the predicted block.

A difference between merge mode and skip mode may be transmission or use of a residual signal. In other words, skip mode may be similar to merge mode except that no residual signal is transmitted or used.

When the skip mode is used, the encoding apparatus 100 transmits information indicating which block's motion information among spatial or temporal candidate blocks is used as the motion information of the target block to the decoding apparatus 200 through a bitstream. can transmit The encoding apparatus 100 may generate entropy-encoded information by performing entropy encoding on such information, and may signal the entropy-encoded information to the decoding apparatus 200 through a bitstream. The decoding apparatus 200 may extract entropy-encoded information from a bitstream and obtain information by performing entropy decoding on the entropy-encoded information.

Also, when the skip mode is used, the encoding device 100 may not transmit other syntax element information such as MVD to the decoding device 200. For example, when the skip mode is used, the encoding apparatus 100 may not signal a syntax element related to at least one of the MVD, the coded block flag, and the transform coefficient level to the decoding apparatus 200.

3-1) Creation of merge candidate list

Skip mode can also use a merge candidate list. In other words, the merge candidate list can be used in both merge mode and skip mode. In this regard, the merge candidate list may be referred to as a "skip candidate list" or a "merge/skip candidate list".

Alternatively, skip mode may use a separate candidate list different from merge mode. In this case, in the description below, the merge candidate list and the merge candidate may be replaced with the skip candidate list and the skip candidate, respectively.

The merge candidate list may be generated before prediction by skip mode is performed.

3-2) Motion vector search using merge candidate list

The encoding apparatus 100 may determine a merge candidate to be used for encoding a target block. For example, the encoding apparatus 100 may perform predictions on a target block using merge candidates of a merge candidate list. The encoding apparatus 100 may use a merge candidate requiring a minimum prediction cost for encoding a target block.

Also, the encoding apparatus 100 may determine whether to use a skip mode in encoding a target block.

3-3) Transmission of inter prediction information

The encoding apparatus 100 may generate a bitstream including inter prediction information required for inter prediction. The decoding apparatus 200 may perform inter prediction on a target block using inter prediction information of a bitstream.

The inter-prediction information may include 1) mode information indicating whether a skip mode is used and 2) a skip index.

The skip index may be the same as the aforementioned merge index.

When the skip mode is used, the target block may be coded without a residual signal. Inter prediction information may not include a residual signal. Alternatively, the bitstream may not include a residual signal.

The decoding apparatus 200 may obtain the skip index from the bitstream only when the mode information indicates that the skip mode is used. As mentioned above, the merge index and skip index may be the same. The decoding apparatus 200 may obtain the skip index from the bitstream only when the mode information indicates that merge mode or skip mode is used.

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

3-4) Skip Mode Inter Prediction Using Inter Prediction Information

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

A motion vector of a target block may be specified by a motion vector of a merge candidate indicated by a skip index, a reference picture index, and a reference direction.

4) Current picture reference mode

The current picture reference mode may mean a prediction mode using a pre-reconstructed region in a target picture to which a target block belongs.

A motion vector may be used to specify the pre-reconstructed area. Whether the target block is encoded in the current picture reference mode can be determined using the reference picture index of the target block.

A flag or index indicating whether the target block is a block encoded in the current picture reference mode may be signaled from the encoding device 100 to the decoding device 200. Alternatively, whether the target block is a block coded in the current picture reference mode may be inferred through a reference picture index of the target block.

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

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

If the target picture exists at an arbitrary position in the reference picture list, a separate reference picture index indicating such an arbitrary position may be signaled from the encoding apparatus 100 to the decoding apparatus 200.

5) Subblock merge mode

A sub-block merge mode may mean a mode for deriving motion information for a sub-block of a CU.

When the sub-block merge mode is applied, motion information of a call sub-block of a target sub-block in a reference image (ie, a sub-block based temporal merge candidate) and/or an affine control point motion vector A subblock merge candidate list may be generated using an affine control point motion vector merge candidate.

6) Triangle partition mode

In the triangular division mode, divided target blocks may be generated by dividing the target block in a diagonal direction. For each divided target block, motion information of each divided target block may be derived, and prediction samples for each divided target block may be derived using the derived motion information. A prediction sample of the target block may be derived through a weighted sum of prediction samples of the divided target blocks.

7) Inter-intra combined prediction mode

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

In the aforementioned modes, the decoding apparatus 200 may perform self-correction on the derived motion information. For example, the decoding apparatus 200 may search a specific area based on a reference block indicated by the derived motion information to search for motion information having a minimum sum of absolute differences (SAD). and the searched motion information may be derived as corrected motion information.

In the aforementioned modes, the decoding apparatus 200 may perform compensation for prediction samples derived through inter prediction using an optical flow.

In the aforementioned AMVP mode, merge mode, and skip mode, motion information to be used for prediction of a target block among motion information in the list may be specified through an index of the list.

In order to improve encoding efficiency, the encoding apparatus 100 may signal only the index of an element that causes the least cost in inter prediction of a target block among elements in the list. The encoding device 100 may encode the index and signal the encoded index.

Accordingly, the aforementioned lists (ie, the predicted motion vector candidate list and the merge candidate list) may have to be derived in the same manner based on the same data in the encoding apparatus 100 and the decoding apparatus 200. Here, the same data may include a reconstructed picture and a reconstructed block. Also, in order to specify an element by index, the order of the elements within the list may need to be constant.

10 shows spatial candidates according to an example.

In Fig. 10, the positions of spatial candidates are shown.

A large block in the middle may represent a target block. Five small blocks may represent spatial candidates.

Coordinates of the target block may be (xP, yP), and the size of the target block may be (nPSW, nPSH).

The spatial candidate A ₀ may be a block adjacent to the lower left corner of the target block. A ₀ may be a block occupying pixels of coordinates (xP - 1, yP + nPSH).

Spatial candidate A ₁ may be a block adjacent to the left of the target block. A ₁ may be the lowest block among blocks adjacent to the left side of the target block. Alternatively, A ₁ may be a block adjacent to the top of A ₀ . A ₁ may be a block occupying pixels of coordinates (xP - 1, yP + nPSH - 1).

Spatial candidate B ₀ may be a block adjacent to the upper right corner of the target block. B ₀ may be a block occupying pixels of coordinates (xP + nPSW, yP - 1).

Spatial candidate B ₁ may be a block adjacent to the top of the target block. B ₁ may be the rightmost block among blocks adjacent to the top of the target block. Alternatively, B ₁ may be a block adjacent to the left side of B ₀ . B ₁ may be a block occupying pixels of coordinates (xP + nPSW - 1, yP - 1).

Spatial candidate B ₂ may be a block adjacent to the upper left corner of the target block. B ₂ may be a block occupying pixels of coordinates (xP - 1, yP - 1).

Determination of the availability of spatial and temporal candidates

In order to include the motion information of the spatial candidate or the motion information of the temporal candidate in the list, it must be determined whether the motion information of the spatial candidate or the motion information of the temporal candidate is available.

Hereinafter, a candidate block may include a spatial candidate and a temporal candidate.

For example, the above determination may be made by sequentially applying steps 1) to 4) below.

Step 1) If the PU including the candidate block is outside the boundary of the picture, the availability of the candidate block may be set to false. "Availability is set to false" may mean the same as "availability is set to unavailability".

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

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

Step 4) If the prediction mode of the PU including the candidate block is an intra prediction mode, availability of the candidate block may be set to false. If the PU containing the candidate block does not use inter prediction, the availability of the candidate block may be set to false.

11 illustrates an order of adding motion information of spatial candidates to a merge list according to an example.

As shown in FIG. 11 , in adding motion information of spatial candidates to the merge list, the order of A ₁ , B ₁ , B ₀ , A ₀ and B ₂ may be used. That is, in the order of A ₁ , B ₁ , B ₀ , A ₀ and B ₂ , motion information of available spatial candidates may be added to the merge list.

Merge list derivation method in merge mode and skip mode

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

Motion information (ie, merge candidates) may be added to the merge list in the order of steps 1) to 4) below.

Step 1) Available spatial candidates among spatial candidates may be added to the merge list. Motion information of available spatial candidates may be added to the merge list in the order shown in FIG. 11 . In this case, if motion information of an available spatial candidate overlaps with other motion information already existing in the merge list, the motion information may not be added to the merge list. Checking whether it overlaps with other motion information existing in the list can be abbreviated as "redundancy check".

A maximum of N pieces of motion information may be added.

Step 2) If the number of pieces of motion information in the merge list is smaller than N and a temporal candidate is available, the motion information of the temporal candidate may be added to the merge list. In this case, if motion information of an available temporal candidate overlaps with other motion information already existing in the merge list, the motion information may not be added to the merge list.

Step 3) If the number of pieces of motion information in the merge list is smaller than N and the type of the target slice is “B”, the combined motion information generated by combined bi-prediction is added to the merge list. can

The target slice may be a slice including the target block.

The combined motion information may be a combination of L0 motion information and L1 motion information. The L0 motion information may be motion information referring only to the reference picture list L0. The L1 motion information may be motion information referring only to the reference picture list L1.

Within the merge list, there may be one or more L0 motion information. Also, within the merge list, there may be one or more L1 motion information.

The combined motion information may be one or more. In generating the combined motion information, which L0 motion information and which L1 motion information among one or more pieces of L0 motion information and one or more pieces of L1 motion information are to be used may be predefined. One or more pieces of combined motion information may be generated in a predefined order by combined bidirectional prediction using pairs of different pieces of motion information in a merge list. One of the pairs of different motion information may be L0 motion information and the other may be L1 motion information.

For example, the combined motion information added with the highest priority may be a combination of L0 motion information having a merge index of 0 and L1 motion information having a merge index of 1. If the motion information with a merge index of 0 is not L0 motion information or the motion information with a merge index of 1 is not L1 motion information, the combined motion information may not be generated and added. Motion information added next may be a combination of L0 motion information having a merge index of 1 and L1 motion information having a merge index of 0. The following specific combinations may follow other combinations in the field of encoding/decoding of video.

In this case, if the combined motion information overlaps with other motion information already existing in the merge list, the combined motion information may not be added to the merge list.

Step 4) If the number of pieces of motion information in the merge list is smaller than N, zero vector motion information may be added to the merge list.

The zero vector motion information may be motion information in which a motion vector is a zero vector.

One or more zero vector motion information may be provided. Reference picture indexes of one or more zero vector motion information may be different from each other. For example, the value of the reference picture index of the first zero vector motion information may be 0. The value of the reference picture index of the second zero vector motion information may be 1.

The number of zero vector motion information may be equal to the number of reference pictures in the reference picture list.

The reference direction of the zero vector motion information may be bidirectional. Both motion vectors may be zero vectors. The number of zero vector motion information may be the smaller of the number of reference pictures in the reference picture list L0 and the number of reference pictures in the reference picture list L1. Alternatively, when the number of reference pictures in the reference picture list L0 and the number of reference pictures in the reference picture list L1 are different from each other, a unidirectional reference direction may be used for a reference picture index applicable to only one reference picture list.

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

If the zero vector motion information overlaps with other motion information already present in the merge list, the zero vector motion information may not be added to the merge list.

The order of steps 1) to 4) described above is merely exemplary, and the order between steps may be interchanged. Also, some of the steps may be omitted according to predefined conditions.

Derivation method of predictive motion vector candidate list in AMVP mode

The maximum number of motion vector predictor candidates in the predictor motion vector candidate list may be predefined. The predefined maximum number is denoted by N. For example, the predefined maximum number may be 2.

Motion information (ie, predicted motion vector candidates) may be added to the predicted motion vector candidate list in the order of steps 1) to 3) below.

Step 1) Among the spatial candidates, available spatial candidates may be added to the predicted motion vector candidate list. Spatial candidates may include a first spatial candidate and a second spatial candidate.

The first spatial candidate may be one of A ₀ , A ₁ , scaled A ₀ and scaled A ₁ . The second spatial candidate may be one of B ₀ , B ₁ , B ₂ , scaled B ₀ , scaled B ₁ , and scaled B ₂ .

Motion information of available spatial candidates may be added to the predicted motion vector candidate list in the order of the first spatial candidate and the second spatial candidate. In this case, when motion information of an available spatial candidate overlaps with other motion information already existing in the motion vector predictor candidate list, the motion information may not be added to the predictor motion vector candidate list. In other words, when the value of N is 2, if the motion information of the second spatial candidate is identical to the motion information of the first spatial candidate, the motion information of the second spatial candidate may not be added to the predicted motion vector candidate list.

A maximum of N pieces of motion information may be added.

Step 2) If the number of pieces of motion information in the predicted motion vector candidate list is smaller than N and a temporal candidate is available, the motion information of the temporal candidate may be added to the predicted motion vector candidate list. In this case, when motion information of an available temporal candidate overlaps with other motion information already existing in the motion vector predictor candidate list, the motion information may not be added to the predictor motion vector candidate list.

Step 3) If the number of pieces of motion information in the predicted motion vector candidate list is smaller than N, zero vector motion information may be added to the predicted motion vector candidate list.

One or more zero vector motion information may be provided. Reference picture indexes of one or more zero vector motion information may be different from each other.

The encoding apparatus 100 and/or the decoding apparatus 200 may sequentially add zero vector motion information to the predictive motion vector candidate list while changing the reference picture index.

When the zero vector motion information overlaps with other motion information already existing in the predicted motion vector candidate list, the zero vector motion information may not be added to the predicted motion vector candidate list.

The description of the zero vector motion information described above for the merge list can also be applied to the zero vector motion information. Redundant descriptions are omitted.

The order of steps 1) to 3) described above is merely exemplary, and the order between steps may be interchanged. Also, some of the steps may be omitted according to predefined conditions.

12 illustrates a process of transformation and quantization according to an example.

As shown in FIG. 12, a quantized level may be generated by performing a transform and/or quantization process on the residual signal.

A residual signal may be generated as a difference between an original block and a prediction block. Here, the prediction block may be a block generated by intra prediction or inter prediction.

The residual signal may be converted into the frequency domain through a transform process that is part of the quantization process.

Transformation kernels used for transformation may include various DCT kernels such as Discrete Cosine Transform (DCT) type 2 (DCT-II) and Discrete Sine Transform (DST) kernels. .

These transform kernels may perform a separable transform or a 2D (2D) non-separable transform on the residual signal. The separable transform may be a transform that performs a one-dimensional (1D) transform on the residual signal in each of a horizontal direction and a vertical direction.

DCT type and DST type adaptively used for 1D conversion may include DCT-V, DCT-VIII, DST-I and DST-VII in addition to DCT-II as shown in Table 3 and Table 4 below, respectively. there is.

[Table 3]

[Table 4]

As shown in Tables 3 and 4, a transform set may be used in deriving the DCT type or DST type to be used for transformation. Each transform set may include a plurality of transform candidates. Each transformation candidate may be a DCT type or a DST type.

Table 5 below shows an example of a transform set applied in the horizontal direction and a transform set applied in the vertical direction according to the intra prediction mode.

[Table 5]

In Table 5, vertical transform set numbers and horizontal transform set numbers applied to the horizontal direction of the residual signal are indicated according to the intra prediction mode of the target block.

As illustrated in Table 5, transform sets 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 a transform included in a transform set corresponding to the intra prediction mode of the target block. Also, the decoding apparatus 200 may perform an inverse transform on the residual signal using a transform included in a transform set corresponding to the intra prediction mode of the target block.

In these transforms and inverse transforms, the set of transforms applied to the residual signal may be determined as exemplified in Tables 3, 4, and 5, and may not be signaled. Transformation indication information may be signaled from the encoding device 100 to the decoding device 200. The transform indication information may be information indicating which transform candidate among a plurality of transform candidates included in a transform set applied to the residual signal is used.

For example, when the size of the target block is 64x64 or less, transform sets each having three transforms may be configured according to the intra prediction mode. An optimal transform method can be selected among all 9 multiple transform methods resulting from a combination of three transforms in the horizontal direction and three transforms in the vertical direction. Encoding efficiency can be improved by encoding and/or decoding the residual signal using such an optimal conversion method.

In this case, information on which transform among transforms belonging to the transform set is used for at least one of the vertical transform and the horizontal transform may be entropy encoded and/or decoded. A truncated unary binarization may be used to encode and/or decode this information.

As described above, a method using various transforms may be applied to a residual signal generated by intra prediction or inter prediction.

Transformation may include at least one of a primary transformation and a secondary transformation. A transform coefficient may be generated by performing a primary transform on the residual signal, and a secondary transform coefficient may be generated by performing a secondary transform on the transform coefficient.

A primary transformation may be named primary. Also, the primary transform may be referred to as an adaptive multiple transform (AMT). As described above, AMT may mean that different transforms are applied to each of the 1D directions (ie, the vertical direction and the horizontal direction).

The secondary transform may be a transform for improving the energy concentration of the transform coefficient generated by the primary transform. Like the first-order transformation, the second-order transformation may be a separable transformation or a non-separable transformation. The non-separable transform may be a non-separable secondary transform (NSST).

The primary transformation may be performed using at least one of a plurality of predefined transformation methods. For example, a plurality of predefined transform methods include a Discrete Cosine Transform (DCT), a Discrete Sine Transform (DST), and a Karhunen-Loeve Transform (KLT) based transform. can include

Also, the primary transform may be a transform having various transform types according to a kernel function defining DCT or DST.

For example, the transform type includes: 1) prediction mode of the target block (eg, one of intra prediction and inter prediction), 2) size of the target block, 3) shape of the target block, 4) intra prediction mode of the target block , 5) components of the target block (eg, one of a luma component and a chroma component) and 6) a partition type applied to the target block (eg, Quad Tree (QT), Binary Tree (BT)). ) and one of a ternary tree (TT)).

For example, the primary transform includes transforms such as DCT-2, DCT-5, DCT-7, DST-7, DST-1, DST-8 and DCT-8 according to the transform kernel presented in Table 6 below. can do. In Table 6, various transform types and transform kernel functions for Multiple Transform Selection (MTS) are illustrated.

MTS may mean that a combination of one or more DCT and/or DST transform kernels is selected for horizontal and/or vertical transform of the residual signal.

[Table 6]

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

A secondary transform may be performed on transform coefficients generated by performing the primary transform.

As in the first-order transform, a set of transforms can be defined in the second-order transform. Methods for deriving and/or determining a set of transforms, such as those described above, may be applied to first-order as well as second-order transforms.

A primary transform and a secondary transform can be determined for a specified object.

For example, the first transform and the second transform may be applied to one or more signal components of a luma component and a chroma component. Whether to apply the primary transform and/or the secondary transform may be determined according to at least one of coding parameters for a target block and/or a neighboring block. For example, whether to apply the first transform and/or the second transform may be determined by the size and/or shape of the target block.

In the encoding device 100 and the decoding device 200, transformation information indicating a transformation method used for a target may be derived by using specified information.

For example, the transformation information may include an index of a transformation to be used for primary transformation and/or secondary transformation. Alternatively, the transform information may indicate that the primary transform and/or the secondary transform are not used.

For example, when the target of the primary transform and the secondary transform is a target block, the transform method(s) applied to the primary transform and/or the secondary transform indicated by transform information may be applied to the target block and/or neighboring blocks. It may be determined according to at least one of the coding parameters for

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

For example, whether or not a primary transform is used for one CU, an index indicating the primary transform, whether or not a secondary transform is used, and an index indicating the secondary transform may be derived as transformation information in the decoding apparatus 200 there is. Alternatively, conversion information indicating whether or not the primary transform is used for one CU, an index indicating the primary transform, whether or not a secondary transform is used, and an index indicating the secondary transform may be signaled.

A quantized transform coefficient (ie, a quantized level) may be generated by performing quantization on a residual signal or a result generated by performing the primary transform and/or the secondary transform.

13 illustrates diagonal scanning according to an example.

14 illustrates horizontal scanning according to an example.

15 illustrates vertical scanning according to an example.

The quantized transform coefficients may be scanned according to at least one of (up-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. A block may be a transform unit.

Each scanning can start at a specified start point and end at a specified end point.

For example, quantized transform coefficients may be changed into a one-dimensional vector form by scanning coefficients of a block using the diagonal scanning of FIG. 13 . Alternatively, the horizontal scanning of FIG. 14 or the vertical scanning of FIG. 15 may be used instead of diagonal scanning according to the block size and/or intra prediction mode.

Vertical scanning may be scanning two-dimensional block form coefficients in a column direction. Horizontal scanning may be scanning two-dimensional block form coefficients in a row direction.

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

As shown in FIGS. 13, 14, and 15, quantized transform coefficients may be scanned along a diagonal, horizontal, or vertical direction.

Quantized transform coefficients may be expressed in block form. A block may include a plurality of sub-blocks. Each sub-block may be defined according to a minimum block size or a minimum block shape.

In scanning, a scanning order according to a type or direction of scanning may be applied to subblocks first. Also, a scanning order according to a scanning direction may be applied to quantized transform coefficients in a sub-block.

For example, as shown in FIGS. 13, 14, and 15, when the size of the target block is 8x8, transform coefficients quantized by primary transform, secondary transform, and quantization of the residual signal of the target block are can be created Thereafter, one of three scanning orders may be applied to the four 4x4 subblocks, and quantized transform coefficients may be scanned according to the scanning order for each 4x4 subblock.

The encoding apparatus 100 may generate an entropy-encoded quantized transform coefficient by performing entropy encoding on the scanned quantized transform coefficients, and may generate a bitstream including the entropy-coded quantized transform coefficients. .

The decoding apparatus 200 may extract entropy-encoded quantized transform coefficients from a bitstream and generate quantized transform coefficients by performing entropy decoding on the entropy-coded quantized transform coefficients. Quantized transform coefficients may be arranged in a 2D block form through inverse scanning. At this time, as a reverse scanning method, at least one of (upper right) diagonal scan, vertical scan, and horizontal scan may be performed.

In the decoding apparatus 200, inverse quantization may be performed on quantized transform coefficients. Depending on whether the second-order inverse transform is performed, the second-order inverse transform may be performed on the result generated by performing the inverse quantization. Also, depending on whether the first-order inverse transform is performed, the first-order inverse transform may be performed on the result generated by performing the second-order inverse transform. A reconstructed residual signal may be generated by performing a first-order inverse transform on a result generated by performing a second-order inverse transform.

For a luma component reconstructed through intra prediction or inter prediction, inverse mapping of a dynamic range may be performed before in-loop filtering.

The dynamic range may be divided into 16 equal pieces, and a mapping function for each piece may be signaled. The mapping function may be signaled at the slice level or tile group level.

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

In-loop filtering, reference picture storage, and motion compensation may be performed in the inversely mapped region.

A prediction block generated through inter prediction may be converted into a mapped region by mapping using a mapping function, and the converted prediction block may be used to generate a reconstructed block. However, since intra prediction is performed in a mapped region, a prediction block generated by intra prediction can be used to generate a reconstructed block without mapping and/or inverse mapping.

For example, when the target block is a residual block of a chroma component, the residual block may be converted into an inversely mapped region by scaling the chroma component of the mapped region.

Whether scaling is available may be signaled at the slice level or tile group level.

For example, scaling can be applied only when mapping for luma components is available, and partitioning of luma components and partitioning of chroma components follow the same tree structure.

Scaling may be performed based on an average of values of samples of a luma prediction block corresponding to the chroma prediction block. In this case, when the target block uses inter prediction, the luma prediction block may mean a mapped luma prediction block.

A value required for scaling may be derived by referring to a look-up table using an index of a piece to which an average of values of samples of the luma prediction block belongs.

By performing scaling on the residual block using the finally derived value, the residual block may be converted into an inversely mapped region. Then, for the chroma component block, reconstruction, intra prediction, inter prediction, in-loop filtering, and reference picture storage may be performed in the inversely mapped region.

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

A prediction block of the target block may be generated based on the block vector. A block vector may indicate displacement between a target block and a reference block. A reference block may be a block within a target image.

In this way, a prediction mode in which a prediction block is generated by referring to a target image may be referred to as an intra block copy (IBC) mode.

IBC mode can be applied to CUs of a specified size. For example, IBC mode can be applied to MxN CUs. Here, M and N may be 64 or less.

The IBC mode may include a skip mode, a merge mode, and an AMVP mode. In the case of skip mode or merge mode, a merge candidate list may be constructed, and a merge index may be signaled so that one merge candidate among merge candidates of the merge candidate list may be specified. A block vector of a specified merge candidate may be used as a block vector of a target block.

In case of AMVP mode, a differential block vector may be signaled. Also, the predicted block vector may be derived from the left neighboring block and the top neighboring block of the target block. Also, an index of which neighboring block is to be used may be signaled.

The prediction block of the IBC mode may be included in the target CTU or the left CTU, and may be limited to a block within a pre-constructed region. For example, the value of the block vector may be limited so that the prediction block of the target block is located within a specified region. The specified area may be an area of three 64x64 blocks that are encoded and/or decoded prior to the 64x64 block including the target block. As the value of the block vector is limited in this way, memory consumption and device complexity according to the implementation of the IBC mode can be reduced.

16 is a structural diagram of an encoding device according to an embodiment.

The encoding device 1600 may correspond to the aforementioned encoding device 100.

The encoding device 1600 includes a processing unit 1610, a memory 1630, a user interface (UI) input device 1650, a UI output device 1660, and storage that communicate with each other through a bus 1690. (1640). Also, the encoding device 1600 may further include a communication unit 1620 connected to the network 1699.

The processor 1610 may be a semiconductor device that executes processing instructions stored in a central processing unit (CPU), memory 1630, or storage 1640. The processing unit 1610 may be at least one hardware processor.

The processing unit 1610 may generate and process signals, data, or information input to the encoding device 1600, output from the encoding device 1600, or used inside the encoding device 1600, signals, Inspection, comparison, and judgment related to data or information can be performed. In other words, in an embodiment, data or information generation and processing, and data or information related inspection, comparison, and judgment may be performed by the processing unit 1610 .

The processing unit 1610 includes an inter prediction unit 110, an intra prediction unit 120, a switch 115, a subtractor 125, a transform unit 130, a quantization unit 140, an entropy encoding unit 150, and an inverse quantization unit. It may include a unit 160, an inverse transform unit 170, an adder 175, a filter unit 180, and a reference picture buffer 190.

An inter prediction unit 110, an intra 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, At least some of 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. Program modules may be included in the encoding device 1600 in the form of an operating system, application program modules, and other program modules.

Program modules may be physically stored on various known storage devices. Also, at least some of these program modules may be stored in a remote storage device capable of communicating with the encoding device 1600 .

Program modules include routines, subroutines, programs, objects, components, and data that perform functions or operations according to an embodiment or implement abstract data types according to an embodiment. A data structure, etc. may be encompassed, but is not limited thereto.

Program modules may include instructions or codes executed by at least one processor of the encoding device 1600 .

The processing unit 1610 includes an inter prediction unit 110, an intra prediction unit 120, a switch 115, a subtractor 125, a transform unit 130, a quantization unit 140, an entropy encoding unit 150, and an inverse quantization unit. Instructions or codes of the unit 160, the inverse transform unit 170, the adder 175, the filter unit 180, and the reference picture buffer 190 may be executed.

Storage may represent memory 1630 and/or storage 1640 . Memory 1630 and storage 1640 may be various forms of volatile or non-volatile storage media. For example, the memory 1630 may include at least one of a ROM 1631 and a RAM 1632 .

The storage unit may store data or information used for the operation of the encoding device 1600 . In an embodiment, data or information of the encoding device 1600 may be stored in a storage unit.

For example, the storage unit may store pictures, blocks, lists, motion information, inter prediction information, and bitstreams.

The encoding device 1600 may be implemented in a computer system including a recording medium that can be read by a computer.

The recording medium may store at least one module required for the encoding device 1600 to operate. The memory 1630 may store at least one module, and the at least one module may be configured to be executed by the processing unit 1610 .

A function related to communication of data or information of the encoding device 1600 may be performed through the communication unit 1620 .

For example, the communication unit 1620 may transmit a bitstream to a decoding apparatus 1700 to be described later.

17 is a structural diagram of a decryption device according to an embodiment.

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

The decryption device 1700 includes a processing unit 1710, a memory 1730, a user interface (UI) input device 1750, a UI output device 1760, and storage that communicate with each other through a bus 1790. (1740). Also, the decoding apparatus 1700 may further include a communication unit 1720 connected to the network 1799.

The processor 1710 may be a semiconductor device that executes processing instructions stored in a central processing unit (CPU), memory 1730, or storage 1740. The processing unit 1710 may be at least one hardware processor.

The processing unit 1710 may generate and process signals, data, or information input to the decoding device 1700, output from the decoding device 1700, or used inside the decoding device 1700, and signals, Inspection, comparison, and judgment related to data or information can be performed. In other words, in an embodiment, data or information generation and processing, and data or information related inspection, comparison, and judgment may be performed by the processing unit 1710 .

The processing unit 1710 includes 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 A unit 260 and a reference picture buffer 270 may be included.

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 At least some of the reference picture buffers 270 may be program modules, and may communicate with an external device or system. Program modules may be included in the decryption device 1700 in the form of an operating system, application program modules, and other program modules.

Program modules may be physically stored on various known storage devices. Also, at least some of these program modules may be stored in a remote storage device capable of communicating with the decryption device 1700 .

Program modules include routines, subroutines, programs, objects, components, and data that perform functions or operations according to an embodiment or implement abstract data types according to an embodiment. A data structure, etc. may be encompassed, but is not limited thereto.

The program modules may include instructions or codes executed by at least one processor of the decoding device 1700 .

The processing unit 1710 includes 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 Instructions or codes of unit 260 and reference picture buffer 270 may be executed.

Storage may represent memory 1730 and/or storage 1740 . Memory 1730 and storage 1740 may be various forms of volatile or non-volatile storage media. For example, the memory 1730 may include at least one of a ROM 1731 and a RAM 1732 .

The storage unit may store data or information used for the operation of the decoding device 1700 . In an embodiment, data or information of the decoding device 1700 may be stored in the storage unit.

For example, the storage unit may store pictures, blocks, lists, motion information, inter prediction information, and bitstreams.

The decryption device 1700 may be implemented in a computer system including a recording medium that can be read by a computer.

The recording medium may store at least one module required for the decoding apparatus 1700 to operate. The memory 1730 may store at least one module, and the at least one module may be configured to be executed by the processing unit 1710 .

A function related to communication of data or information of the decoding device 1700 may be performed through the communication unit 1720 .

For example, the communication unit 1720 may receive a bitstream from the encoding device 1600.

Hereinafter, the processing unit may represent the processing unit 1610 of the encoding device 1600 and/or the processing unit 1710 of the decoding device 1700. For example, for functions related to prediction, the processing unit may represent switch 115 and/or switch 245. In the function related to inter prediction, the processing unit may represent the inter prediction unit 110, the subtractor 125, and the adder 175, and may represent the inter prediction unit 250 and the adder 255. In the function related to intra prediction, the processing unit may represent the intra prediction unit 120, the subtractor 125, and the adder 175, and may represent the intra prediction unit 240 and the adder 255. In the function related to transform, the processing unit may represent the transform unit 130 and the inverse transform unit 170, and may represent the inverse transform unit 230. In the function related to quantization, the processing unit may represent the quantization unit 140 and the inverse quantization unit 160, and may represent the inverse quantization unit 220. In the function related to entropy encoding and/or decoding, the processing unit may represent the entropy encoding unit 150 and/or the entropy decoding unit 210. In terms of filtering-related functions, the processing unit may represent the filter unit 180 and/or the filter unit 260. For functions related to reference pictures, the processing unit may indicate the reference picture buffer 190 and/or the reference picture buffer 270 .

In a typical video encoding/decoding method, a decoder-side motion information derivation method can be used in a limited manner. Therefore, improvement in coding efficiency due to the decoder-end motion information derivation method may be limited.

In embodiments, an encoding/decoding method, apparatus, and recording medium using a motion information search method may be provided in order to improve encoding efficiency in inter prediction.

The processing unit may perform inter prediction on the target block. In this case, the processing unit may derive motion information of the target block from available information when prediction of the target block is performed through the motion information search method.

Encoding efficiency can be improved by minimizing bits used for signaling of coding information required for decoding a target block through the derivation of such motion information.

In embodiments, the coding information may include information required for decoding the target block (transmitted from the encoding device 1600 to the decoding device 1700) described in the embodiments. For example, coding information may include coding parameters.

Information described as being signaled through a bitstream in embodiments may be included in coding information. Also, coding information may be signaled through a bitstream.

In embodiments, available information may refer to information that has already been decoded (or reconstructed) before prediction of the target block is performed.

In embodiments, the available information is a coding parameter, motion information, predicted sample information, reconstructed sample information, and in-loop filtering-performed decoded information at a specific sample position in the target picture. At least one of sample information may be included.

In embodiments, the available information is at least one of coding parameters, motion information, predicted sample information, reconstructed sample information, and decoded sample information on which in-loop filtering is performed of a specific block including a specific sample position in a target picture. may contain one.

In embodiments, a location of a specific sample may refer to a location of a neighboring sample adjacent to the left side, upper side, and/or upper left side of a target block.

In embodiments, the available information is at least one of a coding parameter at a specific sample position in a reference picture of a target block, motion information, predicted sample information, reconstructed sample information, and reconstructed sample information on which in-loop filtering is performed. may contain one.

In embodiments, the available information includes a coding parameter, motion information, predicted sample information, reconstructed sample information, and reconstructed information on which in-loop filtering has been performed. At least one of sample information may be included.

For example, a specific sample location may be indicated from motion information of a target block.

In embodiments, the available information is a coding parameter, motion information, prediction sample information, reconstructed sample information, and a reconstructed sample for which in-loop filtering has been performed at a specific sample position in a col picture of a target block. At least one of the information may be included.

In embodiments, the available information is a coding parameter, motion information, prediction sample information, reconstructed sample information, and a reconstructed sample on which in-loop filtering is performed At least one of the information may be included.

For example, a specific sample location may be indicated from motion information of a target block.

For example, the location of a specific sample may refer to a location of a neighboring sample adjacent to the left side, upper side, and/or upper left side of a collocated block.

Embodiments for implementing the motion information search method are described below.

18 is a flowchart of a method of predicting a target block and generating a bitstream according to an embodiment.

The prediction method of the target block and the bitstream generation method of the embodiment may be performed by the encoding device 1600 . An embodiment may be a part of a method of encoding a target block or a method of video encoding.

Prediction may be one of the prediction methods described above in embodiments. For example, prediction may be inter prediction or intra prediction.

In step 1810, the processing unit 1610 may determine prediction information to be used for encoding of the target block.

Prediction information may include information used for prediction described in the embodiments.

For example, the prediction information may include inter prediction information. For example, the prediction information may include intra prediction information.

For example, the prediction information may include a motion information candidate list and final motion information to be described later. Determination of prediction information may include configuration of a motion information candidate list and determination of final motion information.

In step 1820, encoding of the coding information may be performed to generate coded coding information.

Coding information may refer to signaling/encoding/decoding information described in the embodiments. In other words, the coding information may be information used for the decoding apparatus 1700 to perform prediction corresponding to the prediction performed by the encoding apparatus 1600 .

At step 1830, processing unit 1610 may generate a bitstream.

A bitstream may include information about a target block. Also, the bitstream may include the information described above in the embodiments.

For example, a bitstream may include coded coding information or coding information.

For example, the bitstream may include coding parameters related to a target block and/or attributes of the target block.

Information included in the bitstream may be generated at step 1820, or at least partially at steps 1810 and 1820.

The processing unit 1610 may store the generated bitstream in the storage 1640 . Alternatively, the communication unit 1620 may transmit the bitstream to the decoding device 1700.

A bitstream may include coded information about a target block. The processing unit 1610 may generate encoded information on the target block by performing entropy encoding on the information on the target block.

In step 1840, the processing unit 1610 may perform prediction on the target block using information on the target block and prediction information.

The processing unit 1610 may use coding information in predicting a target block. Alternatively, coding information may be generated to correspond to information used in prediction of the target block.

A prediction block may be generated by predicting the target block. A residual block that is the difference between the target block and the predicted block may be generated. Information on the target block may be generated by applying transform and quantization to the residual block.

Information on the target block may include transform and quantized coefficients of the target block. A reconstructed residual block may be generated by applying inverse quantization and inverse transformation to transform and quantized coefficients of the target block. A reconstructed block that is the sum of the prediction block and the reconstructed residual block may be generated.

19 is a flowchart of a method of predicting a target block using a bitstream according to an embodiment.

A method of predicting a target block using the bitstream of the embodiment may be performed by the decoding apparatus 1700. The embodiment may be part of a decoding method of a target block or a video decoding method.

Prediction may be one of the prediction methods described above in embodiments. For example, prediction may be inter prediction or intra prediction.

In step 1910, the communication unit 1720 may obtain a bitstream. The communication unit 1720 may receive a bitstream from the encoding device 1600. The processing unit 1710 may store the obtained bitstream in the storage 1740 .

The processing unit 1710 may read a bitstream from the storage unit 1740 .

A bitstream may include information about a target block.

Information on the target block may include transform and quantized coefficients of the target block.

Also, the bitstream may include the information described above in the embodiments.

For example, a bitstream may include coded coding information or coding information.

For example, the bitstream may include coding parameters related to a target block and/or attributes of the target block.

A computer-readable recording medium may include a bitstream, and prediction and decoding of a target block may be performed using information on the target block included in the bitstream.

The computer readable recording medium may be a non-transitory computer readable recording medium.

A bitstream may include coded information about a target block. The processing unit 1710 may generate information on the target block by performing entropy decoding on the encoded information on the target block.

In step 1920, the processing unit 1710 may obtain coding information from the bitstream.

The processing unit 1710 may generate coding information by performing decoding on the encoded coding information of the bitstream.

Coding information may refer to signaling/encoding/decoding information described in the embodiments. In other words, the coding information may be information used for the decoding apparatus 1700 to perform prediction corresponding to the prediction performed by the encoding apparatus 1600 .

In step 1930, the processing unit 1710 may determine prediction information to be used for decoding the target block.

Prediction information may include information used for prediction described in the embodiments.

The prediction information in the decoding device 1700 may be the same as the prediction information in the encoding device 1600. In other words, processing unit 1710 may generate prediction information identical to the prediction information used in step 1840 to make the same prediction as the prediction performed in step 1840 .

For example, the prediction information may include inter prediction information. For example, the prediction information may include intra prediction information.

For example, the prediction information may include a motion information candidate list and final motion information to be described later. Determination of prediction information may include configuration of a motion information candidate list and determination of final motion information.

Processing unit 1710 may determine prediction information using the methods used in the embodiments.

The processing unit 1710 may determine prediction information of the target block based on information related to a prediction method obtained from the bitstream.

Prediction information may include inter prediction information. Prediction information may include intra prediction information.

In operation 1940, the processing unit 1710 may perform prediction on the target block using information on the target block and prediction information.

The processing unit 1710 may use coding information in predicting the target block. A prediction block may be generated by predicting the target block.

Information on the target block may include transform and quantized coefficients of the target block. A reconstructed residual block may be generated by applying inverse quantization and inverse transformation to transform and quantized coefficients of the target block. A reconstructed block that is the sum of the prediction block and the reconstructed residual block may be created.

Intra Block Copy (IBC)

The intra block copy mode may refer to a mode in which an area indicated by a block vector of a target block is used as a prediction block of the target block.

The target block may be encoded/decoded in one of intra prediction, inter prediction, and intra block copy modes.

An encoding/decoding method based on prediction using intra block copy is 1) when a luma component and a chroma component have an independent block division structure (ie, a dual tree structure is used). case) and 2) when the luma component and the chroma component have the same block division structure (ie, when a single tree structure is used).

Intra Block Copy mode is a block (eg, reference block or prediction block) from a previously coded/decoded region in a target image using a derived block vector (BV). may be a way to induce

The target image may be an image including the target block. Here, since a block is derived from an image including a target block, intra block copying may correspond to intra prediction.

A block vector may mean an intra block vector.

The previously encoded/decoded area may be a reconstructed image of the target picture or an area within a decoded image. Here, an area within a reconstructed image may mean a reconstructed area. An area within a decoded image may mean a decoded area.

A previously encoded/decoded region in the target image may be a reconstructed region to which at least one of the in-loop filtering is not applied.

In embodiments, in-loop filtering is performed by 1) chroma scaling and luma mapping, 2) deblocking filtering, Adaptive Sample Offset (ASO) and adaptive (adaptive) may include in-loop filtering.

In embodiments, a previously encoded/decoded region within a target image may be a reconstructed/decoded region in which at least one of in-loop filtering is performed.

Adaptive Motion Vector Resolution (AMVR)

In embodiments, the term "resolution" may mean the term "motion vector resolution".

In the adaptive motion vector resolution, the resolution of the motion vector difference may be adjusted in units of blocks.

The adaptive motion vector resolution information may represent the resolution of the motion vector difference. The resolution of the motion vector differential for the target block may be determined through signaling/encoding/decoding of adaptive motion vector resolution information.

Motion vector resolutions applicable to blocks may be the same or different.

For example, resolutions of motion vectors applicable to the target block may be determined based on at least one of a coding parameter, motion information, and mode information of the target block.

Adaptive motion vector resolution can improve coding efficiency by adjusting the resolution of motion vector differences.

For example, the scaled resolution can be one of 16-pels, 8-pels, 4-pels, full-pels, half-pels and quarter-pels, as described above. It is not limited to the listed pels.

When the adjusted resolution is n-pels, if the value of the motion vector difference component changes by 1, the position indicated by the motion vector difference may change by n pixel(s). In other words, if the adjusted resolution of the target block is n-pels, each component of the motion vector difference may indicate a reference block in units of n pixels.

For example, if the motion vector difference to be actually applied to the target block is (a, b) and the adjusted resolution is p-pel, (a/p, b/p) rather than (a, b) will be encoded. can That is, the motion vector difference signaled/encoded in the encoding apparatus 1600 may be (a/p, b/p). The decoding apparatus 1700 may derive the original motion vector difference (a, b) again by multiplying the signaled motion vector difference (a/p, b/p) by p.

Decoder-side motion vector derivation

In one embodiment, the decoder-end motion information derivation method 1) derives initial motion information for a target block, and 2) performs refinement on the initial motion information using a pre-defined operation. By doing so, motion information on the target block can be derived.

In embodiments, improving specific information may mean amending, correcting, or updating specific information. In embodiments, the terms "refinement," "amendment," and "correction" may be used interchangeably. Improved information may be created by performing improvement on specific information.

For example, a decoder-end motion information derivation method may include: 1) a motion information offset and/or motion in a second direction by applying a pre-defined operation to a motion information offset and/or motion vector difference in a first direction; A vector difference may be derived, and 2) motion information may be improved by adding the derived motion information offset and/or the motion vector difference to the motion information in the second direction.

For example, the pre-defined operation may include at least one of mirroring, scaling, and copying, and is not limited to the operations listed above.

For example, when mirroring is applied to a specific motion vector MV, the mirroring result may be -MV.

For example, when scaling is applied to a specific motion vector MV, 1) a picture-order-count (POC) interval between an image including a specific motion vector MV and a reference image indicated by the MV, and 2) A scaled motion vector may be derived by changing the size of a specific motion vector MV based on a POC interval between an image including the target block and a reference image of 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.

For example, when copying is applied to a specific motion vector MV, the result of copying may be MV.

In one embodiment, the decoder-end motion information derivation method may improve motion information by performing a search from a position indicated by initial motion information.

In one embodiment, the decoder-end motion information derivation method may refine the motion information candidate list by performing a search from an initial motion information candidate list. The initial motion information candidate list may include a plurality of (initial) motion information items. Through improvement, at least some of a plurality of pieces of motion information may be improved in the motion information candidate list.

For example, each piece of motion information in the enhanced motion information candidate list is either 1) initial motion information or 2) enhanced motion information generated by performing refinement using a decoder-end motion information derivation method on the initial motion information. can be Also, each piece of motion information is not limited to 1) initial motion information and 2) improved motion information. Initial motion information may refer to one of a plurality of pieces of motion information included in the initial motion information candidate list.

In one embodiment, the decoder-end motion information derivation method may: 1) compare matching costs of a plurality of candidates in the (initial) motion information candidate list; Orders within the motion information candidate list may be adjusted. In other words, reordering of a plurality of candidates may be performed based on matching costs of the plurality of candidates of the motion information candidate list. A plurality of candidates may be a plurality of pieces of motion information.

For example, a plurality of candidates in the (initial) motion information candidate list may be sorted in ascending order of matching costs of the plurality of candidates.

In embodiments, the initial motion information may be 1) motion information to which a decoder-side motion information derivation method is applied and/or 2) motion information to which each search step of a decoder-side motion information derivation method is applied.

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.

At least one of the following processes may be applied to the initial motion information. Motion information to which the following processing is applied may be used as initial motion information of a decoder-level motion information derivation method.

Process 1: Initial motion information can be refined using a decoder-level motion information derivation method.

Process 2: The initial motion information can be refined using a motion information offset.

For example, a motion vector difference may be added to initial motion information.

For example, a motion information offset may be added to initial motion information.

For example, initial motion information or part of the initial motion information may be changed to the same value as the motion information offset.

For example, when an Advanced Motion Vector Prediction (AMVP) mode is used for a target block, initial motion information may be a motion vector predictor.

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.

For example, in performing the decoder-end motion information derivation method, at least one motion information offset may be added to at least one of the improved motion information generated in the search steps.

For example, if the decoder-end motion information derivation method consists of N search steps, first improved motion information may be generated in a first search step. The first motion information offset may be added to the first enhanced motion information. A second search step may be performed on the first enhanced motion information. Thereafter, the second motion information offset may be added to the N−1 th improved motion information generated in the N−1 th search step. An N th search step may be performed on the N−1 th improved motion information.

For example, information on a motion information offset may be signaled/encoded/decoded.

The motion information offset may be determined by a rate-distortion optimization process. In this case, encoding efficiency of the target block may be improved.

For example, the motion information offset may be a pre-defined value. The pre-defined value may be zero. The pre-defined value may be (0, 0).

For example, a motion information offset may mean a motion vector difference.

For example, when bi-directional inter prediction is performed on the target block, at least one of the motion information offsets may be added to motion information in the L0 direction and motion information in the L1 direction, respectively.

For example, at least one of the motion information offsets may be added only to motion information in the LX direction.

For example, when bi-directional inter prediction is performed on a target block, a motion information offset may be added only to motion information in the LX direction.

X can be 0, 1 or a positive integer.

X may be a pre-defined value.

For example, the pre-defined value may be zero.

For example, the pre-defined value may be a value indicating a direction having a lower matching cost among the L0 direction and the L1 direction. Here, the matching cost may be a matching cost for motion information.

For example, the pre-defined value may be a value indicating a direction having a higher matching cost among the L0 direction and the L1 direction. Here, the matching cost may be a matching cost for motion information.

When a predefined value X is used, motion information or a direction (to which a motion information offset is added) may be determined in the decoding apparatus 1700 without signaling/encoding/decoding for the predefined value. By this determination, the amount of bits for signaling can be reduced.

Information indicating a pre-defined value X may be signaled/encoded/decoded. When signaling/encoding/decoding of information representing a pre-defined value X is performed, the pre-defined value X used in each block may be determined through rate-distortion optimization. Prediction performance of the target block may be improved by such a decision.

For example, the motion information offset may be a motion vector.

For example, the motion information offset can be determined using a combination between the angles in the angle list and the distance offsets in the distance offset list. The angle list may include a plurality of angles. The distance offset list may include a plurality of distance offsets.

In embodiments, the angle list may be a list of angles from the X axis/Y axis.

In embodiments, "X axis/Y axis" may be replaced with "horizontal axis/vertical axis" or "horizontal axis/vertical axis".

The angle list may be configured to include angles with a value of "i_ANGLE×π/NUM_ANGLE". i_ANGLE can be an integer 0,1, ... or (2×NUM_ANGLE-1).

The number of angles in the angle list can be 2×NUM_ANGLE.

NUM_ANGLE may be a pre-defined value. NUM_ANGLE can be 4, 8 or 16. NUM_ANGLE can be a positive integer.

Angles constituting NUM_ANGLE and/or the angle list may be determined based on a coding parameter of a target block. For example, the coding parameter may include motion information.

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).

For example, when the affine mode is used for the current prediction block, the value of NUM_ANGLE may be 8 (or a 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).

The values of NUM_ANGLE described above may only be illustrative. Each of the first value, the second value, the third value, and the fourth value may be a specific integer greater than or equal to 1.

For example, the distance offset list may be a list for indicating distances from the position indicated by the current motion vector. Alternatively, the distance offset list may be a list for indicating positions relative to the position indicated by the current motion vector.

For example, the current motion vector may mean the motion vector of the target block.

For example, when a motion vector (of a target block) points to a first position, the second position may be specified by an angle of an angle list and a distance offset of a distance offset list. In this case, the angle of the angle list may indicate an angle between the first line and the 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 location and the second location. A distance offset in the distance offset list may indicate a distance between the first location and the second location.

For example, the distance offset list may be configured to include at least one of 0, 1, 2, 4, and 8. That is to say, the distance offsets in the distance offset list can be all or part of 0, 1, 2, 4 and 8.

For example, the distance offset list may include a value corresponding to multi _offset -Pel as a distance offset. The distance offset of the distance offset list may indicate a value of multi _offset -pel.

For example, multi _offset can be 1/8, 1/4, 1/2, 1, 2, 4, 8, 16, 32 or a positive integer.

For example, distance offsets constituting the distance offset list and/or the size of the distance offset list may be determined based on a coding parameter of a target block or motion information of the target block.

In embodiments, the size of a list may mean the number of elements (or candidates) in the list.

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 list of distance offsets may be {1, 2, 4, 8, 16}.

The values of the distance offsets in the distance offset list described above may only be exemplary. The value of each distance offset may be a specific integer greater than or equal to 1.

For example, a location determined using a combination of a specific angle θ and a specific distance offset ρ may mean a location at a distance ρ in a direction at an angle θ from the X axis/Y axis.

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 pointing to a position at a distance ρ in a direction at which an angle from the X axis/Y axis is θ.

For example, a motion information offset may be added to a reference picture index.

For example, when the value of the reference image index of the target block is the first value and the value of the motion information offset added to the reference image index is the second value, the value of the reference image index of the target block is the first value + the second value. can be changed to The first value may be 0, 1 or a positive integer. The second value may be -2, -1, 0, 1, 2, or an integer.

For example, when the value of the reference image index of the target block is the first value and the value of the motion information offset added to the reference image index is the second value, the value of the reference image index of the target block may be changed to the second value. there is. The first value may be 0, 1 or a positive integer. The second value may be 0, 1 or a positive integer.

For example, in performing the decoder-end motion information derivation method, at least one of the enhanced motion information generated by each search step may be changed to have the same value as at least one motion information offset value.

For example, if the decoder-end motion information derivation method consists of N search steps, the motion vector of the improved motion information generated by the first search step may be changed to the first motion information offset. A second search step may be performed on motion information having a changed motion vector. Thereafter, the reference image index of the improved motion information generated by the N-1 th search step may be changed to the second motion information offset. An N-th search step may be performed on the motion information having the changed reference picture index.

For example, a motion information offset may mean a coding parameter. Alternatively, the motion information offset may mean motion information.

For example, the motion information offset may mean a motion vector or a reference picture index.

For example, when bi-directional inter prediction is used for the target block, motion information in the L0 direction and motion information in the L1 direction may be changed to be the same as the motion information offset for each direction.

For example, only motion information in the LX direction among motion information in a plurality of directions may be changed to be the same as the motion information offset.

For example, when bi-directional inter prediction is used for a target block, only motion information in an LX direction among motion information in a plurality of directions may be changed to be the same as the motion information offset.

The X may be 0, 1 or a positive integer.

The X may be a pre-defined value.

For example, the pre-defined value may be zero.

For example, the pre-defined value may be a value corresponding to a direction having a lower matching cost for motion information among directions L0 and L1.

For example, the pre-defined value may be a value corresponding to a direction having a higher matching cost for motion information among directions L0 and L1.

If a pre-defined value X is used, X can be determined at the decoder-end without signaling/encoding/decoding information indicating the pre-defined value. By this determination, the amount of bits for signaling can be reduced.

Information representing X may be signaled/encoded/decoded. When information representing X is signaled/encoded/decoded, X used in each block may be determined through rate-distortion optimization. Prediction performance of the target block may be improved by such a decision.

For example, the search in the embodiments may be performed using calculation of a cost function to determine the similarity between NUM_TEMPLATE_COMPARE templates.

In embodiments, search may refer to a process of determining motion information that satisfies a specific condition within a pre-defined search range.

For example, motion information satisfying a specific condition may refer to motion information having the lowest matching cost among motion information within a search range.

In embodiments, NUM_TEMPLATE_COMPARE can be 0, 1, 2 or a positive integer.

Enhanced motion information may refer to motion information that satisfies the above specific condition.

The enhanced motion information may mean at least one of 1) motion information determined through a decoder-end motion information derivation method and 2) motion information determined through each search step of the decoder-end motion information derivation method.

In embodiments, the search range may be a specific range centered on a location indicated by the initial motion information. In other words, the center of the search range may be a location indicated by the initial motion information. The specific range may be a range having a predefined width.

For example, the center of the search range may be a point indicated by the initial motion information. The search range may have a rectangle shape with a horizontal length of SR_X and a vertical length of SR_Y. Alternatively, the search range may have a shape of a diamond having 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-described embodiments.

Each of SR_X and SR_Y may be a pre-defined positive integer.

A template may be a subset of pixels included in a specific region.

For example, the template for the target block is 1) a subset of pixels included in TMSIZE_LEFT lines adjacent to the left of the target block and 2) included in TMSIZE_ABOVE lines adjacent to the top of the target block. may include one or more of the subsets of pixels that are However, the relationship between the location of the template and the location of the target block or the method of constructing the template is not limited to the above-described embodiments.

For example, the template for the reference block is 1) a subset of pixels included in TMSIZE_LEFT number of lines adjacent to the left side of the reference block and 2) a subset of pixels included in TMSIZE_ABOVE number of lines adjacent to the top of the reference block. may contain one or more. However, the relationship between the location of the template and the location of the reference block or the method of constructing the template is not limited to the above-described embodiments.

TMSIZE_LEFT and TMSIZE_ABOVE may be pre-defined integers greater than or equal to 0. TMSIZE_LEFT and TMSIZE_ABOVE may be equal to or different from each other.

For example, a cost function for the template of the target block and a cost function for the template of the reference block in the L0 direction and/or the L1 direction may be calculated.

For example, calculation of the cost function for the template of the reference block in the L0 direction and the cost function for the template of the reference block in the L1 direction may be performed.

In embodiments, the cost function is a Sum of Absolute Differences (SAD), a Sum of Absolute Transformed Differences (SATD), a Mean-Removed Sum of Absolute Differences (SATD). of Absolute Differences (MR-SAD), Mean Squared Error (MSE), and Sum of Squared Error (SSE). However, cost functions are not limited to the items listed above.

When the decoder-end motion information derivation method is performed on the target block, a cost function used in the decoder-end motion information derivation method may be pre-defined.

When the decoder-end motion information derivation method is performed on the target block, information on a cost function used in the decoder-end motion information derivation method may be signaled/encoded/decoded.

The template may be all or part of block information of one or more blocks included in a specific area.

For example, the template for the target block is 1) a subset of block information for blocks included in TMSIZE_LEFT lines adjacent to the left side of the target block and 2) blocks included in TMSIZE_ABOVE lines adjacent to the top of the target block. It may include one or more of the subsets of block information for . However, the positional relationship between the template and the target block or the method of constructing the template is not limited to the above-described embodiments.

For example, templates for reference blocks are 1) a subset of block information for blocks included in TMSIZE_LEFT lines adjacent to the left side of the reference block and 2) blocks included in TMSIZE_ABOVE lines adjacent to the top of the reference block. It may include one or more of the subsets of block information for . However, the positional relationship between the template and the reference block or the method of constructing the template is not limited to the above-described embodiments.

Each of TMSIZE_LEFT and TMSIZE_ABOVE may be a pre-defined integer greater than or equal to 0.

In embodiments, block information may mean at least one of target block information, neighboring block information, reference block information, and collocated block information.

Also, block information may include at least one of coding parameters.

Also, the block information may include at least one of information used in inter prediction, intra prediction, transformation, inverse transformation, quantization, inverse quantization, entropy encoding/decoding, and an in-loop filter.

That is, block information includes block size, block depth, block partition information, block shape (square or non-square), information indicating whether the block is partitioned in a quad tree format, and whether the block is partitioned in a binary tree format. , information indicating binary tree type splitting direction (horizontal direction or vertical direction), binary tree type splitting type (symmetric splitting or asymmetric splitting), prediction mode (intra prediction or inter prediction), intra luma prediction mode/direction, intra Chroma prediction mode/direction, intra partitioning information, inter partitioning information, coding block partitioning flag, block partitioning flag, transform block partitioning flag, reference sample filter tap, reference sample filter coefficient, block filter tap, block filter coefficient ), block boundary filter taps, block boundary filter coefficients, motion vectors (motion vectors for at least one of L0, L1, L2 and L3, etc.), motion vector differences (for at least one of L0, L1, L2 and L3, etc.) motion vector difference), inter prediction direction (inter prediction direction for at least one of unidirectional prediction and bi-prediction, etc.), reference picture index (reference picture index for 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 candidate, motion information candidate list, information indicating whether merge mode is used, merge index, merge candidate, merge candidate list, whether skip mode is used information, interpolation filter type, interpolation filter tap, interpolation filter coefficient, motion vector magnitude, motion vector representation precision (integer sample, 1/2 sample, 1/4 sample, 1/8 sample, 1/16 units or resolutions used for expressing motion vectors, such as samples and 1/32 samples), transform type, transform size, information indicating whether a first-order transform is used, information indicating whether a second-order transform is used, and Transform index, secondary transform index, information indicating presence/absence of a residual signal, coding block pattern, coding block flag, quantization parameter, residual quantization parameter, quantization matrix, information indicating whether an intra-loop filter is applied, intra-loop Filter coefficients, intra-loop filter tab, intra-loop filter shape/shape, information indicating whether deblocking filter is applied, deblocking filter coefficient, deblocking filter tab, deblocking filter strength, deblocking filter shape/shape, adaptive Information indicating whether 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 Tap, adaptive loop filter shape/form, binarization/inverse binarization method, context model determination method, context model update method, information indicating whether regular mode is performed, information indicating whether bypass mode is performed, Context bin, bypass bin, significant coefficient flag, last significant coefficient flag, flag indicating whether encoding/decoding of units of coefficient group is applied, position of last significant coefficient, coefficient value greater than 1 flag indicating whether the coefficient value is greater than 2, flag indicating whether the coefficient value is greater than 3, remaining coefficient value information, sign information, reconstructed luma sample, reconstruction Residual chroma sample, residual luma sample, residual chroma sample, luma transform coefficient, chroma transform coefficient, luma quantization level, chroma quantization level, transform coefficient level scanning method, size of motion vector search region at decoder-stage, at decoder-stage The shape of the motion vector search region of , the number of motion vector searches in the decoder-stage, CTU size information, minimum block size information, maximum block size information, maximum block depth information, minimum block depth information, slice identification information, slice division information, Tile identification information, tile type, tile division information, input sample bit depth, reconstructed sample bit depth, residual sample bit depth, transform coefficient bit depth, quantization level bit depth, overlapped block motion compensation Compensation mode indicator, Local Illumination Compensation; LIC) mode indicator and at least one value of a motion information offset, a combined form, or statistics.

For example, the cost function may be a function for determining similarity between motion information and/or coding parameters of templates.

In embodiments, the similarity between a first value and a second value may be determined by performing at least one of 1) a difference between two values, 2) a ratio between two values, and 3) an operation comparing a difference between two values to a particular value. can be judged using

For example, that the motion information (or coding parameters) of the templates are similar may mean a case where the difference between the motion information (or coding parameters) of the templates is less than or equal to a pre-defined positive number THRES_PARAMETER. . However, the criterion for determining similarity is not limited to the above comparison.

For example, that motion information (or coding parameters) of templates are similar may mean a case where the sum of differences between motion information (or coding parameters) of templates is less than or equal to a pre-defined positive number THRES_PARAMETER. . However, the criterion for determining similarity is not limited to the above comparison.

For example, that motion information (or coding parameters) of templates are similar may mean a case where a ratio between motion information (or coding parameters) of templates is less than or equal to a pre-defined positive number WEIGHT_THRES_PARAMETER. . However, the criterion for determining similarity is not limited to the above comparison.

For example, that motion information (or coding parameters) of templates are similar may mean a case where a ratio between motion information (or coding parameters) of templates is greater than or equal to a pre-defined positive number WEIGHT_THRES_PARAMETER. . However, the criterion for determining similarity is not limited to the above comparison.

For example, motion information (or coding parameters) of templates are similar means that the number of identical motion information (or coding parameters) in templates is greater than or equal to a pre-defined positive number THRES_PARAMETER. However, the criterion for judging similarity is not limited to the above comparison.

Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER may be a pre-defined positive integer.

Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER may be set based on motion information (or coding parameters). Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER may change according to motion information (or coding parameters). Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER may be set based on the type of motion information (or coding parameter). Each of THRES_PARAMETER and WEIGHT_THRES_PARAMETER may change according to the type of motion information (or coding parameter).

For example, the enhanced motion information candidate list may include DMVG_NUM pieces of motion information.

DMVG_NUM may be a pre-defined positive integer of 1 or greater.

Information representing DMVG_NUM can be signaled/encoded/decoded.

DMVG_NUM may be equal to the size of the initial motion information candidate list.

DMVG_NUM may be different from the size of the initial motion information candidate list.

For example, DMVD_NUM may be smaller than the size of the initial motion information candidate list.

Each candidate in the enhanced motion information candidate list may be initial motion information or enhanced motion information generated based on the initial motion information.

For example, the improved motion information candidate list may include DMVG_NUM pieces of initial motion information having the lowest matching costs among candidates in the initial motion information candidate list.

In embodiments, "information with the lowest matching costs" may mean "information selected in ascending order of matching costs".

For example, improved motion information may be generated by applying enhancement to candidates of an initial motion information candidate list. An advanced motion information candidate list may be configured to include DMVG_NUM pieces of advanced motion information having the lowest matching costs among the advanced motion information.

For example, when the initial motion information candidate list for the L(1-X) direction is refined, when motion information in the LX direction has already been determined as MVP_LX, each of the initial motion information candidate list for the L(1-X) direction The candidate can be improved so that the cost of bilateral matching with MVP_LX is minimized.

For example, when the initial motion information candidate list for the L(1-X) direction is improved, when motion information in the LX direction has already been determined as MVP_LX, the candidates of the initial motion information candidate list for the L(1-X) direction The order between the two sides may be reordered in ascending order of matching costs with MVP_LX.

X can be 0, 1 or a positive integer.

X may be a pre-defined value. The pre-defined value may be zero.

For example, the pre-defined value may mean a value indicating a direction having a lower matching cost among the L0 direction and the L1 direction. Here, the matching cost for a specific direction may be a matching cost for motion information in a specific direction.

For example, the pre-defined value may mean a value indicating a direction having a higher matching cost among the L0 direction and the L1 direction. Here, the matching cost for a specific direction may be a matching cost for motion information in a specific direction.

It can be signaled/encoded/decoded representing X.

Information representing X may be determined through a rate-distortion optimization process. Through this determination, encoding efficiency of the target block may be improved.

For example, in performing a search in the decoder-end motion information derivation method, a position-based weight w may be used for a matching cost for motion information in each search process.

For example, a matching cost for specific motion information in each search process may be improved by an operation using a weight w. This operation may mean at least one of the four arithmetic operations.

Whether the weight w is used and/or the value of the weight w may be determined based on a position indicated by the initial motion information.

Initial motion information may refer to initial motion information of a decoder-level motion information derivation method. Alternatively, the initial motion information may refer to motion information to be improved in each search process of the decoder-side motion information derivation method.

In embodiments, the weight w may be a value multiplied by a matching cost of specific motion information in a search process.

For example, the weight w may be determined based on a distance between a position indicated by specific motion information and a position indicated by initial motion information.

For example, whether the weight w has a value other than 1 in the search process may be determined based on a distance between a position indicated by specific motion information and a position indicated by initial motion information.

For example, whether to multiply a matching cost of specific motion information by a weight w in a search process may be determined based on a distance between a location indicated by the specific motion information and a location indicated by the initial motion information.

Weight w may be pre-defined.

For example, a matching cost at a position indicated by the initial motion information may be multiplied by a pre-defined weight W. At this time, W may be a value of 1 or less or a value of 1 or more.

For example, when the AMVP mode is used for the target block, W at a position indicated by the initial motion information may be a value greater than or equal to 1.

For example, when merge mode is used for a target block, W at a position indicated by initial motion information may be a value of 1 or less.

For example, when the AMVP mode is used for the target block, the weight W may have a larger value as it is closer to the position indicated by the initial motion information.

For example, when the merge mode is used for the target block, the weight W may have a lower value as it is closer to the position indicated by the initial motion information.

For example, in performing a specific search step of the decoder-end motion information derivation method, when the value of the weight W multiplied by the matching cost for the specific motion information is 1 or less, the initial motion information of the current search step is the specific motion information may increase the likelihood of improvement.

For example, in performing a specific search step of the decoder-end motion information derivation method, when the value of the weight W multiplied by the matching cost for the specific motion information is 1 or more, the initial motion information of the current search step is the specific motion information. The possibility of improvement with information may be reduced.

The decoder-end motion information derivation method may include one or more search steps.

In each search step, improved motion information may be determined through a search from initial motion information in the search step.

In the two different search steps, at least one of initial motion information, a template construction method, a type of cost function for determining similarity, a unit in which motion information is improved, a search range, a search pattern, a search resolution, and a search method may be different from each other.

For example, each search step of the decoder-end motion information derivation method may be regarded as an independent decoder-end motion information derivation method composed of one search step.

For example, a first decoder-end motion information derivation method consisting of L-search steps and a second decoder-end motion information derivation method consisting of M-search steps consist of N-search steps. It can be regarded as a decoder-end motion information derivation method. N may be the sum of L and M.

Types of these search methods may differ from each other according to one or more of 1) search pattern, 2) search resolution, 3) search range, and 4) units from which motion information is derived. However, the criterion for classifying the type of search method is not limited to these conditions.

The search resolution may be one of 4-pels, full-pels, half-pels, and quarter-pels. However, the search resolution is not limited to the aforementioned pels.

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.

The search using the diamond pattern, when (0, 0) indicates the position indicated by the initial motion information, (0, 2×RR), (RR, RR), (2×RR, 0), (RR, - RR), (0, -RR), (-RR, -RR), (-RR, 0), (-RR, RR), and (0, 0) It may mean searching for one or more of the positions there is. RR may mean search resolution and may be a pre-defined positive number.

A search using a cross pattern is performed when (0, 0) indicates the position indicated by the initial motion information, (0, RR), (RR, 0), (0, -RR), (-RR, 0) and It may mean searching for one or more of the positions of (0, 0).

A search using a full-search pattern may mean searching for all locations within a pre-defined search range.

For example, when FS_i has values from -FS_X to FS_X and FS_j has values from -FS_Y to FS_Y, a search using a full-search pattern finds locations of (FS_i×RR, FX_j×RR) It can mean exploring. Here, (0, 0) may be a position indicated by the initial motion information. However, the search range is not limited to the aforementioned positions. Each of FS_X and FS_Y may be a pre-defined positive number.

For example, the decoder-end motion information derivation method may be configured in the order of 1) deriving motion information for the entire block and 2) deriving motion information for sub-blocks of the block. However, a method of deriving motion information performed in each step and an order between steps are not limited to the methods and order described in the embodiments.

For example, the decoder-end motion information derivation method may consist of N search steps. In the first search step, motion information on the first block may be derived. In a second search step located later than the first search step, motion information on the second block may be derived. Here, the unit of the second block may be smaller than the unit of the first block. In other words, the second block may be a sub-block of the first block. Alternatively, the unit of the first block and the unit of the second block may be the same. In other words, the first block and the second block may be the same.

For example, if motion information for the entire block is derived in the first search step, motion information for the entire block or motion information for a sub-block of the block may be derived in the second search step.

A search method in each step may be pre-defined based on a coding parameter or motion information of a target block.

When the decoder-side motion information derivation method is composed of two or more search steps, the initial motion information input in a specific search step after the first search step is improved motion information in a search step prior to the specific search step. can be

For example, the decoder-end motion information derivation method may consist of three search steps. 1) In the first search step, motion information can be improved by using bilateral matching for the entire block. 2) In the second search step, motion information may be improved through bilateral matching in units of subblocks having a size of 16x16. 3) In the third search step, motion information may be improved by using light-flow based motion information prediction in units of sub-blocks having a size of 8x8.

Motion information improved through the decoder-end motion information derivation method of the embodiments may replace the initial motion information. For example, the enhanced motion information may be used to determine motion information for inter prediction in the L0 direction and/or the L1 direction of the target block.

The decoder-side motion information derivation method of the embodiments includes template matching (TM), bilateral matching (BM), decoder-side motion vector refinement (DMVR), symmetric motion vector difference (Symmetric Motion Vector Difference; SMVD) and motion information prediction based on optical-flow. However, the decoder-end motion information derivation method is not limited to the above methods.

Symmetric Motion Vector Difference (SMVD) mode

20 illustrates motion vector differences in a symmetric motion vector difference mode according to an example.

The SMVD mode is a prediction mode using bi-directional prediction and may be one of AMVP modes.

For example, when bi-directional prediction is used for a target block, the symmetric motion vector differential mode may be signaled as a lower mode of the AMVP mode.

For example, when the symmetric motion vector differential mode is used, the reference picture index may be determined through a pre-defined method.

For example, signaling of a reference picture list may be omitted.

For example, when the SMVD mode is used, a pair of reference images may be specified in the L0-direction reference image list and the L1-direction reference image list. Here, the specified reference images may be a first reference image in the L0 reference image list and a second reference image in the L1 reference image list. The first POC interval may be a difference between the POC of the target image and the POC of the first reference image. The second POC interval may be a difference between the POC of the target image and the POC of the second reference image. The target image may be an image including the target block. The first POC interval and the second POC interval may be the same. A direction of the first reference image with respect to the target image and a direction of the second reference image with respect to the target image may be opposite to each other.

In other words, reference picture indices used in SMVD mode may be reference picture indices of reference pictures having the same POC interval.

When the SMVD mode is used, a motion vector prediction index can be signaled/encoded/decoded for reference pictures in each direction, and information on one motion vector difference can be signaled/encoded/decoded. In SMVD mode, two MVP indices and one MVD can be signaled.

When the SMVD mode is used, motion vector differences for reference images in each direction may include encoded/decoded first motion vector differences and second motion vector differences symmetrical to the first motion vector differences.

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.

21 illustrates syntax elements signaled/encoded/decoded in a symmetric motion vector differential mode according to an example.

x0 and y0 may mean that the upper left location of the luma block of the target block is (x0, y0).

sps_smvd_enabled_flag may be a syntax element determined in a sequence parameter set, and may be an indicator indicating whether to activate a symmetric motion vector differential mode in a target sequence.

When sps_smvd_enabled_flag is the first value, the symmetric motion vector differential mode may be deactivated in the target sequence. The first value may be 0 or false.

When sps_smvd_enabled_flag is the second value, the symmetric motion vector differential mode may be activated in the target sequence. The second value may be 1 or true.

inter_pred_idc may be a syntax element for an inter prediction direction. inter_pred_idc may be an inter prediction indicator.

When inter_pred_idc is the first value, unidirectional inter prediction in the L0 direction may be performed on the target block. The first value may be 1 or PRED_L0.

When inter_pred_idc is the second value, unidirectional inter prediction in the L1 direction may be performed on the target block. The second value may be 2 or PRED_L1.

When inter_pred_idc is the third value, bi-directional inter prediction may be performed on the target block. The third value may be 3 or PRED_BI.

inter_affine_flag may be an indicator indicating whether affine mode is performed for a target block.

When inter_affine_flag is the first value, the affine mode may not be performed on the target block. The first value can be 0 or false.

When inter_affine_flag is the second value, affine mode may be performed on the target block. The second value may be 1 or true.

RefIdxSymL0 is a reference picture index in the L0 direction determined for the symmetric motion vector difference mode. RefIdxSymL1 is a reference image index in the L1 direction determined for the symmetric motion vector difference mode.

For example, RefIdxSymL0 may be an index indicating a reference image having the smallest POC interval among reference images in the L0 direction reference image list. Here, the POC interval may be the difference between the POC of the target image and the POC of the reference image.

For example, RefIdxSymL1 may be an index indicating a reference image having the largest POC interval among reference images of the direction reference image list among reference images of the L1 direction reference image list. Here, the POC interval may be the difference between the POC of the target image and the POC of the reference image.

ph_mvd_l1_zero_flag may be a syntax element determined in a picture parameter set, and may be an indicator indicating whether to perform signaling/decoding/decoding for the L1-direction motion vector difference in the target image.

When ph_mvd_l1_zero_flag is the first value, signaling/encoding/decoding of an L1-direction motion vector differential may be performed in the target image. The first value can be 0 or false.

When ph_mvd_l1_zero_flag is the second value, signaling/encoding/decoding of an L1-direction motion vector differential may not be performed in the target image. The second value may be 1 or true.

NumRefIdxActive[X] may indicate the maximum number of reference images usable for performing inter prediction in the LX direction. X can be 0, 1 or a positive integer.

sym_mvd_flag may be an indicator indicating whether to perform the symmetric motion vector differential mode.

For example, when sym_mvd_flag is the first value, the symmetric motion vector differential mode may not be performed on the target block. The first value can be 0 or false.

For example, when sym_mvd_flag is the second value, the symmetric motion vector differential mode may be performed on the target block. The second value may be 1 or true.

MotionModelIdc may indicate 1) whether the affine mode is used for the target block and 2) the number of control point motion vectors used in the affine mode when the affine mode is used for the target mode.

MotionModelIdc having the first value may mean that the affine mode is not performed for the target block. The first value may be 0.

When MotionModelIdc is the second value, it may mean that the affine mode is performed on the target block and two control point motion vectors are used. The second value may be 1.

MotionModelIdc having a third value may mean that an affine mode is performed on the target block and three control point motion vectors are used. The third value may be 2.

mvp_l0_flag may mean a motion information index in the L0 direction.

mvp_l1_flag may mean a motion information index in the L1 direction.

mvd_coding(x0, y0, X, Y) may represent that encoding/decoding of the motion vector differential for the Y-th motion information in the LX direction is performed. When the affine mode is not performed for the target block, Y may always be 0. When the affine mode is performed on the target block, Y may be one of 0, 1 and 2. A range of values that Y can have may be determined based on MotionModelIdc.

MvdL0[x0][y0][0] may mean a horizontal component of the L0-direction motion vector difference. MvdL0[x0][y0][1] may mean a vertical component of the L0-direction motion vector difference.

MvdL1[x0][y0][0] may mean a horizontal component of the L1-direction motion vector difference. MvdL1[x0][y0][1] may mean a vertical component of the L1-direction motion vector difference.

Bilateral Matching (BM)

Bilateral matching is a decoder-end motion information derivation method that improves motion information by using a reference block in the L0 direction and a reference block in the L1 direction as templates and calculating a cost function (or a bilateral matching cost) between the two templates. can

The bilateral matching cost may refer to a value resulting from calculation of a cost function between templates of an L0-direction reference block and an L1-direction reference block used in bilateral matching.

The reference block may mean a reference block indicated by 1) initial motion information, 2) motion information derived in the search process of both sides matching, or 3) motion information finally improved through both sides matching.

Both sides matching may operate only when a pre-defined enabling condition is satisfied.

For example, matching both sides can always work.

For example, both-side matching may be performed when an inter prediction mode is used for a target block and two or more reference blocks are used.

For example, matching on both sides may be performed only when the first direction and the second direction are different from each other and the first POC interval and the second POC interval are the same. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image. The first POC interval may be a difference between the POC of the target image and the POC of the L0 direction reference image. The second POC interval may be a difference between the POC of the target image and the POC of the L1-direction reference image.

For example, matching on both sides may be performed only when the first direction and the second direction are different from each other. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

Here, that the first direction and the second direction are different may mean that Equation 1 below is satisfied.

[Equation 1]

(POC _t - POC0) × (POC _t - POC1) < 0

Here, that the first direction and the second direction are the same may mean that Equation 2 below is satisfied.

[Equation 1]

(POC _t - POC0) × (POC _t - POC1) > 0

POC _t may be the POC of the target image.

POC0 may be the POC of the L0 direction reference image.

POC1 may be the POC of the L1-direction reference image.

Matching both sides may involve one or more search steps.

For example, both-side matching may be configured to sequentially include 1) deriving motion information for the entire block and 2) deriving motion information for sub-blocks of the block. However, the method of deriving motion information performed in each step and the sequence of steps are not limited to the above configuration.

In each search step of bilateral matching, motion information for BM_NUM directions among the motion information for the L0 direction and the motion information for the L1 direction may be improved. BM_NUM can be 0, 1, 2 or a positive integer. BM_NUMs used in both matching steps may be the same or different.

For example, when BM_NUM is 1 in a specific search step, motion information improvement may be performed only on motion information in the LX _BM direction in the specific search step.

For example, if BM_NUM is 1 and X _BM is 0 in a specific search step, only the search in the L0 direction is performed while the template in the L1 direction and motion information in the L1 direction are fixed in the specific search step. can

X _BM can be 0, 1 or a positive integer.

X _BM may be pre-defined.

For example, the X _BM direction may be a direction having a larger POC interval among the L0 direction and the L1 direction. The POC may be the difference between the POC of the reference image (in a specific direction) and the POC of the target image.

For example, the X _BM direction may be a direction having a higher template matching cost among the L0 direction and the L1 direction. Here, the template matching cost in a specific direction may be the template matching cost of motion information in a specific direction.

Information on X _BM can be signaled/encoded/decoded.

The same X _BM can be used in the search steps of both matches. Alternatively, different X _BMs may be used in the search steps of both sides matching.

For example, in performing bi-prediction matching, a cost function used for calculating a bi-prediction with weights of a target block based on a weight index of bi-prediction with weights and/or a weight index of bi-prediction with weights of the target block can be determined.

In embodiments, weighted biprediction may mean Bi-prediciton with CU-level Weights (BCW) at the CU level.

For example, if the first weight and the second weight of the target block are the same, both matching costs may be calculated using SAD or SATD. If the first weight and the second weight of the target block are different from each other, both matching costs may be calculated using MRSAD or MRSATD. Here, the first weight may be a weight in inter-weighted biprediction for the L0 direction. The second weight may be a weight in inter-weighted biprediction for the L1 direction.

BM_NUM can be 0, 1, 2 or a positive integer.

BM_NUM may be pre-defined.

BM_NUM in each search step of bilateral matching can be determined based on coding parameters. Alternatively, BM_NUM may be determined based on at least one of motion information, a search step of both sides matching, a matching cost in the previous search step, a matching cost for initial motion information in the current search step, and BM_NUM in the previous search step.

For example, BM_NUM may be 1 or 2 in the first search step of both sides matching.

For example, BM_NUM of the current search step may be determined based on matching cost in the previous search step.

For example, if a difference between a matching cost for initial motion information in the previous search step and a matching cost for improved motion information in the previous search step is smaller than COSTDIFF_FORBMNUM, BM_NUM in the current search step may be 0.

COSTDIFF_FORBMNUM can be 0, 1, 2, 4, 8, 16 or a positive integer.

For example, COSTDIFF_FORBMNUM may be determined based on the size of the target block. COSTDIFF_FORBMNUM may be the product of the number of pixels in the target block and a specific value. A specific value can be 0, 1, 2, 4, 8 or any positive integer.

For example, if BM_NUM in the previous search step is 0, BM_NUM in the current search step may be 0.

For example, if the matching cost for initial motion information in the current search step is smaller than COSTDIFF_FORBMNUM_INIT, BM_NUM of the target block may be 0.

COSTDIFF_FORBMNUM can be 0, 1, 2, 4, 8, 16 or a positive integer.

For example, COSTDIFF_FORBMNUM_INIT may be determined based on the size of the target block. COSTDIFF_FORBMNUM_INIT may be the product of the number of pixels in the target block and a specific value. A specific value can be 0, 1, 2, 4, 8 or any positive integer.

For example, when motion improvement is performed in the search step of bilateral matching, only when the matching cost for the L0-direction motion information of the initial motion information of the current search step is greater than COSTDIFF_FORBMNUM_INIT, the motion information for the L0-direction motion information Improvements can be made.

For example, when motion improvement is performed in the search step of bilateral matching, only when the matching cost for the L1-direction motion information of the initial motion information of the current search step is greater than COSTDIFF_FORBMNUM_INIT, the motion information for the L1-direction motion information Improvements can be made.

COSTDIFF_FORBMNUM can be 0, 1, 2, 4, 8, 16 or a positive integer.

For example, COSTDIFF_FORBMNUM_INIT may be determined based on the size of the target block. COSTDIFF_FORBMNUM_INIT may be the product of the number of pixels in the target block and a specific value. A specific value can be 0, 1, 2, 4, 8 or any positive integer.

The fact that BM_NUM in a specific search step of both sides matching is 0 may indicate that motion information improvement is not performed in the specific search step. Alternatively, when BM_NUM of a specific search step of both sides matching is 0, it may indicate that the above specific search step is not performed.

For example, when both sides matching is performed, BM_NUM in the motion information derivation step for the entire block may be 1, and BM_NUM may be 2 in the motion information derivation step for the sub-block. In this case, in the motion information derivation step for the entire block, only the motion information in the LX _BM direction can be improved, and in the motion information derivation step for the sub-block, both the motion information in the L0 direction and the L1 direction are improved. It can be.

22 illustrates matching on both sides according to an example.

22 shows a case where BM_NUM is 2 in the step of deriving motion information for all blocks of bilateral matching.

MV0 may be initial motion information in the L0 direction.

MV1 may be initial motion information in the L1 direction.

MV _diff may mean a motion information improvement value derived through matching of both sides.

MV0' and MV1' may be motion information derived through matching of both sides.

In both-side matching, the size of the motion information enhancement value in the L0 direction and the size of the motion information enhancement value in the L1 direction may be the same. The direction of the motion information enhancement value in the L0 direction and the direction of the motion information enhancement value in the L1 direction may be opposite to each other. That is, Equations 3 and 4 below can be established.

[Formula 3]

MV0' = MV0 + MV _diff

[Formula 4]

MV1' = MV1 - MV _diff

Motion information improved through the decoder-end motion information derivation method may mean the sum of initial motion information and a motion information improvement value.

Decoder-side MV Refinement (DMVR)

The decoder-end motion vector enhancement mode may be a mode for performing motion information enhancement using bilateral matching.

The decoder-side motion vector enhancement mode may be a decoder-side motion vector enhancement technique used when POC intervals of reference pictures are the same when bi-prediction is used.

The decoder-end motion vector enhancement mode can be used if certain conditions are met without explicit signaling.

For example, the decoder-end motion vector enhancement mode can be performed only when the bidirectional merge mode is used for the target block and the first POC interval and the second POC interval are the same. However, the conditions for performing the decoder-stage motion vector enhancement mode are not limited to the above conditions. The first POC interval may be a difference between the POC of the target image and the POC of the reference image in the L0 direction. The second POC interval may be a difference between the POC of the target image and the POC of the reference image in the L1 direction.

Template Matching (TM)

23 illustrates template matching according to an embodiment.

In template matching, motion information may be improved through calculation of cost functions for templates of a target block and a reference block.

The template matching cost may refer to a result of calculation using a cost function for a template of a target block and a template of a reference block used in template matching.

The reference block is at least one of 1) a reference block indicated by initial motion information, 2) a reference block indicated by motion information derived in the template matching search process, and 3) a reference block indicated by motion information finally improved through template matching. can be

Motion information improved through template matching may be motion information having the lowest matching cost derived in a template matching search process. However, the motion information derivation method is not limited to the above criteria.

For example, the template of template matching is 1) a subset of pixels included in TMSIZE_LEFT lines adjacent to the left side of the target block (or reference block) and 2) TMSIZE_ABOVE adjacent to the top of the target block (or reference block). It may include one or more of the subsets of pixels included in the lines. However, the relationship between the location of the template and the location of the target block (or reference block) or the method of constructing the template is not limited to the above relationship or method. Here, each of TMSIZE_LEFT and TMSIZE_ABOVE may be a pre-defined value, and may be 0, 1, 2, 3, 4, or a positive integer greater than or equal to 4. TMSIZE_LEFT and TMSIZE_ABOVE may be the same. Alternatively, TMSIZE_LEFT and TMSIZE_ABOVE may be different.

TMSIZE_LEFT and/or TMSIZE_ABOVE may be determined based on coding parameters. TMSIZE_LEFT and/or TMSIZE_ABOVE may be determined based on motion information of the target block.

24 illustrates a first relationship between search patterns and resolutions according to an example.

25 illustrates a second relationship between search patterns and resolutions according to an example.

26 illustrates a third relationship between search patterns and resolutions according to an example.

27 illustrates a fourth relationship between search patterns and resolutions according to an example.

28 illustrates a fifth relationship between search patterns and resolutions according to an example.

29 illustrates a sixth relationship between search patterns and resolutions according to an example.

A search pattern and resolution may be configured based on a coding parameter for a target block or motion information of the target block. A search pattern and resolution may be changed according to a coding parameter for a target block or motion information of the target block.

A search pattern and resolution in the search step of template matching may be configured based on the coding parameter or motion information of the target block according to the order of the tables shown in FIGS. 24 to 29 .

For example, when the AMVP mode is used for the target block and the resolution determined through the adaptive motion vector resolution is 4-pel, after the diamond pattern search using the 4-pel search resolution is performed, the 4-pel search resolution A search of a cross pattern using can be performed.

In the tables of FIGS. 24 and 29, ALT_IF may indicate an index of an adaptive interpolation filter. Interpolation filters may be applied to compute pixel values at sample locations of a particular resolution. The adaptive interpolation filter may be an interpolation filter selected by an index from among a plurality of interpolation filters. In other words, when an adaptive interpolation filter is applied, different interpolation filters may be used according to an index to calculate a pixel value at a sample location of a specific resolution.

For example, a particular resolution may be half-pel. However, the specific resolution is not limited to half-pel.

For example, the interpolation filter determined by the index may be one of a 6-tap interpolation filter and an 8-tap interpolation filter. However, the determination method of the interpolation filter is not limited to the above determination method.

30 illustrates a method of constructing a first template in an affine mode according to an example.

31 illustrates a second template configuration method in an affine mode according to an example.

CPMV may be an affine control point motion vector.

For example, the template matching cost in the affine mode is the sum of the template matching costs of the sub-block templates A0, A1, A2, A3, L0, L1, L2, and L3 or the sub-block templates A0, A1, It may be the average of template matching costs of A2, A3, L0, L1, L2 and L3).

movement information

In embodiments, the motion information may include a motion vector, a reference picture index, an inter prediction indicator, mode information, a merge flag, a motion information index, a reference picture index, a prediction list utilization flag, reference picture list information, a reference picture, At least one of a motion vector candidate, a motion information index, a merge candidate, a merge index, a size of a motion vector difference, a sign of each component of a motion vector difference, an indicator of a superimposed block motion compensation mode, and an indicator of a local illumination compensation mode. can include However, motion information is not limited to the aforementioned information.

Motion information search method

The motion information search method may refer to an inter prediction method including one or more decoder-end motion information derivation methods.

For example, when a motion information search method is performed on a target block, bi-directional inter prediction may always be performed on the target block.

For example, when a motion information search method is performed on a target block, information on an inter prediction direction of the target block may be signaled/encoded/decoded.

In the motion information search method, at least one of constructing a motion information candidate list and determining motion information may be performed.

A motion vector candidate list may be a list including one or more motion vector candidates. Even when only one motion information candidate is used in the motion information search method, it can be considered that a motion information candidate list having a size of 1 is constructed.

For example, in the motion information search method, one or more motion information candidate lists may be configured.

The decoder-end motion information derivation method in the motion information search method may be performed at least once in at least one of constructing a motion information candidate list and determining final motion information.

For example, in the motion information search method, DMVDMODE_NUM number of decoder-end motion information derivation methods may be performed. DMVDMODE_NUM can be 0 or a positive integer. Decoder-end motion information derivation methods may be the same or different.

In the motion information search method, decoder-end motion information derivation methods used for constructing a motion information candidate list may be the same or different.

Decoder-end motion information derivation methods used to determine motion information in the motion information search method may be the same or different.

Decoder-end motion information derivation methods used in the motion information search method may be specified according to a pre-defined method, and may be specified through signaling/encoding/decoding of information about the decoder-end motion information derivation methods. .

For example, when a motion information candidate list is constructed using two or more decoder-end motion information derivation methods, methods for specifying the decoder-end motion information derivation methods may be different from each other.

For example, when two or more decoder-end motion information derivation methods are used to determine final motion information, the methods for specifying the decoder-end motion information derivation methods may be different from each other.

For example, a first method for specifying a decoder-end motion information derivation method used to construct a motion information candidate list and a second method for specifying a decoder-end motion information derivation method used for determining final motion information. may be the same. Alternatively, the first method and the second method may be different from each other.

For example, the decoder-end motion information derivation method includes 1) coding parameter of the target block, 2) motion information of the target block, 3) reference image of the target block, 4) motion vector of the target block, and 5) coding of neighboring blocks. It can be specified using at least one of parameters, 6) motion information of neighboring blocks, 7) reference images of neighboring blocks, and 8) motion vectors of neighboring blocks.

For example, if the target block satisfies the activation condition of both sides matching, both sides matching may be specified as the decoder-end motion information derivation method. Template matching may be specified as a decoder-end motion information derivation method when the target block does not satisfy the activation condition of bilateral matching.

If the target block does not satisfy both the activation condition of bilateral matching and the activation condition of template matching, the decoder-end motion information derivation method may not be performed on the target block. Alternatively, if the target block does not satisfy both the activation condition of both sides matching and the activation condition of template matching, the motion information search method may not be performed on the target block.

In embodiments, conditions for activating both sides matching are: 1) bi-directional inter prediction is used for the target block, 2) the first POC difference and the second POC difference are the same, and 3) the first direction and the second direction are different. it could be Alternatively, the conditions for activating both sides matching are 1) whether bi-directional inter prediction is used for the target block, 2) whether the first POC difference and the second POC difference are the same, and 3) whether the first direction and the second direction are different. may include whether Here, the first POC difference may be a difference between the POC of the target image and the POC of the L0 direction reference image. The second POC difference may be a difference between the POC of the target image and the POC of the L1-direction reference image. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

In embodiments, activation conditions for matching on both sides may be: 1) bi-directional inter prediction is used for the target block, and 2) the first direction and the second direction are different from each other. Alternatively, conditions for activating both sides matching may include 1) whether bi-directional inter prediction is used for the target block and 2) whether the first direction and the second direction are different from each other. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

For example, when the target block satisfies the activation condition of bilateral matching, motion information improvement using bilateral matching may be performed within the motion information search method. If the target block does not satisfy the activation condition of bilateral matching, motion information improvement using bilateral matching may not be performed when the motion information search method is performed.

Construction of motion information candidate list and determination of final motion information

As described above with reference to steps 1810 and 1930, the prediction information may include a motion information candidate list and final motion information, which will be described later. Determination of prediction information may include configuration of a motion information candidate list and determination of final motion information.

For example, final motion information of a motion information search method or motion information derived from the final motion information may be determined as motion information used for inter prediction of a target block.

Encoding/decoding of an image may be performed using motion information used for inter prediction of a target block. For example, encoding/decoding of an image may include one or more of inter prediction, transformation, inverse transformation, quantization, inverse quantization, entropy encoding/decoding, and in-loop filtering. However, encoding/decoding of an image is not limited to the processes listed above.

For example, a reference block for a target block may be determined from the final motion information of the motion information search method or motion information derived from the final motion information.

Encoding/decoding of an image may be performed using the determined reference block. For example, encoding/decoding of an image may include one or more of inter prediction, transformation, inverse transformation, quantization, inverse quantization, entropy encoding/decoding, and in-loop filtering. However, encoding/decoding of video is not limited to the above processes.

Motion information of a block on which a motion information search method is performed may be referred to from neighboring blocks. The motion information referred to by the neighboring block may be at least one of 1) initial motion information of the motion information search method and 2) motion information derived in each search step of the decoder-end motion information derivation method in the motion information search method.

For example, a target block may refer to motion information of neighboring blocks in a target image including the target block. In this case, when the motion information search method is performed on the neighboring block, the motion information referred to from the neighboring block may be initial motion information of the motion information search method.

For example, a target block may refer to motion information of a neighboring block in an image other than the target image including the target block. In this case, when the motion information search method is performed on the neighboring block, the motion information referred to from the neighboring block may be final motion information of the motion information search method.

For example, a target block may refer to motion information of a neighboring block in an image other than the target image including the target block. In this case, when the motion information search method is performed on the neighboring block, the motion information referred to from the neighboring block may be improved motion information in the first search step of the motion information search method.

For example, a target block may refer to motion information of a neighboring block in an image other than the target image including the target block. In this case, when the motion information search method for the neighboring block is performed, one of initial motion information and final motion information in the motion information search method for the neighboring block may be selected. The selected motion information may be determined as motion information referred to from neighboring blocks. For example, this selection may be performed based on matching costs of motion information.

For example, motion information having a lower template matching cost or a lower both-side matching cost among two pieces of motion information may be selected and referred to.

When the motion information search method is performed on the target block, in-loop filtering may be performed on the target block and/or neighboring blocks.

When in-loop filtering is performed on the target block and/or neighboring blocks, information related to the in-loop filter may be determined based on motion information of the target block and/or neighboring blocks.

The information related to the in-loop filter is 1) information indicating whether the in-loop filter is applied, 2) the coefficients of the in-loop filter, 3) the filter tab of the in-loop filter, 4) the shape of the in-loop filter, and 5) It may be one or more of the types of in-loop filters. However, the information related to the in-loop filter is not limited to the information listed above.

The motion information of the target block used to determine the information related to the in-loop filter is derived through 1) initial motion information of the motion information search method and 2) each search step of the decoder-end motion information derivation method in the motion information search method. It may include one or more of the received motion information.

For example, motion information of a target block used to determine information related to an in-loop filter may be initial motion information of a motion information search method.

For example, motion information of a target block used to determine information related to an in-loop filter may be final motion information of a motion information search method.

For example, motion information of a target block used to determine information related to an in-loop filter may be improved motion information in a first search step of a motion information search method.

For example, when the difference between the motion information of the target block and the motion information of neighboring blocks is less than or equal to a specific value, a weak deblocking filter may be used. If a difference between the motion information of the target block and the motion information of neighboring blocks is greater than a specific value, a strong deblocking filter may be used.

Certain values can be positive integers. The specific value may be at least one of 1, 2, 4, 8, and 16. However, the specific value is not limited to the values listed above.

The in-loop filter may be at least one of a deblocking filter and an adaptive in-loop filter. However, the in-loop filter is not limited to the filters listed above.

For example, DMVDMODE_FLAG may be an indicator indicating whether a motion information search method for a target block is performed. When performing encoding/decoding for DMVDMODE_FLAG, a probability model and/or a context model may be used. A probability model and/or a context model used when encoding/decoding DMVDMODE_FLAG of a target block is performed may be determined according to pre-defined conditions.

For example, the probabilistic model and/or the context model may be determined based on weights of inter-weighted biprediction of the target block.

For example, a probabilistic model and/or a context model may be determined according to whether a weight for the L0 direction and a weight for the L1 direction in inter-weighted biprediction are the same. A probabilistic model and/or a context model may be selected from among a plurality of probabilistic models and/or a plurality of context models according to whether the weight for the L0 direction and the weight for the L1 direction are the same.

For example, the probability model and/or the context model may be determined based on 1) whether an affine mode is performed for the target block and/or 2) an affine mode indicator.

For example, different probability models and/or different context models may be determined for a case where the affine mode is performed on the target block and a case where the affine mode is not performed on the target block.

For example, for the case where the value of the affine mode indicator of the target block is 0 (or false) and the value of the affine mode indicator of the target block is 1 (or true), different probability models and /or Different context models may be determined respectively.

For example, the probabilistic model and/or the context model may be determined based on whether the target block satisfies an activation condition of bilateral matching or a part of the above activation condition.

For example, different probability models and/or different context models may be determined for a case in which the target block satisfies the activation condition for both matching and a case in which the target block does not satisfy the activation condition for both matching.

For example, the probability model and/or the context model may be determined based on whether the target block satisfies an activation condition of template matching or a part of the activation condition.

For example, different probability models and/or different context models may be determined for the case where the target block satisfies the template matching activation condition and the case where the target block does not satisfy the template matching activation condition.

The pre-defined conditions are 1) the size of the target block, 2) the size of the motion vector, 3) the enabling condition of template matching, 3) the enabling condition of bilateral matching, 4) the direction of inter prediction, and 5) the inter prediction condition. A condition to use one or more of the following: an index of weighted biprediction, 6) an index of adaptive motion vector resolution (AMVR), 7) an indicator of overlapped block motion compensation mode, and 8) an indicator of local illumination compensation mode. can be However, the pre-defined condition is not limited to the condition using the information listed above.

Inter-weighted biprediction may determine a combination of weights of reference blocks in units of coding blocks when generating a target block using a L0-direction reference block and an L1-direction reference block. For example, the weight of each reference block may be determined by signaling/encoding/decoding an index to a pre-defined table including the weights.

The overlapped block motion compensation mode may be a mode in which at least two prediction blocks are generated and a weighted sum of the prediction blocks is used as a final prediction block. A weighted sum may be applied to a part of a block or to an entire block. For example, a part of a block may be a set of pixels or a set of positions corresponding to a boundary surface of the block.

The local illumination compensation mode derives at least one of a weight and an offset by calculating a correlation between a template of a target block and a template of a reference block, and multiplies part or all of the target block by at least one of the derived weight and the derived offset, or It can be an additive mode.

In embodiments, description that a specific mode is not performed in the target block may mean that an indicator for a specific mode of the target block has a specific value. Certain values can be 0 or false.

In embodiments, description that a specific mode is performed in the target block may mean that an indicator for the specific mode of the target block has a specific value. A specific value can be 1 or true.

In embodiments, description that an indicator for a specific mode of a target block has a specific value may mean that the specific mode is not performed for the target block. Certain values can be 0 or false.

In embodiments, description that an indicator for a specific mode of a target block has a specific value may mean that the specific mode is performed for the target block. A specific value can be 1 or true.

In embodiments, whether a motion information search method is performed for a target block may be determined by signaling/encoding/decoding for an indicator DMVDMODE_FLAG.

In embodiments, the description that DMVDMODE_FLAG is the first value may mean that the motion information search method is not performed for the target block. The first value can be 0 or false.

In embodiments, the description that DMVDMODE_FLAG is the second value may mean that the motion information search method is performed for the target block. The second value may be 1 or true.

In embodiments, the description that the motion information search method for the target block is not performed may mean that DMVDMODE_FLAG for the target block is a specific value. Certain values can be 0 or false.

In embodiments, description that the motion information search method for the target block is performed may mean that DMVDMODE_FLAG for the target block is a specific value. A specific value can be 1 or true.

For example, a motion information search method for a target block may be performed only when the target block satisfies a pre-defined condition.

For example, signaling/encoding/decoding for DMVDMODE_FLAG may be performed only when the target block satisfies a pre-defined condition.

For example, a motion information search method for a target block may be performed whenever the target block satisfies a pre-defined condition.

The pre-defined conditions are 1) size of target block, 2) size of motion vector, 3) template matching activation condition, 4) bilateral matching activation condition, 5) inter prediction direction, 6) inter weighted biprediction It may be a condition to use one or more of an index, 7) an index of adaptive motion vector resolution, 8) an indicator of an overlapped block motion compensation mode, and 8) an indicator of a local illumination compensation mode. Alternatively, the pre-defined condition may be a condition using coding parameters. However, the pre-defined conditions are not limited to conditions using only the information listed above.

For example, inter-weighted biprediction may be used for blocks. At this time, the motion information search method may be performed for a block only when the weight of the block in the L0 direction and the weight of the block in the L1 direction are the same. In other words, the motion information retrieval method can be performed only for blocks having the same weights. Here, weights may be weights in inter-weighted biprediction.

For example, signaling/encoding/decoding of an inter-weighted biprediction index may be performed only when the motion information search method is not performed for the target block.

For example, DMVDMODE_FLAG may be an indicator of a motion information search method. Signaling/encoding/decoding of an inter-weighted biprediction index can be performed only when DMVDMODE_FLAG for a target block is a specific value. For example, a particular value can be 0 or false.

Alternatively, it is possible to encode/decode whether the motion information search method is performed only when the weights of the inter-weighted bi-prediction in the L0 and L1 directions of the target block are the same.

When the motion information search method is performed on the target block, the weight for the L0 direction and the weight for the L1 direction of inter-weighted biprediction may be set to be the same.

For example, the motion information retrieval method for the target block may be performed only when the adaptive motion vector resolution is not used or when the adaptive motion vector resolution is the same as the default resolution.

Signaling/encoding/decoding of an adaptive motion vector resolution index may be performed only when the motion information search method is not performed for the target block.

Alternatively, information indicating whether the motion information search method is performed may be signaled/encoded/decoded only when the adaptive motion vector resolution is not used for the target block or when the adaptive motion vector resolution is equal to the basic resolution.

In embodiments, the basic resolution may mean a motion vector resolution when adaptive motion vector resolution is not applied.

When the motion information search method is performed for the target block, the adaptive motion vector resolution may not be used for the target block, and the adaptive motion vector resolution may be set to be the same as the basic resolution.

For example, whether to perform the motion information search method may be determined based on whether bi-directional inter prediction is performed on a block. The motion information search method may be performed only for blocks on which bidirectional inter prediction is performed.

Signaling/encoding/decoding for inter-prediction directions may be performed only when the motion information search method is not performed for the target block.

For example, information indicating whether the motion information search method is performed may be signaled/encoded/decoded only when biprediction is used for the target block.

For example, when a motion information search method is performed on a target block, bi-directional inter prediction may be performed on the target block.

For example, the motion information search method can be performed only when the affine mode is not used for the target block. Alternatively, the motion information search method may be performed only when an indicator indicating whether the affine mode is performed for the target block has a value of 0 or false.

For example, signaling/encoding/decoding of an indicator indicating whether the affine mode is performed may be performed only when the motion information search method is not performed for the target block.

For example, an indicator indicating whether the motion information search method is performed only when the affine mode is not performed for the target block may be signaled/encoded/decoded.

For example, the motion information search method may be performed only when the first direction and the second direction are different from each other. The first direction may be a direction from the target image to L0_REFPIC_MINPOC. The second direction may be a direction from the target image to L1_REFPIC_MINPOC. Alternatively, the motion information search method may be performed only when the first direction and the second direction are different from each other and the first POC interval and the second POC interval are different from each other. The first POC interval may be the difference between the POC of the target image and the POC of LO_REFPIC_MINPOC. The second POC interval may be the difference between the POC of the target image and the POC of L1_REFPIC_MINPOC.

For example, when a motion information search method is performed on a target block, signaling/encoding/decoding of a reference picture index may be omitted. Signaling/encoding/decoding of the reference picture index may be performed only when the motion information search method is not performed for the target block.

When the motion information search method is performed on the target block, L0-direction and L1-direction reference images of the target block may be L0_REFPIC_MINPOC and L1_REFPIC_MINPOC, respectively.

For example, in performing the motion information search method, signaling/encoding/decoding of DMVDMODE_FLAG may be performed only when the first direction and the second direction are different from each other. The first direction may be a direction from the target image to L0_REFPIC_MINPOC. The second direction may be a direction from the target image to L1_REFPIC_MINPOC. Alternatively, in performing the motion information search method, signaling/encoding/decoding of DMVDMODE_FLAG may be performed only when the first direction and the second direction are different from each other and the first POC interval and the second POC interval are different from each other. The first POC interval may be the difference between the POC of the target image and the POC of LO_REFPIC_MINPOC. The second POC interval may be the difference between the POC of the target image and the POC of L1_REFPIC_MINPOC.

L0_REFPIC_MINPOC may be a reference picture having the smallest POC interval among reference pictures in the L0 direction reference picture list. The POC interval of the reference image may be a difference between the POC of the target image and the POC of the reference image.

The L1_REFPIC_MINPOC may be a reference picture having the smallest POC interval among reference pictures in the L1 direction reference picture list. The POC interval of the reference image may be a difference between the POC of the target image and the POC of the reference image.

Whether or not the motion information retrieval method is performed on the target block depends on sequence level, picture level, tile level, tile group level, slice level, coding tree unit (CTU) level, and coding unit (CU) level. and a prediction unit (PU) level. However, the level at which the above determination is made is not limited to the levels listed above.

For example, information indicating whether a motion information search method is performed may be first signaled/encoded/decoded when mergeFlag is signaled/encoded/decoded and then the signaled/encoded/decoded mergeFlag is a first value. mergeFlag may be an indicator indicating whether merge mode is used for a target block. The first value can be 0 or false. When mergeFlag is the first value, AMVP mode may be performed on the target block.

For example, information indicating whether a motion information search method is performed is first signaled/encoded/decoded when mergeFlag is signaled/encoded/decoded and then the signaled/encoded/decoded mergeFlag is the second value. mergeFlag may be an indicator indicating whether merge mode is used for a target block. The second value may be 1 or true. When mergeFlag is the first value, AMVP mode may be performed on the target block.

For example, final motion information of a motion information search method may be determined as motion information used for inter prediction of a target block. Alternatively, motion information derived from the final motion information of the motion information search method may be determined as motion information used for inter prediction of the target block. Encoding/decoding of an image may be performed using motion information used for inter prediction of a target block. For example, the encoding/decoding process may include one of inter prediction, transformation, inverse transformation, quantization, inverse quantization, entropy encoding/decoding, and in-loop filtering. However, encoding/decoding of an image is not limited to the processes listed above.

Motion information of a block to which a motion information search method is applied may be referred to from neighboring blocks. Motion information referred to by neighboring blocks may be at least one of initial motion information of a motion information retrieval method or motion information derived in each search step of a motion information retrieval method.

Construction of motion information candidate list and determination of final motion information

Below, a method of constructing a motion information candidate list in a motion information search method will be described.

For example, when a motion information candidate list is constructed in a motion information search method, a motion information candidate may be derived from initial motion information.

For example, when a motion information candidate list is constructed in the motion information search method, reordering may be performed on candidates in the motion information candidate list.

For example, each candidate in the motion information candidate list may be initial motion information or motion information generated by improving the initial motion information.

For example, a motion information candidate list may be constructed using only motion information satisfying a specific condition.

Here, the specific condition may include an activation condition (or part of an activation condition) of a decoder-end motion information derivation method.

For example, the decoder-end motion information derivation method may be one of decoder-end motion information derivation methods used in the final motion information determining step of the motion information search method.

For example, if the decoder-end motion information derivation method performed to construct the motion information candidate list is a decoder-end motion information derivation method that can be performed only for bi-directional prediction, the motion information candidate list is It may consist only of motion information candidates having bi-directional motion information.

For example, when the decoder-end motion information derivation method performed to construct the motion information candidate list is bilateral matching, the motion information candidate list may be composed of only motion information candidates having bi-directional motion information.

For example, when a motion information candidate list is constructed, the motion information candidate list may be constructed using only motion information satisfying conditions 1) to 3) below.

Condition 1) Motion information indicates bi-directional inter prediction.

Condition 2) The first POC difference and the second POC difference of the motion information are the same. Here, the first POC difference may be a difference between the POC of the target image and the POC of the LO direction. The second POC difference may be a difference between the POC of the target image and the POC of the L1 direction.

Condition 3) The first direction and the second direction of motion information are different from each other. Here, the first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

For example, when the motion information candidate list is constructed, the motion information candidate list may be constructed using only motion information satisfying conditions 1) and 3) above.

The motion information candidate list may be configured identically for the L0 and L1 directions. Alternatively, the motion information candidate list may be configured differently for the L0 and L1 directions.

For example, each motion information candidate in the motion information candidate list may include at least one of L0-direction motion information and L1-direction motion information. That is, each motion information candidate in the motion information candidate list may be L0-direction unidirectional motion information, L1-direction unidirectional motion information, or bidirectional motion information.

For example, one integrated motion information candidate list for the L0 direction and the L1 direction may be configured.

For example, when the motion information specified from the motion information candidate list includes both L0-direction motion information and L1-direction motion information, bi-directional inter prediction may be performed on the current block.

For example, when the reference direction of the motion information specified from the motion information candidate list is bi-directional, bi-directional inter prediction may be performed on the target block.

For example, when the motion information specified from the motion information candidate list includes only one motion information among L0-direction motion information and L1-direction motion information, unidirectional inter prediction may be performed on the target block.

For example, when the reference direction of the motion information specified from the motion information candidate list is unidirectional, unidirectional inter prediction for a direction indicated by the motion information specified for the current block may be performed.

For example, motion information candidate lists may be configured differently in the L0 direction and the L1 direction.

For example, the motion information candidate list for a specific direction TEMP_DIR may include unidirectional motion information and/or bidirectional motion information for the TEMP_DIR direction.

For example, motion information in a TEMP_DIR direction may be determined from motion information specified from a motion information candidate list for a specific direction TEMP_DIR.

If the specified motion information includes both motion information in the L0 direction and motion information in the L1 direction, only TEMP_DIR direction motion information among the L0 direction motion information and L1 direction motion information of the specified motion information is used to generate TEMP_DIR direction Motion information of may be determined.

If the reference direction of the specified motion information is bidirectional, motion information in the TEMP_DIR direction may be determined using only motion information in the TEMP_DIR direction of the specified motion information.

For example, in the LX direction, the motion information candidate list can be constructed using the same method as the method of constructing the merge candidate list, and in the L(1-X) direction, the method of constructing the AMVP mode candidate list and A motion information candidate list may be constructed using the same method.

For example, in the LX direction, the motion information candidate list can be constructed using the same method as the method of constructing the AMVP mode candidate list, and in the L(1-X) direction, the method of constructing the merge candidate list and A motion information candidate list may be constructed in the same manner.

In one embodiment, a method of constructing each list may be classified by 1) the order of referring candidates in the list, 2) motion information referred to from neighboring blocks, and 3) motion information candidates constituting the list.

In embodiments, a description that the manner of constructing the first list and the manner of constructing the second list are the same indicates the first detailed manners used to construct the first list and the second detailed manners used to construct the second list. It can mean that the methods are the same. Here, the detailed methods may include 1) an order of referring candidates in the list, 2) motion information referred to from neighboring blocks, and 3) motion information candidates constituting the list.

In embodiments, a description that the manner of constructing the first list and the manner of constructing the second list are the same indicates the first detailed manners used to construct the first list and the second detailed manners used to construct the second list. It may not be limited only in the sense that all of the schemes are identical to each other. In other words, at least one of 1) the order of referring candidates in the list, 2) motion information referred to from neighboring blocks, and 3) motion information candidates constituting the list is equally applied to the construction of the first list and the construction of the second list. If so, the two lists can be considered to have been constructed in the same manner. Here, the description that the motion information candidates constituting the list are the same may mean that all of the motion information candidates constituting the list are the same. Alternatively, the description that the motion information candidates constituting the list are the same may mean that some of the motion information candidates constituting the list are the same.

For example, when a motion information search method is performed and bidirectional inter prediction is performed on a target block, motion information on an LX direction may be determined by signaling/encoding/decoding of motion information.

For example, motion information for the L(1-X) direction may be determined by signaling/encoding/decoding the motion information.

For example, motion information for the L(1-X) direction may be determined using a decoder-end motion information derivation method without performing signaling/encoding/decoding.

For example, motion information for the L(1-X) direction may be determined using motion information for the LX direction.

For example, the motion information for the L(1-X) direction may be a candidate having the lowest matching cost among candidates in the motion information candidate list for the L(1-X) direction. Here, the matching cost of the candidate may be the matching cost of both the candidate and the motion information in the LX direction.

For example, motion information signaled/encoded/decoded may be the same as motion information signaled/encoded/decoded to determine motion information of unidirectional prediction in AMVP mode. Signaled/encoded/decoded motion information may include a part of signaled/encoded/decoded motion information to determine motion information of unidirectional prediction in AMVP mode.

For example, the signaled/encoded/decoded motion information may include one or more of a reference picture index, a motion vector differential, and a motion information index.

For example, when the target block satisfies the template matching activation condition, signaling/encoding/decoding may not be performed on the motion information index in the LX direction, and the motion information index in the LX direction is can be determined The LX-direction motion information index may be an index indicating a candidate having the lowest template matching cost among candidates in the LX-direction motion information candidate list.

For example, X can be 0, 1 or a positive integer. X may be a pre-defined value.

For example, the pre-defined value may be zero.

For example, the pre-defined value may be a value indicating a direction having a lower matching cost among the L0 direction and the L1 direction. Here, the matching cost may be a matching cost for motion information.

For example, the pre-defined value may be a value indicating a direction having a higher matching cost among the L0 direction and the L1 direction. Here, the matching cost may be a matching cost for motion information.

X may be determined through signaling/encoding/decoding.

For example, when the motion information search method is performed, if bi-directional inter prediction is used for the target block, motion information or a part of the motion information may be signaled/encoded/decoded only in the LX direction.

For example, motion information for the L(1-X) direction may be determined through signaling/encoding/decoding of the motion information.

For example, motion information for the L(1-X) direction may be determined using a decoder-end motion information derivation method.

For example, motion information for the L(1-X) direction may be determined using motion information for the LX direction.

For example, the motion information for the L(1-X) direction may be a candidate having the lowest matching cost among candidates in the motion information candidate list for the L(1-X) direction. Here, the matching cost of the candidate may be the matching cost of both the candidate and the motion information in the LX direction.

For example, when the target block satisfies the template matching activation condition, the motion information index in the LX direction may be determined using the template matching cost. The LX-direction motion information index may be an index indicating a candidate having the lowest template matching cost among candidates in the LX-direction motion information candidate list.

For example, when a motion information candidate list is constructed in the motion information search method, the decoder-end motion information derivation method may be performed only when the motion information index is less than or equal to DMVD_IDXTHRES.

For example, when a motion information candidate list is constructed in the motion information search method, the decoder-end motion information derivation method may be performed only when the motion information index is greater than or equal to DMVD_IDXTHRES.

DMVD_IDXTHRES can be 0, 1, 2 or a positive integer.

DMVD_IDXTHRES may be a pre-defined value.

For example, when an initial motion information candidate list is constructed in a motion information search method, an initial motion information candidate list is based on whether a target block satisfies an activation condition (or part of an activation condition) of a specific decoder-end motion information derivation method. The maximum size of the motion information candidate list may be determined.

For example, when an improved motion information candidate list is determined in a motion information search method, based on whether a target block satisfies an activation condition (or part of an activation condition) of a specific decoder-end motion information derivation method, The maximum size of the enhanced motion information candidate list may be determined.

For example, when a target block satisfies an activation condition of a specific decoder-end motion information derivation method, the maximum size of the motion information candidate list may be MAXIDX_ENABLED. If the target block does not satisfy the activation condition of the specific decoder-end motion information derivation method, the maximum size of the motion information candidate list may be MAXIDX_DISABLED. The motion information candidate list may be an initial motion information candidate list or an improved motion information candidate list.

For example, when a motion information candidate list is constructed in a motion information search method, motion information based on whether a target block satisfies an activation condition (or part of an activation condition) of a specific decoder-end motion information derivation method. The number of candidates constituting the candidate list may be determined.

For example, when a motion information candidate list is constructed in a motion information search method, based on whether a target block satisfies an activation condition (or part of an activation condition) of a specific decoder-end motion information derivation method, motion The number of candidates constituting the information candidate list may be determined.

For example, in a motion information search method, motion information may be specified from a motion information candidate list using a motion information index. If the target block satisfies the activation condition of the specific decoder-end motion information derivation method, the motion information index may have a value within the range from 0 to MAXIDX_ENABLED. If the target block does not satisfy the activation condition of the specific decoder-end motion information derivation method, the motion information index may have a value within the range from 0 to MAXIDX_DISABLED. The motion information candidate list may be an initial motion information candidate list or an improved motion information candidate list.

For example, in a motion information search method, motion information may be specified from a motion information candidate list using a motion information index. A range for a motion information index value and/or a maximum value of a motion information index may be determined based on whether the target block satisfies an activation condition (or part of the activation condition) of a specific decoder-end motion information derivation method. there is. The motion information candidate list may be an initial motion information candidate list or an improved motion information candidate list.

For example, in the motion information search method, a motion information candidate list may be configured. When a target block satisfies the activation condition of a specific decoder-end motion information derivation method, an initial motion information candidate list may be constructed using MAXIDX_ENABLED motion information candidates. When the target block does not satisfy the activation condition of the specific decoder-end motion information derivation method, an initial motion information candidate list may be constructed using MAXIDX_DISABLED motion information candidates. The motion information candidate list may be an initial motion information candidate list or an improved motion information candidate list.

For example, in a motion information search method, motion information may be specified from a motion information candidate list using a motion information index. If the target block satisfies the activation condition of the specific decoder-end motion information derivation method, the maximum value of the motion information index may be a value within the range from 0 to MAXIDX_ENABLED. If the target block does not satisfy the activation condition of the specific decoder-end motion information derivation method, the maximum value of the motion information index may be a value within the range from 0 to MAXIDX_DISABLED. The motion information candidate list may be an initial motion information candidate list or an improved motion information candidate list.

A specific decoder-end motion information derivation method may be pre-defined.

A specific decoder-end motion information derivation method may be one of template matching and bilateral matching. However, the specific decoder-end motion information derivation method is not limited to the methods listed above.

MAXIDX_ENABLED may be a pre-defined value and may be 0, 1, 2, 3, 6, 12, 48, 96 or a positive integer.

MAXIDX_DISABLED can be a pre-defined value and can be 0, 1, 2, 3, 6, 12, 48, 96 or a positive integer.

32 is a first flowchart illustrating a method of selecting a method of constructing a motion information candidate list according to an example.

33 is a second flowchart illustrating a method of selecting a method of constructing a motion information candidate list according to an example.

34 is a third flowchart illustrating a method of selecting a method of constructing a motion information candidate list according to an example.

35 is a fourth flowchart illustrating a method of selecting a method of constructing a motion information candidate list according to an example.

First, refer to FIG. 32 .

In step 3210, it may be determined whether the motion information search method of the embodiment is performed.

When the motion information search method of the embodiment is performed, step 3220 may be performed.

When the motion information search method of the embodiment is not performed, step 3290 may be performed.

At step 3220, the value of MvdL1ZeroFlag may be checked.

If MvdL1ZeroFlag is true, step 3230 may be performed.

If MvdL1ZeroFlag is false, step 3240 may be performed.

At step 3230, mergeDir may be set to 1. The L1-direction motion information candidate list may be constructed in the same way as the merge candidate list.

At step 3240, the value of XZeroFlag may be checked.

If XZeroFlag is true, step 3250 may be performed.

If XZeroFlag is false, step 3260 may be performed.

At step 3250, mergeDir may be set to 1. The L1-direction motion information candidate list may be constructed in the same way as the merge candidate list.

At step 3260, mergeDir may be set to zero. The motion information candidate list in the L0 direction may be constructed in the same way as the merge candidate list.

In step 3290, prediction modes other than the motion information search method may be used.

Next, refer to FIG. 33 .

Steps 3210, 3220, 3230, 3240, 3250, 3260, and 3290 described above with reference to FIG. 32 may correspond to steps 3310, 3320, 3330, 3340, 3350, 3360, and 3390 of FIG. 33, respectively. . Redundant descriptions are omitted.

At step 3330, mergeDir may be set to 1. The L1-direction motion information candidate list may be constructed in the same way as the merge candidate list.

At step 3350, mergeDir may be set to zero. The motion information candidate list in the L0 direction may be constructed in the same way as the merge candidate list.

At step 3360, mergeDir may be set to 1. The L1-direction motion information candidate list may be constructed in the same way as the merge candidate list.

Next, refer to FIG. 34.

Steps 3210, 3220, 3230, 3240, 3250, 3260, and 3290 described above with reference to FIG. 32 may correspond to steps 3410, 3420, 3430, 3440, 3450, 3460, and 3490 of FIG. 34, respectively. . Redundant descriptions are omitted.

The steps of FIG. 34 may further include step 3415.

According to the determination in step 3410, when the motion information search method of the embodiment is performed, step 3415 may be performed.

At step 3415, InterDir may be set to 3. If InterDir is 3, then both predictions can be performed. affineFlag can be set to false. smvdModeFlag can be set to false. BCW idx can be set to BCW_DEFALT. AMVR idx may be set to IMV_OFF.

After step 3415 is performed, step 3420 may be performed next.

Next, refer to FIG. 35 .

Steps 3410, 3415, 3420, 3430, 3440, 3450, 3460, and 3490 described above with reference to FIG. Each can respond. Redundant descriptions are omitted.

At step 3530, mergeDir may be set to 1. The L1-direction motion information candidate list may be constructed in the same way as the merge candidate list.

At step 3550, mergeDir may be set to zero. The motion information candidate list in the L0 direction may be constructed in the same way as the merge candidate list.

At step 3560, mergeDir may be set to 1. The L1-direction motion information candidate list may be constructed in the same way as the merge candidate list.

Below, the conditions, information and processing shown in FIGS. 32 to 34 will be described in more detail.

MVDL1ZeroFlag may be an indicator indicating whether signaling/encoding/decoding is performed on the information on the motion vector difference in the L1 direction. For example, MVDL1ZeroFlag may be determined in a picture parameter set. However, MVDL1ZeroFlag is not limited to being determined only in the picture parameter set.

For example, when MVDL1ZeroFlag is equal to the first value, it may indicate that signaling/encoding/decoding is performed for the L1-direction motion vector difference in the target image. The first value can be 0 or false. However, the first value is not limited to 0 or false.

For example, when MVDL1ZeroFlag is equal to the second value, it may indicate that signaling/encoding/decoding of an L1-direction motion vector differential is not performed in the target image. The second value may be 1 or true. However, the second value is not limited to 1 or true.

InterDir may be an indicator for a reference direction.

When InterDir is the first value, unidirectional inter prediction in the L0 direction may be performed with respect to the target block. The first value may be 1. However, the first value is not limited to 1.

When InterDir is the second value, unidirectional inter prediction in the L1 direction may be performed on the target block. The second value may be 2. However, the second value is not limited to 2.

When InterDir is the third value, bi-directional inter prediction may be performed on the target block. The third value may be 3. However, the third value is not limited to 3.

AffineFlag may be an indicator indicating whether affine mode is performed. smvdModeFlag is an indicator indicating whether symmetric motion vector differential mode is performed.

When affineFlag is equal to the first value, it may mean that the affine mode is not performed on the target block. The first value can be 0 or false. However, the first value is not limited to 0 or false.

When affineFlag is equal to the second value, it may mean that the affine mode is performed on the target block. The second value may be 1 or true. However, the second value is not limited to 1 or true.

When smvdModeFlag is equal to the first value, it may mean that the symmetric motion vector differential mode is not performed on the target block. The first value can be 0 or false. However, the first value is not limited to 0 or false.

When smvdModeFlag is equal to the second value, it may mean that the symmetric motion vector differential mode is performed on the target block. The second value may be 1 or true. However, the second value is not limited to 1 or true.

BCW idx may be an index of inter-weighted biprediction.

AMVR idx may be an index of adaptive motion vector resolution.

An L0-direction weight and an L1-direction weight in inter-weighted biprediction for a target block may be determined based on the BCW idx. The prediction block of the target block may be generated by applying a weighted sum using the L0-direction weight and the L1-direction weight to the L0-direction reference block and the L1-direction reference block of the target block. The prediction block of the target block may be a weighted sum of the L0-direction reference block and the L1-direction reference block. In the weighted sum, the L0-direction weight may be applied to the L0-direction reference block, and the L1-direction weight may be applied to the L1-direction reference block.

BCW idx equal to BCW_DEFAULT may mean that the L0-direction weight and the L1-direction weight of inter-weighted bi-prediction of the target block are equal.

BCW idx equal to BCW_DEFAULT may mean that the L0-direction weight and the L1-direction weight of inter-weighted bi-prediction of the target block are equal. Alternatively, BCW idx equal to BCW_DEFAULT may mean that inter-weighted biprediction is not performed in the target block.

BCW_DEFAULT can be 0, 2 or a positive integer.

IMV_OFF can be 0, 1 or a positive integer.

A resolution in the adaptive motion vector resolution of the target block may be determined based on the AMVR idx.

AMVR idx equal to IMV_OFF may mean that adaptive motion vector resolution is not used.

AMVR idx equal to IMV_OFF may mean that the resolution of the adaptive motion vector resolution is the same as the basic resolution. Here, the basic resolution may mean a motion vector resolution when adaptive motion vector resolution is not applied.

Other modes may mean other prediction methods and/or other prediction modes other than the motion information search method.

For example, when the motion information candidate list for the L0 direction and the motion information candidate list for the L1 direction are configured differently from each other, signaling/encoding/decoding is performed according to the methods described with reference to FIGS. 34 to 35 It can be.

In this case, XZeroFlag may be an indicator used to determine a motion information candidate list construction method for each direction of the L0 direction and the L1 direction.

“mergeDir = 0” or “L0 is a merge” may mean that the motion information candidate list in the L0 direction is constructed according to the same manner as the merge candidate list is constructed. At this time, the L1-direction motion information candidate list may be constructed according to the same method as the method of constructing the AMVP mode candidate list.

“mergeDir = 1” or “L1 is merge” may mean that the L1-direction motion information candidate list is constructed according to the same method as the merge candidate list. At this time, the motion information candidate list in the L0 direction may be constructed according to the same method as the method of constructing the AMVP mode candidate list.

For example, motion information or part of the motion information may be signaled/encoded/decoded only in the L(1 - XZeroFlag) direction of the L0 direction and the L1 direction. In this case, signaling/encoding/decoding may be performed according to the methods described with reference to FIGS. 34 to 35 .

XZeroFlag can be signaled/encoded/decoded.

XZeroFlag may be an indicator indicating whether signaling/encoding/decoding is performed on which direction of motion information among the L0-direction motion information and the L1-direction motion information.

“mergeDir = 0” or “L0 is a merge” may indicate that signaling/encoding/decoding is performed only for motion information in the L1 direction among motion information in the L0 direction and motion information in the L1 direction.

“mergeDir = 1” or “L1 is a merge” may indicate that signaling/encoding/decoding is performed only for motion information in the L0 direction among motion information in the L0 direction and motion information in the L1 direction.

In embodiments, the motion information candidate for each direction may include at least one of motion information in the L0 direction and motion information in the L1 direction.

For example, the motion information candidate in the L0 direction may include motion information in the L0 direction or motion information in the L0 and L1 directions.

For example, even if the motion information candidate includes both motion information in the L0 direction and motion information in the L1 direction, only motion information in the L0 direction of the motion information candidate may be used. Even though the motion information candidate includes both motion information in the L0 direction and motion information in the L1 direction, only motion information in the L1 direction of the motion information candidate may be used.

For example, reordering may be performed on the order of candidates in the motion information candidate list.

For example, reordering can be performed using a decoder-end motion information derivation method. However, the method of reordering is not limited to the decoder-end motion information derivation method.

For example, the same decoder-end motion information derivation method can be applied to the L0 direction and the L1 direction.

For example, different decoder-end motion information derivation methods may be applied to the L0 direction and the L1 direction, respectively.

For example, candidates of the motion information candidate list may be rearranged in an ascending order of template matching costs in the LX direction. In the L(1-X) direction, candidates in the motion information candidate list may be reordered in ascending order of matching costs on both sides. X can be 0, 1 or a positive integer. Information representing X may be signaled/encoded/decoded.

For example, candidates in the motion information candidate list may be reordered in ascending order of matching costs of the above candidates.

For example, the matching cost may be a template matching cost.

For example, motion information candidate lists for a plurality of directions may be configured differently from each other. Here, when the motion information MVP_LX for the LX direction is specified, candidates of the L(1-X) direction motion information candidate list can be reordered using MVP_LX.

For example, each motion information candidate in the motion information candidate list may include at least one of L0-direction motion information and L1-direction motion information. That is, each motion information candidate in the motion information candidate list may be L0-direction unidirectional motion information, L1-direction unidirectional motion information, or bidirectional motion information.

For example, when the motion information specified from the motion information candidate list includes both L0-direction motion information and L1-direction motion information, bi-directional inter prediction may be performed on the current block. For example, when the reference direction of the motion information specified from the motion information candidate list is bi-directional, bi-directional inter prediction may be performed on the target block.

For example, when the motion information specified from the motion information candidate list includes only one motion information among L0-direction motion information and L1-direction motion information, unidirectional inter prediction may be performed on the target block. For example, when the reference direction of the motion information specified from the motion information candidate list is unidirectional, unidirectional inter prediction for a direction indicated by the motion information specified for the current block may be performed.

For example, MVP_L(1-X) _i may indicate the i-th motion information candidate of the L(1-X) direction motion information candidate list. For each MVP_L(1-X) _i , a matching cost for motion information (MVP_LX, MVP_L(1-X) _i ) or motion information (MVP_L(1-X) _i , MVP_LX) may be calculated. Candidates in the L(1-X) direction motion information candidate list may be reordered in ascending order of matching cost. Here, the reordering method is not limited to the above method.

X can be 0, 1 or a positive integer.

Information representing X may be signaled/encoded/decoded.

For example, MVP_LX _i may indicate the i-th candidate motion information included in the motion information candidate list in the LX direction.

For example, motion information candidate lists for a plurality of directions may be configured differently from each other. In this case, when motion information MVP_LX for the LX direction is specified, an L(1-X) direction motion information candidate list may be reconstructed using MVP_LX.

For example, MVP_L(1-X) _i may indicate each motion information candidate in the motion information candidate list in the L(1-X) direction. For each MVP_L(1-X) _i , the following processing may be applied. When motion information (MVP_LX, MVP_L(1-X) _i ) or motion information (MVP_L(1-X) _i , MVP_LX) does not satisfy a specific condition, MVP_L from the motion information candidate list in the L(1-X) direction (1-X) _i can be eliminated. At this time, X may be 0, 1 or a positive integer. Information representing X may be signaled/encoded/decoded. The reconstruction method is not limited to the above method. For example, the aforementioned specific condition may be an activation condition (or part of an activation condition) of a decoder-level motion information derivation method.

MVP_L(1-X) _i may represent each motion information candidate in the motion information candidate list in the L(1-X) direction. For each MVP_L(1-X) _i, the following processing may be applied. When motion information (MVP_LX, MVP_L(1-X) _i ) or motion information (MVP_L(1-X) _i , MVP_LX) satisfies a specific condition, MVP_L(1-X) _i moves in the L(1-X) direction. can be used as a motion information candidate for At this time, X may be 0, 1 or a positive integer. Information representing X may be signaled/encoded/decoded. The reconstruction method is not limited to the above method. For example, the aforementioned specific condition may be an activation condition (or part of an activation condition) of a decoder-level motion information derivation method.

For example, motion information candidate lists for a plurality of directions may be configured differently from each other. In this case, when motion information MVP_LX for the LX direction is specified, an L(1-X) direction motion information candidate list may be constructed using MVP_LX.

The motion information candidate list in the L(1-X) direction always includes motion information (MVP_LX, MVP_L(1-X) _i ) or motion information (MVP_L(1-X) _i for each motion information candidate MVP_L(1-X) i. X) _i , MVP_LX) may be configured to satisfy a specific condition. Here, MVP_L(1-X) _i may represent each motion information candidate in the motion information candidate list in the L(1-X) direction. X can be 0, 1 or a positive integer. Information representing X may be signaled/encoded/decoded. The reconstruction method is not limited to the above method. For example, the aforementioned specific condition may be an activation condition (or part of an activation condition) of a decoder-level motion information derivation method.

Information of neighboring blocks may be referred to in order to determine each motion information candidate MVP_L(1-X) _i . For each motion information candidate MVP_L(1-X) _i , the motion information (MVP_LX, MVP_L(1-X) _i ) or the motion information (MVP_L(1-X) _i , MVP_LX) always satisfies a specific condition. A motion information candidate list in the L(1-X) direction may be constructed by using only the neighboring blocks that do. For example, the aforementioned specific condition may be an activation condition (or part of an activation condition) of a decoder-level motion information derivation method.

For example, the aforementioned specific conditions are 1) whether bi-directional inter prediction is used for the target block, 2) whether the first POC difference and the second POC difference are the same, and 3) whether the first direction and the second direction are different from each other. It may be a condition related to whether or not The first POC difference may be a difference between the POC of the target image and the POC of the reference image in the L0 direction. The second POC difference may be a difference between the POC of the target image and the POC of the reference image in the L1 direction. The first direction may be a direction from the target image to the reference image in the L0 direction. The second direction may be a direction from the target image to the reference image in the L1 direction.

For example, an explanation that satisfies a specific condition is that 1) bi-directional inter prediction is used for the target block, 2) the first POC difference and the second POC difference are the same, and 3) the first direction and the second direction are different from each other. can mean

For example, the aforementioned specific conditions may be conditions related to 1) whether bi-directional inter prediction is used for the target block and 2) whether the first direction and the second direction are different from each other. The first direction may be a direction from the target image to the reference image in the L0 direction. The second direction may be a direction from the target image to the reference image in the L1 direction.

For example, the description that a specific condition is met may mean that 1) bi-directional inter prediction is used for the target block, and 2) the first direction and the second direction are different from each other.

Motion information for the LX direction can be specified by using a pre-defined index. Alternatively, the motion information in the LX direction may be specified by signaling/encoding/decoding of a coding parameter for determining the motion information in the LX direction.

For example, the pre-defined index may be 0.

For example, the pre-defined index may be an index value of a motion information candidate having the lowest matching cost among motion information candidates in the LX direction.

For example, the pre-defined index may be an index value of a motion information candidate having the highest matching cost among motion information candidates in the LX direction.

36 illustrates reconstruction of a motion information candidate list according to an embodiment.

36 shows the reconstruction of the L(1-X) direction motion information candidate list.

The target image may be an image including the target block.

The past reference image may be an image having a smaller POC than the POC of the target image.

The future reference image may be an image having a larger POC than the POC of the target image.

The LX direction reference block can be specified by 1) a pre-defined index and/or 2). Motion information MVP_LX may be determined by signaling/encoding/decoding.

Candidates 0 to 4 may mean motion information candidate MVP_L1 _i in the previous L(1-X) direction of reconstruction. i may be an integer of 0 or more and 4 or less. The motion information candidate list in the L(1-X) direction may be reconstructed so that the motion information (MVP_L0, MVP_L1 _i ) satisfies the activation condition (or part of the activation condition) of matching on both sides.

Accordingly, among the motion information candidates in the L(1-X) direction shown in FIG. 36, candidate 0 and candidate 1 may be removed from the motion information candidate list in the L(1-X) direction. This is because the direction of MVP_LX and the direction of candidate 0 to the reference picture are the same, and the direction of MVP_LX and the direction of candidate 1 to the reference picture are the same.

As a result, the motion information candidate list in the L(1-X) direction can be reconstructed to include only candidate 2, candidate 3, and candidate 4. Alternatively, as a result, the motion information candidate list in the L(1-X) direction may be reconstructed to include candidate 2, candidate 3, candidate 4, and at least one new motion information candidate.

In the embodiments, the case where the LX direction reference image is a past reference image has been described. The LX direction reference image may not be limited to past reference images. The LX direction reference image may be a future reference image.

In embodiments, conditions for activating both sides matching are 1) whether bi-directional inter prediction is used for the target block, 2) whether the first POC difference and the second POC difference are the same, and 3) whether the first direction and the second direction are It may be a condition related to whether they are different or not. The first POC difference may be a difference between the POC of the target image and the POC of the reference image in the L0 direction. The second POC difference may be a difference between the POC of the target image and the POC of the reference image in the L1 direction. The first direction may be a direction from the target image to the reference image in the L0 direction. The second direction may be a direction from the target image to the reference image in the L1 direction.

In embodiments, conditions for activation of matching on both sides may be conditions related to 1) whether bi-directional inter prediction is used for the target block and 2) whether the first direction and the second direction are different from each other. The first direction may be a direction from the target image to the reference image in the L0 direction. The second direction may be a direction from the target image to the reference image in the L1 direction.

A value representing X may be pre-defined.

For example, the pre-defined value may be zero.

For example, the pre-defined value may be a value indicating a direction having a lower matching cost among the L0 direction and the L1 direction. Here, the matching cost may be a matching cost for motion information.

For example, the pre-defined value may be a value indicating a direction having a higher matching cost among the L0 direction and the L1 direction. Here, the matching cost may be a matching cost for motion information.

A value representing X can be signaled/encoded/decoded.

Motion information for the LX direction can be specified by using a pre-defined index. Alternatively, the motion information in the LX direction may be specified by signaling/encoding/decoding of a coding parameter for determining the motion information in the LX direction.

For example, the pre-defined index may be 0.

For example, the pre-defined index may be an index value of a motion information candidate having the lowest matching cost among motion information candidates in the LX direction.

For example, the pre-defined index may be an index value of a motion information candidate having the highest matching cost among motion information candidates in the LX direction.

For example, the motion information candidate list may be reconstructed using only some motion information candidates among motion information candidates of the motion information candidate list. Alternatively, some of the candidates of the motion information candidate list may be removed from the list.

When such reorganization or removal is performed, the maximum value of the index may decrease. By decreasing the maximum value of the index, the number of bits required for index-indicating information may be reduced in a process of signaling/encoding/decoding of index-indicating information.

For example, a candidate having an index equal to or greater than MVPLIST_NUM may be removed from each motion information candidate list. At this time, MVPLIST_NUM may be pre-defined. Alternatively, different MVPLIST_NUMs may be used for motion information candidate lists. MVPLIST_NUM can be 1, 2 or a positive integer.

Determination of final motion information in the motion information search method

The final motion information determined in the motion information search method or motion information derived from the final motion information may be determined as motion information used for inter prediction of the target block. An image encoding/decoding process may be performed using motion information used for inter prediction of a target block. The encoding/decoding process may include one or more of inter prediction, transformation, inverse transformation, quantization, inverse quantization, entropy encoding/decoding, and in-loop filtering. However, encoding/decoding of video is not limited to the above processes.

For example, a reference block for a target block may be determined using final motion information determined through a motion information search method or motion information derived from the final motion information. An image encoding/decoding process may be performed using the determined reference block. The encoding/decoding process may include one or more of inter prediction, transformation, inverse transformation, quantization, inverse quantization, entropy encoding/decoding, and in-loop filtering. However, encoding/decoding of video is not limited to the above processes.

For example, the final motion information may refer to motion information determined from a motion information candidate list. Alternatively, the final motion information may refer to motion information derived from motion information determined from a motion information candidate list.

In the motion information search method, at least one piece of final motion information may be determined.

For example, final motion information may be determined from a motion information candidate list.

The motion information candidate list may mean an initial motion information candidate list or an improved motion information candidate list.

For example, final motion information may be determined using a motion information index.

For example, the motion information index may be a pre-defined value.

The pre-defined value may be the smallest index value, eg the pre-defined value may be 0.

When a pre-defined value is used, the amount of bits for signaling can be reduced. Encoding efficiency can be improved by reducing the amount of bits.

For example, the motion information index may be an index determined through signaling/encoding/decoding.

The motion information index may be determined through a rate-distortion optimization process. By such a decision, encoding efficiency of the target block may be improved.

For example, the motion information index may be an index derived using a decoder-end motion information derivation method.

The motion information index may be an index value of a motion information candidate having the smallest matching cost among motion information candidates in the motion information candidate list.

The decoder-end motion information derivation method may be one or more of bilateral matching and template matching. However, the type of decoder-end motion information derivation method is not limited to bilateral matching and/or template matching.

For example, the same decoder-end motion information derivation method can be used for the L0 and L1 directions.

For example, different decoder-end motion information derivation methods may be used for the L0 and L1 directions, respectively.

For example, when the motion information index is determined, the motion information index in the LX direction may be determined by comparing template matching costs of motion information candidates in the LX direction. The motion information index in the L(1-X) direction may be determined by comparing matching costs of both sides of motion information candidates in the L(1-X) direction.

X may be a pre-defined value.

X can be 0, 1 or a positive integer.

Information representing X may be signaled/encoded/decoded.

The cost of matching both sides of the motion information candidate in the L(1-X) direction may be the cost of matching both sides between the motion information candidate in the L(1-X) direction and the motion information in the LX direction.

For example, final motion information may be specified using a decoder-end motion information derivation method.

For example, the specified motion information may be a motion information candidate having the lowest matching cost among motion information candidates in a motion information candidate list.

The decoder-end motion information derivation method may be one or more of bilateral matching and template matching. However, the type of decoder-end motion information derivation method is not limited to bilateral matching and/or template matching.

For example, the same decoder-end motion information derivation method can be used for the L0 and L1 directions.

For example, different decoder-end motion information derivation methods may be used for the L0 and L1 directions, respectively.

For example, when the motion information is specified, the motion information in the LX direction can be specified using the template matching cost of the motion information candidate in the LX direction, and the motion information in the L(1-X) direction is L(1-X ) can be specified using the matching cost of both sides of the directional motion information candidate.

X can be 0, 1 or a positive integer.

Information representing X may be signaled/encoded/decoded.

The cost of matching both sides of the motion information candidate in the L(1-X) direction may be the cost of matching both sides between the motion information candidate in the L(1-X) direction and the motion information in the LX direction.

For example, when motion information is specified, a template matching cost may be used for the LX direction, and a bilateral matching cost may be used for the L(1-X) direction. X can be 0, 1 or a positive integer. Information representing X may be signaled/encoded/decoded.

The same final motion information determination method may be used for the L0 direction and the L1 direction.

For example, the final motion information in the L0 direction and the final motion information in the L1 direction may be determined by signaling/encoding/decoding a motion information index.

For example, the final motion information for the L0 direction and the final motion information for the L1 direction may be determined through signaling/encoding/decoding of the motion information index for each direction.

Different methods of determining final motion information may be used in the L0 direction and the L1 direction, respectively.

For example, for the LX direction, motion information MVP_LX may be determined using a pre-defined method and/or an index determined by signaling/encoding/decoding, and MVP_LX determines motion information in the L(1-X) direction. can be used for

For example, MVP_L(1-X) _i may indicate each motion information candidate in the motion information candidate list in the L(1-X) direction. For each MVP_L(1-X) _i, the following processing may be applied. Motion information (MVP_LX, MVP_L(1-X) _i ) or motion information (MVP_L(1-X) _i , MVP_LX) satisfies the activation condition (or part of the activation condition) of a specific decoder-end motion information derivation method. In this case, MVP_L(1-X) _i may be used to determine L(1-X) direction motion information.

The specific decoder-end motion information derivation method may be one of decoder-end motion information derivation methods used in the motion information search method of the target block. For example, the decoder-end motion information derivation method may be bilateral matching.

For example, the final motion information in the LX direction may be derived using only N candidates having the smallest index among candidates in the motion information candidate list in the LX direction. N can be 1, 2 or a positive integer. X can be 0, 1 or a positive integer. Information representing X may be signaled/encoded/decoded.

For example, the LX-direction motion information candidate list may be reconstructed to include only N motion information candidates having the smallest index among motion information candidates of the LX-direction motion information candidate list. N can be 1, 2 or a positive integer. X can be 0, 1 or a positive integer. Information representing X may be signaled/encoded/decoded.

For example, final motion information may be determined by using a pre-defined method for the LX _signal direction and/or signaling/encoding/decoding of motion information. Then, for the L (1-LX _signal ) direction, candidates in the motion information candidate list may be reordered in ascending order of matching costs for specific motion information. Here, when the LX _signal is 0, specific motion information may be (MVP_L0, MVP_L1 _i ). When the LX _signal is 1, specific motion information may be (MVP_L0 _i , MVP_L1). The final motion information in the L (1-LX _signal ) direction may be determined by a pre-defined method and/or signaling/encoding/decoding of the motion information. However, a method of determining motion information in each direction is not limited to the above method.

For example, the final motion information in the L (1-X _signal ) direction may be reconstructed to include only N motion information candidates having the smallest index among motion information candidates of a motion information candidate list in the L (1X _signal ) direction. . N can be 1, 2 or a positive integer. X can be 0, 1 or a positive integer. Information representing X may be signaled/encoded/decoded.

For example, the L (1-X _signal ) direction motion information candidate list is reconfigured to include only N motion information candidates having the smallest index among the motion information candidates of the L (1-X _signal ) direction motion information candidate list. It can be. N can be 1, 2 or a positive integer. X can be 0, 1 or a positive integer. Information representing X may be signaled/encoded/decoded.

X _signal may be a pre-defined value. Alternatively, the X _signal may be determined by signaling/encoding/decoding information representing the X _signal .

Processing applied to the final motion information of the motion information search method

One or more of the processes to be described below may be applied to the final motion information of the motion information search method. Motion information to which a process to be described later is applied may be used as final motion information of a motion information search method.

The final motion information can be refined using a decoder-end motion information derivation method.

The final motion information may be refined using a motion information offset.

For example, a motion vector difference may be added to final motion information.

For example, a motion information offset may be added to final motion information.

For example, the final motion information or part of the final motion information may be changed to the same value as the motion information offset.

For example, final motion information of a motion information retrieval method may be determined for each direction of the L0 direction and the L1 direction. Next, improved motion information may be generated by performing refinement on the final motion information in at least one of the L0 direction and the L1 direction using the decoder-end motion information derivation method. Next, the improved motion information may be used as final motion information.

For example, a direction in which motion information enhancement is applied may be a pre-defined direction.

For example, the direction in which motion information enhancement is applied may be a direction having a higher template matching cost among the L0 direction and the L1 direction. Here, the template matching cost in a specific direction may be the template matching cost of motion information in a specific direction.

For example, a direction in which motion information enhancement is applied may be a direction having a lower template matching cost among the L0 direction and the L1 direction. Here, the template matching cost in a specific direction may be the template matching cost of motion information in a specific direction.

For example, motion information improvement using a decoder-end motion information derivation method is always performed for both the final motion information of the motion information retrieval method in the L0 direction and the final motion information of the motion information retrieval method in the L1 direction. can

For example, information on a direction in which motion information is to be improved may be signaled/encoded/decoded.

For example, when the matching cost of the final motion information of the motion information search method is greater than or equal to THRES_FOR_AFTERREFINE, refinement on the final motion information may be performed using the decoder-end motion information derivation method, and the improved motion information corresponds to the final motion can be used as information. THRES_FOR_AFTERREFINE may be a product of the number of pixels in the target block and a specific value. A specific value can be 0, 1, 2, 4, 8 or any positive integer.

For example, the decoder-end motion information retrieval method used to improve the motion information of the motion information retrieval method may be at least one of template matching, bilateral matching, and light-flow based motion information prediction technology.

For example, the X _signal may be determined by a pre-defined method or signaling/encoded/decoded information. With respect to the determined X _signal , motion information in the direction of the LX _signal may be determined by 1) comparison with template matching cost and/or signaling/encoding/decoding of the motion information in the direction of the LX _signal . Next, among motion information candidates in the L (1-X _signal ) direction, a motion information candidate having the lowest matching cost may be set as motion information in the L (1-X _signal ) direction. Here, the matching cost of the motion information candidate may be the cost of matching both sides of the motion information in the LX _signal direction. Alternatively, two motion information candidates having the lowest matching cost among motion information candidates in the L (1-X _signal ) direction may be selected. Here, the matching cost of the motion information candidate may be the cost of matching both sides of the motion information in the LX _signal direction. One motion information candidate among the two selected motion candidates may be specified. Here, an index may be used to specify a motion information candidate. Signaling/encoding/decryption of the index may be performed. One specified motion information candidate may be determined as motion information in the L (1-X _signal ) direction.

For example, information signaled/encoded/decoded for motion information in the LX _signal direction may include one or more of a motion information index, a reference picture index, and an inter prediction indicator. However, information signaled/encoded/decoded for motion information in the LX _signal direction is not limited to the information listed above.

For example, a motion information candidate having the lowest template matching cost among motion candidates in the LX _signal direction may be specified as the motion information candidate in the LX _signal direction.

For example, when final motion information of a motion information retrieval method is determined using a decoder-end motion information derivation method, at least one of used decoder-end motion information derivation methods may be pre-defined.

For example, if the target block satisfies the activation condition of both sides matching, both sides matching may be specified as the decoder-end motion information derivation method. Template matching may be specified as a decoder-end motion information derivation method when the target block does not satisfy the activation condition of bilateral matching. However, the type of decoder-end motion information derivation method and the method for specifying the decoder-end motion information derivation method are not limited to the above examples.

In embodiments, conditions for activating both sides matching are: 1) bi-directional inter prediction is used for the target block, 2) the first POC difference and the second POC difference are the same, and 3) the first direction and the second direction are different. it could be Alternatively, the conditions for activating both sides matching are 1) whether bi-directional inter prediction is used for the target block, 2) whether the first POC difference and the second POC difference are the same, and 3) whether the first direction and the second direction are different. may include whether Here, the first POC difference may be a difference between the POC of the target image and the POC of the L0 direction reference image. The second POC difference may be a difference between the POC of the target image and the POC of the L1-direction reference image. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

In embodiments, activation conditions for matching on both sides may be: 1) bi-directional inter prediction is used for the target block, and 2) the first direction and the second direction are different from each other. Alternatively, conditions for activating both sides matching may include 1) whether bi-directional inter prediction is used for the target block and 2) whether the first direction and the second direction are different from each other. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

For example, when bidirectional inter prediction is used for a target block, in at least one search step among search steps of a specific decoder-end motion information derivation method, motion information improvement is performed only for the LX direction of the final motion information of the motion information search method. can be performed

X may be a pre-defined value.

For example, X may be a direction having a higher matching cost among the L0 direction and the L1 direction. Here, the matching cost in a specific direction may be a matching cost for motion information in a specific direction.

X can be 0, 1 or a positive integer.

Information representing X may be signaled/encoded/decoded.

For example, when final motion information is determined in the motion information search method, 1) whether a decoder-side motion information derivation method is performed, 2) the number of decoder-side motion information derivation methods used, and 3) a decoder-side motion information derivation method used. However, one or more types of motion information derivation methods may be determined based on a motion information index value. Alternatively, at least one of 1) whether a decoder-side motion information derivation method is performed, 2) the number of decoder-side motion information derivation methods used, and 3) a type of decoder-side motion information derivation method used is a motion information index value can be independent.

For example, when the motion information index for a specific direction is the first value, improvement of motion information using a decoder-end motion information derivation method for the specific direction may not be performed. The first value may be 0. However, the first value is not limited to 0.

For example, when the motion information index for a specific direction is the second value, motion information improvement using a decoder-end motion information derivation method for the specific direction may be performed. The second value may be 1, 2 or a positive integer. However, the second value is not limited to the values listed above.

For example, when the motion information index for a specific direction is a first value, motion information may be improved using a first decoder-end motion information derivation method for the specific direction. The first value may be 0. However, the first value is not limited to 0. The first decoder-end motion information derivation method may be bilateral matching. However, the first decoder-end motion information derivation method is not limited to bilateral matching.

For example, when the motion information index for a specific direction is the second value, motion information may be improved using the first decoder-end motion information derivation method for the specific direction. The second value can be 1, 2 or a positive integer. However, the second value is not limited to the values listed above. The second decoder-end motion information derivation method may be template matching. However, the second decoder-end motion information derivation method is not limited to template matching.

For example, when the index of motion information in a specific direction is the first value, matching costs for motion information in a specific direction may not be calculated. The first value may be 0. However, the first value is not limited to 0.

For example, when the index of motion information in a specific direction is the second value, matching costs for motion information in a specific direction may be calculated. The second value may be 1, 2 or a positive integer. However, the second value is not limited to the values listed above.

For example, when the index of motion information in a specific direction is a first value, a matching cost used to calculate matching costs for motion information in a specific direction may be a template matching cost. The first value may be 0. However, the first value is not limited to 0.

For example, when the index of motion information in a specific direction is a first value, matching costs used to calculate matching costs for motion information in a specific direction may be matching costs for both sides. The first value may be 0. However, the first value is not limited to 0.

For example, when motion information refinement is performed on final motion information in a motion information search method, 1) whether a decoder-end motion information derivation method is performed, 2) the number of decoder-end motion information derivation methods used, and 3) One or more types of used decoder-end motion information derivation methods may be determined based on a motion information index value. Alternatively, at least one of 1) whether a decoder-side motion information derivation method is performed, 2) the number of decoder-side motion information derivation methods used, and 3) a type of decoder-side motion information derivation method used is a motion information index value can be independent.

For example, when the motion information index for a specific direction is the first value, improvement of motion information using a decoder-end motion information derivation method may not be performed for the specific direction. The first value may be 0. However, the first value is not limited to 0.

For example, when the motion information index for a specific direction is the second value, motion information improvement using a decoder-end motion information derivation method for the specific direction may be performed. The second value may be 1, 2 or a positive integer. However, the second value is not limited to the values listed above.

For example, when the motion information index for a specific direction is a first value, motion information may be improved using a first decoder-end motion information derivation method for the specific direction. The first value may be 0. However, the first value is not limited to 0. The first decoder-end motion information derivation method may be bilateral matching. However, the first decoder-end motion information derivation method is not limited to bilateral matching.

For example, when the motion information index for a specific direction is the second value, motion information may be improved using the first decoder-end motion information derivation method for the specific direction. The second value can be 1, 2 or a positive integer. However, the second value is not limited to the values listed above. The second decoder-end motion information derivation method may be template matching. However, the second decoder-end motion information derivation method is not limited to template matching. For example, when the motion information index for a specific direction is the second value, the first decoder-end motion information derivation method for a specific direction Improvement of motion information may be performed using The second value can be 1, 2 or a positive integer. However, the second value is not limited to the values listed above. The second decoder-end motion information derivation method may be template matching. However, the second decoder-end motion information derivation method is not limited to template matching.

For example, in the motion information search method, when a motion information candidate list is constructed or when final motion information is determined, MVD_NUM_STEP motion vector differentials may be added to the motion information.

For example, when a motion information retrieval method is performed, a decoder-end motion information derivation method may be performed after motion vector difference is added to motion information.

The motion information to which the motion vector difference is added is 1) each candidate of the initial motion information candidate list, 2) each candidate of the improved motion information candidate list, 3) final motion information determined from the motion information candidate list, or 4) initial motion information or final motion information. When the decoder-side motion information derivation method is applied to the motion information, it may be one of motion information derived from search steps of the decoder-side motion information derivation method. However, the motion information to which the motion vector difference is added is not limited to the motion information described above.

MVD_NUM_STEP can be 0 or a positive integer.

MVD_NUM_STEP may be a pre-defined value.

For example, MVD_NUM_STEP may be determined as 0, 1, or a positive integer according to an activation condition of a decoder-level motion information derivation method.

For example, the decoder-end motion information derivation method may be at least one of bilateral matching and template matching. However, the decoder-end motion information derivation method is not limited to bilateral matching and/or template matching.

When MVD_NUM_STEP is 0, signaling/encoding/decoding of a motion vector differential for a target block may not be performed. That is, signaling/encoding/decoding of the motion vector differential for the target block can be performed only when MVD_NUM_STEP is equal to or greater than 1.

For example, signaling/encoding/decoding of the motion vector difference may be performed only when the motion information index of the target block is greater than or equal to MVD_IDXTHRES.

For example, signaling/encoding/decoding of the motion vector differential may be performed only when the motion information index of the target block is less than or equal to MVD_IDXTHRES.

For example, MVD_NUM_STEP motion vector differences may be added to the motion information only when the motion information search index is greater than or equal to MVD_IDXTHRES.

For example, MVD_NUM_STEP motion vector differences may be added to the motion information only when the motion information search index is less than or equal to MVD_IDXTHRES.

MVD_IDXTHRES can be 0, 1, 2 or a positive integer.

MVD_IDXTHRES may be a pre-defined value.

For example, a motion vector difference may be signaled/encoded/decoded only when a target block does not satisfy an activation condition of a specific decoder-end motion information derivation method.

The decoder-end motion information derivation method may be template matching and/or bilateral matching. However, the decoder-end motion information derivation method is not limited to template matching and/or bilateral matching.

In embodiments, activation conditions of a specific decoder-end motion information derivation method are: 1) bi-directional inter prediction is used for a target block, 2) the first POC difference and the second POC difference are the same, and 3) the first direction and The second direction may be different from each other. Alternatively, the activation conditions of the specific decoder-end motion information derivation method are 1) whether bi-directional inter prediction is used for the target block, 2) whether the first POC difference and the second POC difference are the same, and 3) the first direction and It may include whether the second directions are different from each other. Here, the first POC difference may be a difference between the POC of the target image and the POC of the L0 direction reference image. The second POC difference may be a difference between the POC of the target image and the POC of the L1-direction reference image. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

In embodiments, activation conditions of a specific decoder-end motion information derivation method may be that 1) bi-directional inter prediction is used for a target block, and 2) the first direction and the second direction are different from each other. Alternatively, the activation conditions of the specific decoder-end motion information derivation method may include 1) whether bi-directional inter prediction is used for the target block and 2) whether the first direction and the second direction are different from each other. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

The motion vector difference may be added only to motion information in the LX direction among motion information in the L0 direction and motion information in the L1 direction.

X can be 0, 1 or a positive integer.

X may be a pre-defined value.

X may be determined through signaling/encoding/decoding.

For example, when the motion vector difference in L0 and the motion vector difference in direction L1 are signaled/encoded/decoded respectively, the motion vector difference in each direction may be added to motion information in each direction.

For example, when only one motion vector difference MVD_ONLY is signaled/decoded, MVD_ONLY may be added to the L0-direction motion vector, and -MVD_ONLY may be added to the L1-direction motion vector.

For example, when only one motion vector difference MVD_ONLY is signaled/encoded/decoded, MVD_ONLY can be added only to the LX_MVD direction.

X_MVD can be 0 or 1. However, the value of X_MVD is not limited to 0 and/or 1.

X_MVD may be a pre-defined value.

For example, LX_MVD may be a direction having a lower matching cost among the L0 direction or the L1 direction. Here, the matching cost of a specific direction may be a matching cost of motion information in a specific direction.

For example, LX_MVD may be a direction having a higher matching cost among the L0 direction or the L1 direction. Here, the matching cost of a specific direction may be a matching cost of motion information in a specific direction.

X_MVD can be determined through signaling/encoding/decoding.

For example, a motion vector difference may be added only to motion information in the LX direction. Motion information in the L(1-X) direction may be improved by a decoder-end motion information derivation method using the motion information in the LX direction to which the motion vector difference is added.

The motion information in the improved L(1-X) direction may be motion information having the lowest matching cost among motion information within a search range of the decoder-side motion information derivation method. Here, the matching cost may be a cost of matching both sides with motion information in the LX direction.

The motion information in the improved L(1-X) direction may be motion information having the lowest template matching cost among motion information within a search range of the decoder-side motion information derivation method.

For example, when the motion information in the L(1-X) direction is improved by the decoder-end motion information derivation method, the motion information in the LX direction may not be improved.

For example, when motion information in the L(1-X) direction is improved by the decoder-end motion information derivation method, motion information in the LX direction may also be improved.

For example, when a decoder-side motion information derivation method is performed, enhancement may be performed only on motion information in the L(1-X) direction in some search steps among search steps of the decoder-side motion information derivation method. there is.

For example, when the decoder-side motion information derivation method is performed, enhancement may be performed only on motion information in the LX direction in some of the search steps of the decoder-side motion information derivation method.

X can be 0, 1 or a positive integer.

X may be a pre-defined value.

X may be determined through signaling/encoding/decoding.

For example, a motion vector difference may be added only to motion information in the LX direction. Next, motion information in the L0 direction and/or motion information in the L1 direction may be improved by the decoder-end motion information derivation method.

For example, motion information may be improved only in a direction having a lower template matching cost among the L0 direction and the L1 direction.

For example, motion information may be improved only in a direction having a higher template matching cost among the L0 direction and the L1 direction.

When the motion information in the L(1-X) direction is improved by the decoder-end motion information derivation method, the motion information in the LX direction can also be improved.

For example, when the decoder-side motion information derivation method is performed, enhancement may be performed only on motion information in the L(1-X) direction in some search steps of the decoder-side motion information derivation method.

For example, in performing the decoder-side motion information derivation method, only motion information improvement in the LX direction may be performed in some search steps of the decoder-side motion information derivation method.

X can be 0, 1 or a positive integer.

X may be a pre-defined value.

X may be determined through signaling/encoding/decoding.

For example, when four arithmetic operations using motion information offsets are applied to motion information of a target block, four arithmetic operations using motion information offsets may be applied to motion information only in the LX direction.

X can be 0, 1 or a positive integer.

X may be a pre-defined value.

X may be determined through signaling/encoding/decoding.

For example, when the motion information of the target block is changed to the same value as the motion information offset, only the motion information for the LX direction may be changed to the same value as the motion information offset.

X can be 0, 1 or a positive integer.

X may be a pre-defined value.

X may be determined through signaling/encoding/decoding.

37 illustrates setting of flags according to values of motion information indices according to an example.

For example, when the L0-direction motion information index and the L1-direction motion information index of the target block are signaled/encoded/decoded, one or more of IDENTICAL_MVP_FLAG and MVP_LX_FLAG may be signaled/encoded/decoded. IDENTICAL_MVP_FLAG may be an indicator indicating whether a motion information index in the L0 direction and a motion information index in the L1 direction are the same. MVP_LX_FLAG may be a motion information index in a pre-defined LX direction.

X in MVP_LX_FLAG can be 0, 1 or a positive integer.

In FIG. 37, a signaling/encoding/decoding method for a L0-direction motion information index and an L1-direction motion information index using IDENTICAL_MVP_FLAG and MVP_LX_FLAG is exemplified.

When signaling/encoding/decoding for IDENTICAL_MVP_FLAG and/or MVP_LX_FLAG is performed, signaling/encoding/decoding using a probability model or a context model may be performed.

When signaling/encoding/decoding is performed on the L0-direction motion information index and the L1-direction motion information index using the above syntax elements, the L0-direction motion information index and the L1-direction motion information index are often the same. , the above-described signaling/encoding/decoding method may have higher coding efficiency than a method of separately signaling/encoding/decoding the L0-direction motion information index and the L1-direction motion information index.

38 illustrates a first setting for an index according to values of motion information indexes according to an example.

39 illustrates a second setting for an index according to values of motion information indexes according to an example.

When the motion information index in the L0 direction and the motion information index in the L1 direction of the target block are signaled/encoded/decoded, after signaling/encoding/decoding of MVP_L01_IDX is performed, the motion information index in the L0 direction and the motion information index in the L1 direction from MVP_L01_IDX A combination of motion information indices of can be derived.

38 and 39 illustrate examples of signaling/encoding/decoding methods for motion information indexes in the L0 direction and the motion information index in the L1 direction using MVP_L01_IDX.

For example, values of combinations indicated by value 1 of MVP_L01_IDX may be determined through a pre-defined method. Here, the combination values may be a motion information index value in the L0 direction and a motion information index value in the L1 direction.

For example, the values of the combination indicated by the value 2 of MVP_L01_IDX may be determined through a pre-defined method. Here, the combination values may be a motion information index value in the L0 direction and a motion information index value in the L1 direction.

For example, the pre-defined method may be a method of comparing matching costs for motion information specified by combinations of motion information indices.

For example, a combination of motion information indices (a, b) may represent a combination in which a motion information index in the L0 direction is a and a motion information index in the L1 direction is b. A lower MVP_L01_IDX index may be assigned to a combination having a lower matching cost among combination (0, 1) and combination (1, 0).

When signaling/encoding/decoding for MVP_L01_IDX is performed, signaling/encoding/decryption using a probability model or a context model may be performed.

When signaling/encoding/decoding is performed on the L0-direction motion information index and the L1-direction motion information index using the above syntax elements, each of the L0-direction motion information index and the L1-direction motion information index has a lower index. If the case of having a value occurs frequently, the above-described signaling/encoding/decoding scheme will have higher coding efficiency than a scheme in which the L0-direction motion information index and the L1-direction motion information index are individually signaled/encoded/decoded. can

In performing signaling/encoding/decoding for MVP_L01_IDX, one or more of the following binarization, debinarization, and entropy encoding/decoding methods may be used.

- Signed 0-th order Exp_Golomb binarization/inverse binarization method (abbreviated as se(v))

- Signed k-th order Exp_Golomb binarization/inverse binarization method (abbreviated as sek(v))

- 0-th order Exp_Golomb binarization/inverse binarization method for unsigned positive integers (abbreviated as ue(v))

- k-th order Exp_Golomb binarization/inverse binarization method for unsigned positive integers (abbreviated as uek(v))

- Fixed-length binarization/inverse binarization methods (abbreviated as f(n))

- truncated Rice binarization/inverse binarization method or truncated unary binarization/inverse binarization method (abbreviated as tu(v))

- truncated binary binarization/inverse binarization method (abbreviated as tb(v))

- context-adaptive arithmetic encoding/decoding method (abbreviated as ae(v))

- bit string in bytes (abbreviated as b(8))

- signed integer binarization/inverse binarization method (abbreviated as i(n))

- Unsigned positive integer binarization/inverse binarization method (abbreviated as u(n)) ('u(n)' may mean fixed-length binarization/inverse binarization method.)

- Unary binarization/inverse binarization method

40 illustrates a first setting for an index according to values of reference picture indexes according to an example.

41 illustrates a second setting for an index according to values of reference picture indexes according to an example.

For example, in the motion information search method, reordering using a decoder-end motion information derivation method may be applied to candidates of a reference picture list.

For example, a reference block may be specified from a motion vector in the L0 direction and a reference picture in a reference picture list in the L0 direction. A reference block may be specified from a motion vector in the L1 direction and a reference picture in the reference picture list in the L1 direction. For each reference image list, the order of candidates of the reference image list may be sorted in ascending order of matching cost for a specified reference block.

For example, REFIDX_L01 may be signaled/encoded/decoded to signal/encode/decode the L0-direction reference picture index and the L1-direction reference picture index for the target block. A combination of a reference picture index in the L0 direction and a reference picture index in the L1 direction can be derived from REFIDX_L01.

40 and 41 illustrate examples of signaling/encoding/decoding methods for reference picture indexes in the L0 direction and reference picture indexes in the L1 direction using REFIDX_L01.

For example, values of a combination indicated by a value 1 of REFIDX_L01 may be determined through a pre-defined method. Here, the values of the combination may be a reference image index value in the L0 direction and a reference image index value in the L1 direction.

For example, the values of the combination indicated by the value 2 of REFIDX_L01 may be determined through a pre-defined method. Here, the values of the combination may be a reference image index value in the L0 direction and a reference image index value in the L1 direction.

For example, the pre-defined method may be a method of comparing matching costs for reference blocks in the L0 direction and reference blocks in the L1 direction specified by combinations of reference picture indices.

For example, the combination (a, b) of the reference picture index in the L0 direction and the reference picture index in the L1 direction may represent a combination in which the reference picture index in the L0 direction is a and the reference picture index in the L1 direction is b. A lower REFIDX_L01 may be assigned to a combination having a lower matching cost among combination (0, 1) and combination (1, 0).

When signaling/encoding/decoding for REFIDX_L01 is performed, signaling/encoding/decryption using a probability model or a context model may be performed.

When signaling/encoding/decoding is performed for the reference picture index in the L0 direction and the reference picture index in the L1 direction using the above-described syntax elements, each of the reference picture index in the L0 direction and the reference picture index in the L1 direction is a low index. If the case with a value occurs frequently, the above-described signaling/encoding/decoding scheme will have higher coding efficiency than the scheme of separately signaling/encoding/decoding the L0-direction reference picture index and the L1-direction reference picture index. can

In performing signaling/encoding/decoding of REFIDX_L01, one or more of the following binarization, debinarization, and entropy encoding/decoding methods may be used.

- Signed 0-th order Exp_Golomb binarization/inverse binarization method (abbreviated as se(v))

- Signed k-th order Exp_Golomb binarization/inverse binarization method (abbreviated as sek(v))

- 0-th order Exp_Golomb binarization/inverse binarization method for unsigned positive integers (abbreviated as ue(v))

- k-th order Exp_Golomb binarization/inverse binarization method for unsigned positive integers (abbreviated as uek(v))

- Fixed-length binarization/inverse binarization methods (abbreviated as f(n))

- truncated Rice binarization/inverse binarization method or truncated unary binarization/inverse binarization method (abbreviated as tu(v))

- truncated binary binarization/inverse binarization method (abbreviated as tb(v))

- context-adaptive arithmetic encoding/decoding method (abbreviated as ae(v))

- bit string in bytes (abbreviated as b(8))

- signed integer binarization/inverse binarization method (abbreviated as i(n))

- Unsigned positive integer binarization/inverse binarization method (abbreviated as u(n)) ('u(n)' may mean a fixed-length binarization/inverse binarization method.

- Unary binarization/inverse binarization methods.

42 illustrates setting of flags according to values of reference picture indexes according to an example.

For example, when a reference picture index in the L0 direction and a reference picture index in the L1 direction for the target block are signaled/encoded/decoded, one or more of IDENTICAL_REFIDX_FLAG and REFIDX_LX_FLAG may be signaled/encoded/decoded. IDENTICAL_REFIDX_FLAG may be an indicator indicating whether a reference picture index in the L0 direction and a reference picture index in the L1 direction are the same. MEFIDX_LX_FLAG may be a pre-defined reference picture index in the LX direction.

X in REFIDX_LX_FLAG can be 0, 1 or a positive integer.

In FIG. 42, a signaling/encoding/decoding method for a L0-direction reference picture index and an L1-direction reference picture index using IDENTICAL_REFIDX_FLAG and REFIDX_LX_FLAG is exemplified.

When signaling/encoding/decoding for IDENTICAL_REFIDX_FLAG and/or REFIDX_LX_FLAG is performed, signaling/encoding/decryption using a probability model or a context model may be performed.

When signaling/coding/decoding is performed on the L0-direction reference picture index and the L1-direction reference picture index using the above syntax elements, the L0-direction reference picture index and the L1-direction reference picture index are often the same. If this occurs, the above-described signaling/encoding/decoding method may have higher coding efficiency than a method of individually signaling/encoding/decoding the L0-direction reference picture index and the L1-direction reference picture index.

Determination of the sign of each component of the motion vector difference

For example, when a sign for each component of a motion vector difference is determined, a list of combinations may be constructed by a pre-defined method. Each combination in the list may be a combination of signs of components of the motion vector difference. Next, the signs of the components can be determined using the index specifying one of the combinations in the list.

An index for motion vector difference may be a pre-defined index. An index for motion vector difference may be an index having the lowest value (eg, 0). Alternatively, an index for motion vector difference may be signaled/encoded/decoded.

Each component of the components of the motion vector difference may be a component in a vertical direction or a component in a horizontal direction.

In performing signaling/encoding/decoding of an index for motion vector difference, signaling/entropy encoding/decoding using a probability model or a context model may be performed.

Here, an index for motion vector difference may be an index for determining a sign of motion vector difference.

When the sign for each component of the motion vector difference is determined, a list is constructed as described above, and information indicating an index for the list can be signaled/encoded/decoded. When such a list and index are used, a probability that an index having a low value is selected as an index for motion vector difference of a target block may increase. Thus, signaling/encoding/decoding efficiency can be improved.

In performing signaling/encoding/decoding of an index for motion vector difference, one or more of the following binarization, debinarization, and entropy encoding/decoding methods may be used.

- Signed 0-th order Exp_Golomb binarization/inverse binarization method (abbreviated as se(v))

- Signed k-th order Exp_Golomb binarization/inverse binarization method (abbreviated as sek(v))

- 0-th order Exp_Golomb binarization/inverse binarization method for unsigned positive integers (abbreviated as ue(v))

- k-th order Exp_Golomb binarization/inverse binarization method for unsigned positive integers (abbreviated as uek(v))

- Fixed-length binarization/inverse binarization methods (abbreviated as f(n))

- truncated Rice binarization/inverse binarization method or truncated unary binarization/inverse binarization method (abbreviated as tu(v))

- truncated binary binarization/inverse binarization method (abbreviated as tb(v))

- context-adaptive arithmetic encoding/decoding method (abbreviated as ae(v))

- bit string in bytes (abbreviated as b(8))

- signed integer binarization/inverse binarization method (abbreviated as i(n))

- Unsigned positive integer binarization/inverse binarization method (abbreviated as u(n)) ('u(n)' may mean fixed-length binarization/inverse binarization method.)

- Unary binarization/inverse binarization method

Magnitudes of motion vector difference components may be determined as (mag_X, mag_Y). When the magnitudes of the components are determined, a combination of components having the changed signs may be added to the motion vector difference candidate list while changing the sign of each component. The combination may include a vertical component of the motion vector difference and a horizontal component of the motion vector difference. Here, the magnitude sets of combinations may be equal. The combinational magnitude set may be the magnitudes of the components of the motion vector difference. Here, code sets of combinations may be different from each other. The set of codes of the combination may be the codes of the components of the motion vector difference.

For example, the motion vector difference candidates of the motion vector difference candidate list may include one or more of (mag_X, mag_Y), (-mag_X, mag_Y), (mag_X, -mag_Y), and (-mag_X, -mag_Y). . However, motion vector difference candidates are not limited to the combinations listed above.

Reordering of candidates in motion vector difference candidate list using decoder-end motion information derivation method

For example, reordering using a decoder-end motion information derivation method may be performed on candidates of the motion vector difference candidate list.

For example, candidates of the motion vector difference candidate list may be sorted in ascending order of matching cost.

For example, the matching cost may be a template matching cost and/or a bilateral matching cost. However, the matching cost is not limited to the template matching cost and/or both matching costs.

For example, when a target block satisfies an activation condition for both sides matching, a cost for both sides matching may be used. When the target block does not satisfy the activation condition of both sides matching, a template matching cost may be used.

A template used for calculating the matching cost may be changed based on motion information (or coding parameters) of the target block. Alternatively, the template used for calculating the matching cost may be constant regardless of motion information (or coding parameters) of the target block.

For example, when a specific mode is performed with respect to a target block, a matching cost may be calculated using only a template for a reference block in the L0 direction. When a specific mode is not performed with respect to the target block, a matching cost may be calculated using both the L0-direction template and the L1-direction template.

For example, the specific mode may be one of 1) symmetric motion vector difference mode, 2) merge mode including motion vector difference, or 3) affine merge mode including motion vector difference. However, the specific mode is not limited to the modes listed above.

In embodiments, a merge mode including motion vector differences (ie, MMVD mode) may be a merge mode that uses only restricted motion vector differences using a pre-defined table. The motion vector difference may be determined by signaling/encoding/decoding of an index for a table.

In embodiments, the table may be at least one of 1) a table for magnitudes of motion vector differences, 2) a table for signs of motion vector difference components, and 3) a table for directions of motion vector differences. For example, a table for directions of motion vector differences may include k candidate directions. k may be an integer. Each candidate can point in a direction corresponding to 2×i×π/K _MAX . K_MAX may be a pre-defined positive integer. i may be an integer greater than or equal to 0 and less than or equal to (k-1).

In embodiments, an affine merge mode including motion vector differences (ie, an affine MMVD mode) may be an affine merge mode that uses only restricted motion vector differences using a pre-defined table. A motion vector difference for at least one of the affine control point motion vectors of the affine mode may be signaled/encoded/decoded. The table may be the same as a pre-defined table of merge mode including motion vector difference. Alternatively, the table may be different from the pre-defined table of the merge mode including the motion vector difference.

For example, when a target block satisfies an activation condition (or part of an activation condition) of a specific decoder-end motion information derivation method, a matching cost may be calculated using only a template for a reference block in the L0 direction. . If the target block does not satisfy the activation condition (or part of the activation condition) of the specific decoder-end motion information derivation method, the matching cost may be calculated using both the template for the L0 direction and the template for the L1 direction. there is.

For example, when a target block does not satisfy an activation condition (or part of an activation condition) of a specific decoder-end motion information derivation method, a matching cost may be calculated using only a template for a reference block in the L0 direction. there is. If the target block satisfies the activation condition (or part of the activation condition) of the specific decoder-end motion information derivation method, the matching cost may be calculated using both the template for the L0 direction and the template for the L1 direction. .

A specific decoder-end motion information derivation method may be one or more of bilateral matching and template matching. However, the type of a specific decoder-end motion information derivation method is not limited to bilateral matching and/or template matching.

In embodiments, conditions for activating both sides matching are: 1) bi-directional inter prediction is used for the target block, 2) the first POC difference and the second POC difference are the same, and 3) the first direction and the second direction are different. it could be Alternatively, the conditions for activating both sides matching are 1) whether bi-directional inter prediction is used for the target block, 2) whether the first POC difference and the second POC difference are the same, and 3) whether the first direction and the second direction are different. may include whether Here, the first POC difference may be a difference between the POC of the target image and the POC of the L0 direction reference image. The second POC difference may be a difference between the POC of the target image and the POC of the L1-direction reference image. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

In embodiments, activation conditions for matching on both sides may be: 1) bi-directional inter prediction is used for the target block, and 2) the first direction and the second direction are different from each other. Alternatively, conditions for activating both sides matching may include 1) whether bi-directional inter prediction is used for the target block and 2) whether the first direction and the second direction are different from each other. The first direction may be a direction from the target image to the L0 direction reference image. The second direction may be a direction from the target image to the L1-direction reference image.

43 illustrates a method of determining a sign for each component of a motion vector difference according to an embodiment.

For example, MVP may indicate motion information of a target block.

For example, the motion vector difference candidates may be (mag_X, mag_Y), (-mag_X, mag_Y), (mag_X, -mag_Y), and (-mag_X, -mag_Y).

For example, motion information A, motion information B, motion information C, and motion information D may be generated by adding motion vector difference candidates to motion information of a target block.

For example, the motion information A may be the sum of the motion information of the target block and the first motion vector difference candidate (ie, (mag_X, mag_Y)). Motion information A may be the motion vector MVP + (-mag_X, -mag_Y).

For example, the motion information B may be the sum of the motion information of the target block and the second motion vector difference candidate (ie, (-mag_X, mag_Y)). Motion information B may be the motion vector MVP + (mag_X, -mag_Y).

For example, the motion information C may be the sum of the motion information of the target block and the third motion vector difference candidate (ie, (mag_X, -mag_Y)). Motion information C may be the motion vector MVP + (-mag_X, mag_Y). Motion information D may be the motion vector MVP + (mag_X, mag_Y).

For example, the motion information D may be the sum of the motion information of the target block and the fourth motion vector difference candidate (ie, (-mag_X, -mag_Y)).

For example, candidates in the motion vector candidate list may be reordered by comparing matching costs of motion information A, motion information B, motion information C, and motion information D.

For example, the matching cost of motion vector A is greater than the matching cost of motion vector B, the matching cost of motion vector B is greater than the matching cost of motion vector C, and the matching cost of motion vector C is the matching cost of motion vector D. If it is greater than the cost, it can be reordered in the order of (-mag_X, -mag_Y), (mag_X, -mag_Y), (-mag_X, mag_Y), (mag_X, mag_Y). In other words, the candidates can be reordered so that the candidate corresponding to the motion vector with the lower matching cost is positioned earlier in the list.

A motion vector difference candidate can be specified through an index from the motion vector difference candidate list to which this reordering is applied, and the sign of each component of the motion vector difference can be determined from the specified motion vector difference candidate.

For example, in the above-described embodiments, when the index for the motion vector difference for the target block is 1, the motion vector difference for the target block may be (mag_X, -mag_Y).

The index may be a pre-defined value. For example, the index may be 0.

The index may be determined through signaling/encoding/decoding.

44 illustrates a first code for signaling/encoding/decoding motion vector differences according to an example.

45 illustrates a second code for signaling/encoding/decoding motion vector differences according to an example.

46 illustrates a third code for signaling/encoding/decoding motion vector differences according to an example.

47 illustrates a fourth code for signaling/encoding/decoding motion vector differences according to an example.

48 illustrates a fifth code for signaling/encoding/decoding motion vector differences according to an example.

44 to 48, examples of codes for signaling/encoding/decoding motion vector differences have been described. Here, the motion vector difference information may include one or more of information described below as illustrated in FIGS. 44 to 48 .

MVD_NUM_ATSAMETIME may be a pre-defined positive integer. For example, MVD_NUM_ATSAMETIME can be 1, 2 or 3.

The information on the motion vector difference may include an indicator indicating whether an absolute value of the motion vector difference is greater than a specific value. (For example, abs_mvd_greater0_flag and abs_mvd_greater1_flag) A specific value may be an integer greater than or equal to 0. For example, abs_mvd_greater0_flag may be an indicator indicating whether an absolute value of a motion vector difference is greater than 0.

Information on the motion vector difference may include 1) an absolute value of the motion vector difference, 2) a sum of the absolute value of the motion vector difference and a specific value, or 3) a difference between the absolute value of the motion vector difference and a specific value. (eg, abs_mvd_minus2) A specific value can be an integer. For example, abs_mvd_minus2 may indicate an absolute value of a motion vector difference and a difference between 2 motion vectors.

mvd[refList] may indicate a motion vector difference in the direction of refList or an absolute value of the motion vector difference.

Information on the motion vector difference may include indices for determining an index of a motion vector difference sign combination. (eg sign_idx_hor and sign_idx_ver)

For example, when information on motion vector differences is signaled/encoded/decoded, context and/or probability information used in signaling/encoding/decoding may change based on motion information (or coding parameters) of a target block. there is. Alternatively, context and/or probability information used in signaling/encoding/decoding may be constant regardless of motion information (or coding parameters) of a target block.

For example, the motion information and/or coding parameter may be one or more of a magnitude of a motion vector difference and an affine mode indicator.

Signaling/encoding/decoding process of information on motion vector difference

A signaling/encoding/decoding process of information on motion vector difference will be described below. The signaling/encoding/decoding process may be performed in the order of 1) to 4) described below. However, the signaling/encoding/decoding process is not limited to the order of 1) to 4) below.

A list of possible MVD candidates may be generated based on 1) the code combination and 2) the absolute value of the MVD.

2) MV can be derived by adding MVD to MV predictor, and template matching cost can be calculated.

3) Reordering of candidates in the MVD list may be performed based on the derived template matching cost.

4) The true (ie, final) MVD can be retrieved based on the signaled MVD code prediction index.

Below, descriptions of a motion information search method according to an example are disclosed.

- Initial motion information may be bi-directional motion information. In this case, initial motion information may be constructed using an AMVP predictor for one direction. For another direction, initial motion information may be constructed using a merge predictor.

- A valid pair of merge reference blocks (or AMVP reference blocks) may have one reference picture from the past relative to the target picture and one reference picture from the future relative to the target picture. .

- AMVR parameter and BCW parameter may be set to default values IMV_OFF and BCW_DEFAULT.

- The AMVP part may be signaled as a regular one-way AMVP. The AMVP part may mean motion information of a direction in which an AMVP predictor is used when initial motion information is constructed.

- The merge part can be implicitly derived by minimizing the bilateral matching cost between the AMVP predictor and the merge predictor. The merge part may mean motion information of a direction in which a merge predictor is used when initial motion information is constructed.

- Affine and SMVD may not be used together in a mode of embodiments.

- A valid merge candidate may have a predictor in the (1-LX) direction. Alternatively, only merge candidates with predictors in the (1-LX) direction may be valid. A valid merge candidate may mean at least one of a merge predictor candidate and an initial motion information candidate of a merge part.

- The AMVP part of the mode of embodiments may be signaled as regular uni-directional AMVP. That is, the reference index and the MVD may be signaled, and the MVP index (or motion information index) may be signaled when template matching is disabled. If template matching is activated (ie, TM_AMVP is activated), the AMVP part may have a derived MVP index (or motion information index). In this case, the MVP index (or motion information index) Signaling may not be performed

- The merge index of the merge part (or the motion information index of the merge part) may not be signaled. Among the candidates in the motion information candidate list of the merge part, a candidate having the smallest template matching cost (or smallest two-sided matching cost) may be selected as a merge predictor (or motion information of the merge part).

- If the selected merge predictor and AMVP predictor satisfy the DMVR condition, both matching motion vector refinement may be applied to the merge motion vector candidate and the AMVP MVP as a starting point. Otherwise, if the template matching functionality is enabled, motion information improvement using template matching may be applied to a merge predictor or an AMVP predictor having a higher matching cost. Here, the explanation that the AMVP predictor satisfies the DMVR condition is that 1) there is at least one reference picture from the past relative to the target picture and at least one reference picture from the future relative to the target picture, and 2) two This may mean that distances from reference pictures to the target picture are the same.

- At least one decoder-end motion information derivation method may be performed on the selected merge predictor (or motion information of the merge part) and the AMVP predictor (or motion information of the AMVP part). For example, the decoder-end motion information derivation method may be at least one of bilateral matching, a partial search step of bilateral matching, template matching, a partial search step of template matching, and an optical-flow based motion information correction method. can For example, the decoder-end motion information derivation method may be an optical-flow based motion information correction method in units of 8x8 subblocks. For example, the decoder-end motion information derivation method may be a Bi-Directional Optical Flow (BDOF) in units of 8x8 sub-blocks.

- When template matching is activated, first the AMVP MVP can be improved by template matching, and then by both sides matching. The final MV of the AMVP portion can be derived by applying the TM enhanced MVD, BM enhanced MVD and signal MVD to the AMVP MVP.

Below, steps performed by a motion information search method according to an example will be described. The motion information search method may be performed in the order of 1) to 4) described below. However, the execution of the motion information search method is not limited to the order of 1) to 4) below.

1) AMVP part having signaled information may be determined.

2) The merge part may be determined using a matching cost between the AMVP predictor and the merge predictor. (The search may not be performed. Only the BM cost at the predictor position may be used.)

3) If the DMVR condition is satisfied, next improvement using DMVR can be performed. If the DMVR condition is not met (if the TM is active) then improvement using the TM can be performed.

4) MVD can be added to the AMVP part.

motion vector search mode

In the motion vector search of the embodiments, a motion vector search mode described later may be used.

When the motion vector search mode is selected, two reference picture indices may be signaled through a bitstream, and an MVP for each reference picture of the two reference pictures may be signaled through a bitstream.

When the motion vector search mode is selected, an MVP for each reference picture of two reference pictures may be signaled through a bitstream.

In this case, the reference picture index may be derived without signaling using a specific rule.

For example, the reference picture index may be a specified value IDX. IDX can be 0 or 1. However, IDX is not limited to 0 and/or 1.

In the L0-direction reference image list and the L1-direction reference image list, a pair of reference images may be specified. Here, the specified reference images may be a first reference image in the L0 reference image list and a second reference image in the L1 reference image list. The first POC interval may be a difference between the POC of the target image and the POC of the first reference image. The second POC interval may be a difference between the POC of the target image and the POC of the second reference image. The target image may be an image including the target block. The first POC interval and the second POC interval may be the same. A direction of the first reference image with respect to the target image and a direction of the second reference image with respect to the target image may be opposite to each other.

The reference picture index may indicate a first reference picture and a second reference picture. A reference picture index may be selected so that the first POC interval of the first reference image and the second POC interval of the second reference image are the same.

For example, the reference picture index may be selected to indicate reference pictures having the same POC interval for the target picture in list 0 and list 1. If there are two or more pairs of reference pictures having the same POC interval, a reference picture index having the smallest POC interval may be selected.

When the motion vector search mode is used, signaling of MVD can be omitted. In this case, the MVD may be derived through motion vector search in the decoding apparatus 1700.

When the motion vector search mode is used, the MVD can be transmitted only for either List 0 or List 1.

For example, when MVD is transmitted for list_X, motion vector search is performed using the motion vector for list_X (MVP_X + MVD) and the motion vector MVP_(1-X) for list_(1-X) as starting points. can start X can be 0 or 1.

For example, when MVD is transmitted for list_X, the motion vector for list_X (MVP_X + MVD) and the motion vector for list_(1-X) (MVP_(1-X) - MVD) are used as starting points. A motion vector search may be initiated.

For example, X can always be 0. X can always be 1.

For example, X may be specified through signaling. For example, X may be specified for a CU through signaling to the CU. However, the unit in which X is signaled is not limited to the CU.

When the motion vector search mode is used, a weight may be assigned to a motion vector from which search starts.

For example, a matching cost of a motion vector position where a search starts may be multiplied by a weight W. In this case, W may be a value of 1 or less. W can be a value greater than or equal to 1.

For example, when the motion vector search mode is one of the AMVP modes, W may be a value greater than or equal to 1. However, the value of W is not limited to 1 or more.

For example, when the motion vector search mode is one of the merge modes, W may be 1 or less. However, the value of W is not limited to 1 or more.

When the motion vector of the target block is referred to in neighboring blocks in one picture, either the MVP of the current block or the motion vector improved through motion vector search may be referred to. However, the referenced information is not limited to the information listed above.

The enhanced motion vector may be an enhanced motion vector for a block. The enhanced motion vector may be an enhanced motion vector for a sub-block. In other words, a unit to which enhancement is applied may be a block or a sub-block.

A motion vector retrieval method may be one or more of template matching and bilateral matching. However, the motion vector retrieval method is not limited to template matching and/or both sides matching.

When the motion vector retrieval method is performed, a motion vector may be fixed for one direction among bidirectional directions, and motion search may be performed only for the other direction.

For example, when the motion vector search method is bilateral matching, the motion vector for list X may be fixed, and search may be performed only for motion vectors of list (1-X). Here, X may be 0 or 1.

For example, X can always be 0. X can always be 1.

For example, X may be specified through signaling.

For example, when a motion vector search is performed in several stages, the motion vector for list X is fixed until a specific search stage, and the motion vector for list (1-X) may be searched only. From the subsequent search step, motion vector search may be performed on list 0 and list 1.

For example, X can always be 0. X can always be 1.

For example, X may be specified through signaling.

For example, several steps of motion vector search may be a motion vector search step in units of blocks or a motion vector search step in units of sub blocks. However, the unit for motion vector search is not limited to blocks and/or sub-blocks.

Motion vector search may be performed in units of blocks and units of subblocks.

The motion vector search area may be SR. The search area of the above-described embodiments may be applied to the motion vector search area.

For example, SR can be one of 2, 4 and 8. However, SR is not limited to the values listed above.

For example, SR may be equal to the search range of the DMVR mode.

For example, SR may be flexibly changed according to the size of MVP.

After the motion vector search mode is performed, motion vector improvement methods performed by the encoding apparatus 1600 and the decoding apparatus 1700 may be performed. For example, at least one of BDOF, LIC, and OBMC may be performed as a motion vector improvement method. However, the motion vector enhancement method is not limited to the methods listed above.

Signaling of information indicating whether motion vector search mode is performed

Whether the motion vector search mode is performed may be determined through signaling through a specific conditional statement or bitstream.

The specific conditional statement may be a conditional statement regarding one or more of block size, MVP size, whether the SMVD mode is applied, and whether the DMVR mode can be applied. However, the condition of the conditional statement is not limited to the items listed above.

When the target block does not satisfy the condition allowing the motion vector search mode to be applied, signaling related to the motion vector search mode may be omitted and the motion vector search mode may not be performed.

When a motion vector search mode is applied to a target block, signaling of information about one or more other modes may be omitted.

For example, when the motion vector search mode can be performed only when the BCW index is BCW_DEFAULT and the AMVR mode index is 0, signaling of the BCW index and the AMVR mode index can be omitted.

For example, when the motion vector search mode is performed, at least one of LICFlag and OBMCFlag may always be true. LICFlag may be a flag indicating whether the LIC mode is used. OBMCFlag may be a flag indicating whether OBMC mode is used.

For example, when the motion vector search mode is performed, at least one of LICFlag and OBMCFlag may always be false.

As a detailed mode of the MVD mode, whether the motion vector search mode is performed can be determined by signaling through a specific conditional statement or bitstream as a detailed mode of the SMVD mode.

When there are a plurality of motion vector search methods, one motion vector search method among the plurality of motion vector search methods may be specified by signaling through a specific conditional statement or bitstream. For example, when POC intervals between two reference pictures and a target picture are the same, a motion vector may be searched using bilateral matching. If the PCT intervals are different, the motion vector may be found using template matching. However, the specific conditions for the search method are not limited to the conditions described above.

The motion vector search region SR may be specified by signaling through a specific conditional statement or bitstream.

For example, a specific conditional statement may include conditions for the size of a target block or SRs in neighboring blocks. However, the specific conditional statement is not limited to the conditions described above.

For example, SR may be signaled through a bitstream at a video level, a sequence level, a picture level, a slice level, and a CU level. However, the level at which SR is signaled is not limited to the aforementioned levels.

49 illustrates a signaling method of information for a motion vector search mode according to an example.

In motion vector search mode and SMVD mode, identical_mvp_flag and mvp_L0_flag may be signaled for signaling of mvp indexes of list 0 and list 1. The identical_mvp_flag may indicate whether a value of mvp_L0 and a value of mvp_L1 are the same.

mvp_L0 may be mvp_L0_flag. mvp_L1 may be mvp_L1_flag.

mvp_both_zero_flag, mvp_both_one_flag, and mvp_L0_flag can be signaled for signaling about the mvp index of list 0 and mvp index of list 1 in motion vector search mode and SMVD mode.

mvp_both_zero_flag may indicate whether both mvp_L0 and mvp_L0 are 0. If mvp_both_zero_flag is 1, both mvp_L0 and mvp_L1 may be set to 0. If mvp_both_zero_flag is 1, since values of mvp_L0 and mvp_L1 are determined by mvp_both_zero_flag, mvp_both_one_flag and mvp_L0_flag may not be signaled.

If mvp_both_zero_flag is 0, mvp_both_one_flag may be signaled next. Here, if mvp_both_one_flag is 1, both mvp_L0 and mvp_L1 may be set to 1. If mvp_both_one_flag is 1, since values of mvp_L0 and mvp_L1 are determined by mvp_both_one_flag, mvp_L0_flag may not be signaled.

When mvp_both_zero_flag is 0 and mvp_both_one_flag is 0, one of mvp_L0 and mvp_L1 may be 0 and the other may be 1. At this time, mvp_L0_flag may be signaled. When mvp_L0_flag is signaled, mvp_L1_flag may be set to 1 - mvp_L1_flag.

In the motion vector search mode and the SMVD mode, when the mvp index of list 0 and the mvp index of list 1 are signaled, MVP candidates can be rearranged by comparing matching costs for mvp combinations.

For example, when the candidates of list 0 are mvp0_0 and mvp0_1 and the candidates of list 1 are mvp1_0 and mvp1_1, a combination consisting of one of the candidates of list 0 and one of the candidates of list 1 can be configured. Here, the combination may be an mvp combination. Four combinations (i.e., (mvp0_0, mvp1_0), (mvp0_1, mvp1_0), (mvp0_0, mvp1_1), and (mvp0_1, mvp1_1)) will be constructed using the two candidates from list 0 and the two candidates from list 1. can

The costs of motion vector search modes can be compared for (mvp0_0, mvp1_0), (mvp0_1, mvp1_0), (mvp0_0, mvp1_1) and (mvp0_1, mvp1_1), in ascending order of costs (mvp0_0, mvp1_0), (mvp0_1, Indexes of 0 to 3 may be assigned to mvp1_0), (mvp0_0, mvp1_1), and (mvp0_1, mvp1_1), respectively. In other words, an index of 0 may be assigned to a combination having the lowest cost among (mvp0_0, mvp1_0), (mvp0_1, mvp1_0), (mvp0_0, mvp1_1), and (mvp0_1, mvp1_1). However, the number of mvp candidates in the list is not limited to two.

For example, after reordering is performed, signaling of the mvp index may be omitted, and a combination having the lowest cost among combinations may be selected. The cost may be the matching cost of the motion vector search mode. Inter prediction may be performed on the target block using the selected combination.

For example, matching cost can be calculated using either template matching or two-sided prediction. However, the method of calculating the matching cost is not limited to template matching and/or bilateral prediction.

A motion vector search method of the motion vector search mode and a motion vector search method used to obtain a matching cost may be different from each other.

Block size to which motion vector search mode is applied

The motion vector search method may be applied to the target block when the horizontal length or vertical length of the target block is greater than or equal to Min_bm_smvd_size. For example, Min_bm_smvd_size may be 0 or 8. However, Min_bm_smvd_size is not limited to 0 or 0.

The motion vector search method may be applied to the target block when the horizontal length or vertical length of the target block is equal to or less than Max_bm_smvd_size. For example, Max_bm_smvd_size may be 128. However, Max_bm_smvd_size is not limited to 128.

The motion vector search method may be applied to the target block when the total number of samples of the target block is greater than or equal to Min_bm_smvd_sample. For example, Min_bm_smvd_sample may be 0 or 128. However, Min_bm_smvd_sample is not limited to 0 or 128.

The motion vector search method may be applied to the target block when the total number of samples of the target block is equal to or less than Max_bm_smvd_sample. For example, Max_bm_smvd_sample may be 256. However, Max_bm_smvd_sample is not limited to 256.

Signaling/encoding/decoding of coding information related to inter prediction

50 may be a first code representing coding information signaled in relation to a block division structure according to an example.

51 may be a second code representing coding information signaled in relation to a block division structure according to an example.

The second code of FIG. 51 may follow the first code of FIG. 50 .

Below, the signaling/encoding/decoding of coding information in step 1820 and step 1930 is described in more detail.

인터 예측을 수행하기 위해 사용되는 코딩 정보는 pred_mode_flag, sps_smvd_enabled_flag, inter_pred_idc, inter_affine_flag, ph_mvd_l1_zero_flag, NumRefIdxactive, sym_mvd_flag, MotionModelIdc, mvp_l0_flag, mvp_l1_flag, MvdL0, MvdL1, cu_skip_flag, merge_flag, merge_idx, cu_cbf, tu_cbf_luma, tu_cbf_cb, tu_cbf_cr, 움직임 It may include one or more of information and motion vector difference for performing an information search method.

The coding information for performing the motion information retrieval method is a motion vector difference, a motion information offset, an indicator specifying a method for deriving the decoder-end motion information to be used in the target block, and motion information improvement when a motion information candidate list is constructed. It may include one or more of an indicator indicating a direction in which final motion information is performed, an indicator indicating a direction in which motion information improvement is performed when final motion information is determined, BM_NUM, a motion information index for determining final motion information, and a reference image index.

BM_NUM may be an indicator indicating how many directions to perform motion information improvement when a specific search step of a specific decoder-end motion information derivation method of a motion information search method is performed. BM_NUM can be plural. A plurality of BM_NUMs may be respectively applied to different decoder-end motion information derivation methods. Alternatively, a plurality of BM_NUMs may be applied to different search steps of decoder-level motion information derivation methods, respectively.

For example, when encoding of coding information related to inter prediction is performed or decoding of coding information is performed, a plurality of pieces of coding information for performing a motion information search method may be signaled/encoded/decoded. Coding information for performing the motion information retrieval method may be information on different decoder-end motion information derivation methods of the motion information retrieval method or information on different search steps in the decoder-end motion information derivation method. there is.

pred_mode_flag may be information indicating whether inter prediction mode is applied. pred_mode_flag may be signaled/encoded/decoded for one or more units of a coding block, a prediction block, and a coding unit. For example, when the information indicating whether the inter prediction mode is applied is a first value (eg, 0), it may be indicated that the inter prediction mode is applied. When the information indicating whether the inter prediction mode is applied is the second value (eg, 1), it may indicate that the inter prediction mode is not applied.

sps_smvd_enabled_flag may be a syntax element determined at the sequence parameter set level. sps_smvd_enabled_flag may be an indicator indicating whether a symmetric motion vector differential mode is activated in a target sequence.

When sps_smvd_enabled_flag is the first value, the symmetric motion vector differential mode may be deactivated in the target sequence. The first value can be 0 or false.

When sps_smvd_enabled_flag is the second value, the symmetric motion vector differential mode may be activated in the target sequence. The second value may be 1 or true.

inter_pred_idc may be a syntax element for an inter prediction direction. inter_pred_idc may be an inter prediction indicator.

When inter_pred_idc is the first value, unidirectional inter prediction in the L0 direction may be performed on the target block. The first value may be 1 or PRED_L0.

When inter_pred_idc is the second value, unidirectional inter prediction in the L1 direction may be performed on the target block. The second value may be 2 or PRED_L1.

When inter_pred_idc is the third value, bi-directional inter prediction may be performed on the target block. The third value may be 3 or PRED_BI.

inter_affine_flag may be an indicator indicating whether affine mode is performed for a target block.

When inter_affine_flag is the first value, the affine mode may not be performed on the target block. The first value can be 0 or false.

When inter_affine_flag is the second value, affine mode may be performed on the target block. The second value may be 1 or true.

RefIdxSymL may be a reference image index in the L0 direction determined for the symmetric motion vector difference mode. RefIdxSymL1 may be a reference image index in the L1 direction determined for the symmetric motion vector difference mode. For example, RefIdxSymL0 may be an index indicating a reference image having the smallest POC interval among reference images in the L0 direction reference image list. The POC interval of the reference image may be a difference between the POC of the target image and the POC of the reference image. For example, RefIdxSymL1 may be an index indicating a reference image having the smallest POC interval among reference images in the L1 direction reference image list.

ph_mvd_l1_zero_flag may be a syntax element determined at the picture parameter set level. ph_mvd_l1_zero_flag may be an indicator indicating whether to perform signaling/encoding/decoding for the L1-direction motion vector difference with respect to the target image.

When ph_mvd_l1_zero_flag is the first value, signaling/encoding/decoding of the L1-direction motion vector difference in the target image may be performed. The first value can be 0 or false.

When ph_mvd_l1_zero_flag is the second value, signaling/encoding/decoding of the L1-direction motion vector difference in the target image may not be performed. The second value may be 1 or true.

sym_mvd_flag may be an indicator indicating whether to perform the symmetric motion vector differential mode.

For example, when sym_mvd_flag is the first value, the symmetric motion vector differential mode may not be performed on the target block. The first value can be 0 or false.

For example, when sym_mvd_flag is the second value, the symmetric motion vector differential mode may be performed on the target block. The second value may be 1 or true.

MotionModelIdc may be a syntax element indicating 1) whether the affine mode is used for the target block and 2) the number of control point motion vectors used in the affine mode when the affine mode is used for the target mode.

MotionModelIdc having the first value may mean that the affine mode is not performed for the target block. The first value may be 0.

When MotionModelIdc is the second value, it may mean that the affine mode is performed on the target block and two control point motion vectors are used. The second value may be 1.

MotionModelIdc having a third value may mean that an affine mode is performed on the target block and three control point motion vectors are used. The third value may be 2.

mvp_l0_flag may be a motion information index in the L0 direction.

mvp_l1_flag may be a motion information index in the L1 direction.

mvd_coding(x0, y0, X, Y) may indicate that signaling/encoding/decoding of motion vector difference for the Y-th motion information in the LX direction is performed. When the affine mode is not performed for the target block, Y may always be 0. When the affine mode is performed on the target block, Y may be one of 0, 1 and 2. The value range of Y may be determined based on the MotionModelIdc value.

MvdL0[x0][y0][0] may mean a horizontal component of the L0-direction motion vector difference. MvdL0[x0][y0][1] may mean a vertical component of the L0-direction motion vector difference.

MvdL1[x0][y0][0] may mean a horizontal component of the L1-direction motion vector difference. MvdL1[x0][y0][1] may mean a vertical component of the L1-direction motion vector difference.

cu_skip_flag may be skip mode indication information indicating whether skip mode is used. cu_skip_flag may be signaled/encoded/decoded in at least one unit of a coding block and a prediction block. For example, when the skip mode indication information has a first value (eg, 1), it may indicate that the skip mode is used. When the skip mode indication information is the second value (eg, 0), it may not indicate that the skip mode is used. At this time, cu_skip_flag may indicate the use of inter skip mode.

merge_flag may be merge mode indication information indicating whether merge mode is used. merge_flag may be signaled/encoded/decoded in at least one unit of a coding block and a prediction block. For example, when the merge mode indication information has a first value (eg, 1), it may indicate that merge mode is used. When the merge mode indication information has a second value (eg, 0), it may not indicate that merge mode is used. At this time, merge_flag may indicate the use of inter merge mode.

merge_idx may be information indicating a merge candidate in a merge candidate list. merge_idx may be signaled/encoded/decoded in units of at least one of a coding block and a prediction block. Also, merge_idx may mean merge index information. In addition, merge_idx may indicate a block from which a merge candidate is derived among blocks reconstructed spatially adjacent to a target block. Also, merge_idx may indicate at least one piece of motion information included in a merge candidate. For example, if the merge index information is a first value (eg, 0), the first merge candidate in the merge candidate list may be indicated. If the merge index information is a second value (eg, 1), a second merge candidate in the merge candidate list may be indicated. If the merge index information is a third value (eg, 2), a second merge candidate in the merge candidate list may be indicated. Similarly, if the merge index information is one of the third to Nth values, a merge candidate corresponding to the value of the merge index information may be indicated according to the order of merge candidates in the merge candidate list. Here, N may be a positive integer including 0. At this time, merge_idx may indicate a merge index when inter merge mode is used. That is, the merge candidate list may mean a motion information candidate list. A merge candidate may mean a block information candidate.

A block vector difference may be a difference between a motion vector of an AMVP mode and a motion vector of a motion information candidate. A motion vector differential may be signaled/encoded/decoded for the target block. A prediction block for the target block may be derived using the motion vector difference.

cu_cbf, tu_cbf_luma, tu_cbf_cb, and tu_cbf_cr may indicate whether quantized transform coefficients exist in the residual block.

cu_cbf may be information indicating whether a quantized transform coefficient exists.

When the block division structure of the luma component and the block division structure of the chroma component are the same, cu_cbf may be information indicating whether a quantized transform coefficient of the luma component block and a quantized transform coefficient of the chroma component block exist. When the luma component and the chroma component each have a block division structure independent of each other, cu_cbf may be information indicating whether a quantized transform coefficient of the luma component block or the chroma component block exists.

When cu_cbf is a first value (eg, 1), it may mean that a quantized transform coefficient of a block exists. When cu_cbf is the second value (eg, 0), it may mean that there is no quantized transform coefficient of the block.

tu_cbf_luma may indicate whether a quantized transform coefficient of a luma component block exists.

tu_cbf_cr may indicate whether a quantized transform coefficient of the chroma component Cr exists.

tu_cbf_cb may indicate whether a quantized transform coefficient of the chroma component Cb exists.

When tu_cbf_luma is a first value (eg, 1), it may mean that a quantized transform coefficient of a luma component block exists.

When tu_cbf_luma is a second value (eg, 0), it may mean that a quantized transform coefficient of a luma component block exists.

When tu_cbf_cr is a first value (eg, 1), it may mean that a quantized transform coefficient of the chroma cr component block exists.

When tu_cbf_cr is a second value (eg, 0), it may mean that a quantized transform coefficient of the chroma cr component block exists.

When tu_cbf_cb is a first value (eg, 1), it may mean that a quantized transform coefficient of a chroma cb component block exists.

When tu_cbf_cb is a second value (eg, 0), it may mean that a quantized transform coefficient of the chroma cb component block exists.

At least one of the coding information for performing the motion information search method includes a parameter set, a header, a brick, a Coding Tree Unit (CTU), a Coding Unit (CU), and a Prediction Unit Signaling/encoding/decoding may be performed in at least one of a PU), a transform unit (TU), a coding block (CB), a prediction block, and a transform block.

At this time, at least one of the parameter set, header, brick, CTU, CU, PU, TU, CB, PB, or TB is a video parameter set, a decoding parameter set, and a sequence parameter set. parameter set), adaptation parameter set, picture parameter set, picture header, sub-picture header, slice header, tile group header ( tile group header), tile header, brick, CTU, CU, PU, TU, CB, PB, or TB.

Here, prediction using the motion information search method is performed using coding information for performing the motion information search method in at least one of the signaled parameter set, header, brick, CTU, CU, PU, TU, CB, PB, and TB. can be performed

For example, when at least one of coding information for performing a motion information search method is signaled/encoded/decoded in a sequence parameter set, the encoding device 1600 and the decoding device 1700 have the same syntax element value in a sequence unit. Prediction using the motion information search method may be performed using at least one of coding information related to the motion information search method having .

As another example, when at least one of the coding information for performing the motion information search method is signaled/encoded/decoded in a slide header, the encoding device 1600 and the decoding device 1700 generate the same syntax element value in slice units. Prediction using the motion information search method may be performed using at least one of coding information related to the motion information search method.

As another example, when at least one of coding information for performing the motion information search method is signaled/encoded/decoded in an adaptation parameter set, the encoding device 1600 and the decoding device 1700 use the same syntax element in the adaptation parameter set. Prediction using the motion information search method may be performed using at least one of coding information related to the motion information search method having a value.

As another example, when at least one of the coding information for performing the motion information search method is signaled/encoded/decoded in a specific block, the encoding apparatus 1600 and the decoding apparatus 1700 generate the same syntax element value in the specific block. Prediction using the motion information search method may be performed using at least one of coding information related to the motion information search method. A specific block may be a CU, CB, PU, PB, TU or TB.

Here, at least one of the coding information for performing the motion information search method may be derived using at least one of coding parameters of a target tile, a target slice, a target sequence, a target image, a target block, CTB, and CTU. .

If at least one of the coding information for performing the motion information retrieval method does not exist in the bitstream, at least one of the coding information for performing the motion information retrieval method is inferred as a first value (eg, 0) ( can be inferred.

An adaptation parameter set may mean a parameter set that can be referenced and shared in different pictures, different subpictures, different slices, different tile groups, different tiles, or different bricks.

In addition, different adaptation parameter sets may be referred to in subpictures, slices, tile groups, tiles, or bricks in a picture, respectively, and information in the referenced adaptation parameter set may be used.

In addition, different adaptation parameter sets may be referred to by using identifiers of different adaptation parameter sets in subpictures, slices, tile groups, tiles, or bricks within a picture.

In addition, different adaptation parameter sets may be referred to by using identifiers of different adaptation parameter sets in slices, tile groups, tiles, or bricks in a subpicture, respectively.

Also, different adaptation parameter sets may be referred to by using identifiers of different adaptation parameter sets in tiles or bricks in a slice, respectively.

Also, in bricks within a tile, different adaptation parameter sets may be respectively referred to by using identifiers of different adaptation parameter sets.

A parameter set or header of a subpicture may include information about an adaptation parameter set identifier. An adaptation parameter set corresponding to the adaptation parameter set identifier using the adaptation parameter set identifier may be used for the subpicture.

A parameter set or header of a tile may include information about an adaptation parameter set identifier. An adaptation parameter set corresponding to the adaptation parameter set identifier using the adaptation parameter set identifier may be used for the tile.

A parameter set or header of a brick may include information about an adaptation parameter set identifier. An adaptation parameter set corresponding to the adaptation parameter set identifier using the adaptation parameter set identifier may be used for the brick.

A picture can be divided into one or more tile rows and one or more tile columns.

A subpicture can be divided into one or more tile rows and one or more tile columns within the picture. A subpicture may be a rectangular or square area within a picture. A subpicture may contain one or more CTUs. Also, one or more tiles, one or more bricks, and/or one or more slices may be included in one subpicture.

A tile may be an area having a rectangular/square shape within a picture. A tile may contain one or more CTUs. Also, a tile can be divided into one or more bricks.

A brick may mean one or more CTU rows in a tile. A tile can be divided into one or more bricks. Each brick can have one or more CTU rows. A tile that is not divided into two or more bricks may also mean a brick.

A slice may include one or more tiles within a picture. Also, a slice may include one or more bricks within a tile.

In the embodiments, the description that "specific information is pre-defined" refers to the specific information in the encoding device 1600 and the decoding device 1700 through the same rules and/or methods in the encoding device 1600 and the decoding device 1700. It may mean that information is determined with the same value/method. This specific information may be one or more of coding parameters, block information, search patterns, search sequences, and whether or not a specific mode is used. However, the specific information is not limited to the information listed above.

Entropy encoding/decoding may be performed at one or more levels of sequence level, picture level, tile level, tile group level, slice level, CTU level, CU level, and PU level. However, units in which entropy encoding/decoding is performed are not limited to the levels listed above.

In embodiments, the motion information (MI_L0, MI_L1) may mean motion information having motion information of MI_L0 in the L0 direction and motion information of MI_L1 in the L1 direction.

In the embodiments, the description of “improvement of motion information” means that improved motion information is derived as an intermediate search step and/or a final search step of a decoder-level motion information derivation method are performed on the motion information. can

In embodiments, a neighboring block means a block temporally or spatially close to the target block.

Expressions of “temporally adjacent” and “temporally neighboring” may mean that a specific entity belongs to a col image or an adjacent image of a target image. The adjacent image may be an image whose interval from the target image is smaller than POC_THRES. The interval between the adjacent image and the target image may be a difference between the POC of the adjacent image and the POC of the target image.

POC_THRES can be a positive number greater than or equal to 1. POC_THRES may be pre-defined. Information representing POC_THRES can be signaled/encoded/decoded.

Expressions of “spatially adjacent” and “spatially neighboring” may mean that a distance between a specific object and a pre-defined location of a target block is smaller than SPATIAL_THRES.

For example, the pre-defined position of the target block may be one of upper left, center, upper right, lower left, and lower right of the target block. However, the pre-defined location of the target block is not limited to the locations listed above.

For example, SPATIAL_THRES can be a positive number greater than or equal to 1. SPATIAL_THRES may be pre-defined. Information indicating SPATIAL_THRES can be signaled/encoded/decoded.

In embodiments, the "matching cost" may be a resultant value determined by a cost function. The cost function may be a cost function between templates in the decoder-end motion information derivation method. For example, the matching cost may be at least one of a bilateral matching cost and a template matching cost. However, the matching cost is not limited to both the matching cost and/or the template matching cost.

In embodiments, "matching cost for specific motion information" is 1) matching cost between templates determined from specific motion information and 2) from motion information derived using a decoder-end motion information derivation method from specific motion information. It may be one of matching costs between determined templates.

In embodiments, a “motion vector” may be one of 1) a motion vector of inter prediction and 2) a block vector (BV) of intra block copy (IBC).

In embodiments, the terms "bi-prediction", "bi-directional prediction", "inter bi-prediction", "bi-directional inter prediction" )" may be the same, and the terms may be used interchangeably.

In the encoding apparatus 1600, 1) determination of neighboring blocks to be included in the motion information candidate list, 2) construction, reconstruction, and determination of the motion information candidate list, 3) determination of motion information (determination of motion information index, decoder-end motion information) Derivation of index using derivation method, motion information derivation using decoder-side motion information derivation method), 4) operation with motion information offset, 5) operation adding motion vector difference to motion information, 6) decoder-side motion information Improvement of motion information using a derivation method, 7) Determination of a neighboring block using at least one of the above-described embodiments for the signaling/encoding/decoding process of motion information, 8) Construction, reconstruction and determination of a motion information candidate list , 9) motion information determination, 10) operation using motion information offset, 11) operation adding motion vector difference to motion information, and 12) improvement of motion information using decoder-end motion information derivation method can be performed. .

Also, in the decoding apparatus 1700, 1) determination of a neighboring block using at least one of the above-described embodiments, 2) construction, reconstruction, and determination of a motion information candidate list, 3) determination of motion information, and 4) motion information offset It is possible to perform an operation using , 5) an operation of adding a motion vector differential to motion information, 6) improvement of motion information using a decoder-end motion information derivation method, and 7) signaling/decoding of motion information.

Embodiments have been described for the case of using two reference picture lists. However, inter prediction to which embodiments are applied is not limited to inter prediction using two reference picture lists. Embodiments can also be used when using NUM_REFPICLIST reference picture lists. For example, NUM_REFPICLIST can be 1, 2, 3 or a positive integer.

Embodiments have been described for the case of using one or two motion information candidate lists. However, inter prediction to which embodiments are applied is not limited to inter prediction using one or two motion information candidate lists. Embodiments may be applied even when NUM_MILIST motion information candidate lists are used. NUM_MILIST can be, for example, 1, 2, 3 or a positive integer.

Embodiments may be applied according to the size of at least one of a coding block, a prediction block, a block, and a unit. The size herein may mean a minimum size and/or a maximum size for which embodiments are applied, and may mean a fixed size to which embodiments are applied. Also, the first embodiment may be applied to the first size, and the second embodiment may be applied to the second size. That is, the embodiments may be applied in a complex manner according to the sizes of objects. In addition, the embodiments may be applied when the size of the object is greater than or equal to the minimum size and less than or equal to the maximum size. That is, the embodiments may be applied only when the size of an object is included within a specific range.

In addition, the embodiments may be applied only when the size of the object is greater than or equal to the minimum size and less than or equal to the maximum size. Here, each of the minimum size and maximum size may be a size of one of a block and a unit. That is, a block to which a minimum size restriction is applied may be different from a block to which a maximum size restriction is applied. For example, the embodiments may be applied only when the size of the target block is greater than or equal to the minimum size of the block and less than or equal to the maximum size of the block.

For example, the embodiments may be applied only when the size of the target block is greater than or equal to 8x8. For example, embodiments may be applied only when the size of a target block is greater than or equal to 16x16. For example, the embodiments may be applied only when the size of the target block is greater than or equal to 32x32. For example, embodiments may be applied only when the size of a target block is greater than or equal to 64x64. For example, embodiments may be applied only when the size of a target block is greater than or equal to 128x128. For example, embodiments may be applied only when the size of a target block is 4x4. For example, the embodiments may be applied only when the size of the target block is 8x8 or smaller. For example, the embodiments may be applied only when the size of the target block is 16x16 or smaller. For example, embodiments may be applied only when the size of a target block is greater than or equal to 8x8 and less than or equal to 16x16. For example, embodiments may be applied only when the size of a target block is greater than or equal to 16x16 and less than or equal to 64x64.

The embodiments may be selectively applied according to a temporal layer. A separate identifier may be signaled to identify a temporal layer to which embodiments may be applied. Embodiments may be applied to a temporal layer specified by an identifier. The identifier herein may be defined to indicate a minimum layer and/or a maximum layer to which an embodiment may be applied, or may be defined to indicate a specific layer to which an embodiment is applied.

For example, the embodiments may be applied only when the temporal layer of the target image is the lowest layer. For example, the embodiments may be applied only when the temporal layer identifier of the target image is 0. For example, the embodiments may be applied only when the temporal layer identifier of the target image is 1 or more. For example, the embodiments may be applied only when the temporal layer of the target image is the highest layer.

In the above-described embodiments, motion information of a target block, an intra prediction mode for a block, an inter prediction mode, a reference picture list, a color component, a size, a shape, a motion information index, a type of a cost function, and derivation of decoder-end motion information Based on at least one of coding parameters such as a type of method, 1) determining a neighboring block to be included in a motion information candidate list, 2) constructing, reconstructing, and determining a motion information candidate list, 3) determining motion information, 4) Determination of whether motion information enhancement using the decoder-side motion information derivation method is performed, 5) Determination of the direction in which motion information derivation using the decoder-side motion information derivation method is to be performed, 6) Whether motion information offset is used 7) Determination of motion information on which an operation with a motion information offset is to be performed, 8) Determination of whether to signal/encode/decode motion vector differentials, 9) Determination of whether to signal/encode/decode motion information offsets, 10) Determination of whether to signal/encode/decode the motion information index, 11) Determination of the type of decoder-end motion information derivation method to be used for improving motion information in a target block, 12) Determination of the type of motion information signaling/encoding/decoding to be used At least one of a determination of a probabilistic model or a contextual model may be performed.

As described in the embodiment, the reference picture set used in the process of reference picture list construction and reference picture list modification refers to L0 reference picture list and L1 reference picture set. One or more reference image lists among an image list, an L2 reference image list, and an L3 reference image list may be used.

According to embodiments, motion information of a target block may be used to calculate boundary strength in a deblocking filter. At this time, one or more pieces of motion information may be used. A maximum of N pieces of motion information may be used. N may be a positive integer greater than or equal to 1. For example, N can be 2, 3 or 4.

The unit of the motion vector is 16-pel, 8-pel, 4-pel, integer-pel, and 1/2-pel. 1/2-pel increments, 1/4-pel increments, 1/8-pel increments, 1/16-pel increments , 1/32-pel units, and 1/64-pel units. At least one of the aforementioned pel units may be selectively used as a unit of a motion vector in a signaling/encoding/decoding process for a target block.

A slice type to which embodiments are applied may be defined. Embodiments may be applied based on the slice type.

A block to which embodiments are applied may have a square shape or a non-square shape.

Indicators, indexes, and flags encoded by the encoding device 1600 and decoded by the decoding device 1700, 1) determining neighboring blocks to be included in the motion information candidate list, 2) constructing, reconstructing, and determining the motion information candidate list, 3) Determination of motion information (determination of motion information index, derivation of index using decoder-end motion information derivation method and motion information derivation using decoder-end motion information derivation method), 4) operation with motion information offset, 5) operation of adding motion vector difference to motion information, 6) improvement of motion information using decoder-stage motion information derivation method, 7) motion vector difference, 8) reference picture index, 9) syntax elements related to motion information index, etc. For at least one of them, one or more of the following binarization, debinarization, and encoding/decoding methods may be used.

- Signed 0-th order Exp_Golomb binarization/inverse binarization method (abbreviated as se(v))

- Signed k-th order Exp_Golomb binarization/inverse binarization method (abbreviated as sek(v))

- 0-th order Exp_Golomb binarization/inverse binarization method for unsigned positive integers (abbreviated as ue(v))

- k-th order Exp_Golomb binarization/inverse binarization method for unsigned positive integers (abbreviated as uek(v))

- Fixed-length binarization/inverse binarization methods (abbreviated as f(n))

- truncated Rice binarization/inverse binarization method or truncated unary binarization/inverse binarization method (abbreviated as tu(v))

- truncated binary binarization/inverse binarization method (abbreviated as tb(v))

- context-adaptive arithmetic encoding/decoding method (abbreviated as ae(v))

- bit string in bytes (abbreviated as b(8))

- signed integer binarization/inverse binarization method (abbreviated as i(n))

- Unsigned positive integer binarization/inverse binarization method (abbreviated as u(n)) ('u(n)' may mean fixed-length binarization/inverse binarization method.)

- Unary binarization/inverse binarization method

Not only one limited embodiment among the embodiments is applied to the signaling/encoding/decoding process of the target block. A specific embodiment or at least one combination of embodiments may be applied to a signaling/encoding/decoding process for a target block.

The above embodiments may be performed in the same method and/or a corresponding method in the encoding device 1600 and the decoding device 1700. Also, a combination of one or more of the above embodiments may be used in encoding and/or decoding an image.

The order in which the above embodiments are applied may be different in the encoding device 1600 and the decoding device 1700. Alternatively, the order in which the above embodiments are applied may be the same (at least partially) in the encoding device 1600 and the decoding device 1700.

The above embodiments may be performed for each of the luma signal and chroma signal. The above embodiments may perform the same with respect to the luma signal and the chroma signal.

The shape of a block to which the above embodiments are applied may have a square shape or a non-square shape.

Whether to apply and/or perform at least one of the above embodiments may be determined based on a condition for a block size. In other words, at least one of the above embodiments may be applied and/or performed when a condition for a block size is satisfied. The condition may include a minimum block size and a maximum block size. The block may be one of the blocks described above in the embodiments and the units described above in the embodiments. A block to which the minimum block size is applied and a block to which the maximum block size is applied may be different.

For example, when the size of a block is equal to or greater than the minimum size and/or when the size of a block is equal to or less than the maximum size, the above-described embodiment may be applied and/or performed. When the size of the block is greater than the minimum size and/or when the size of the block is less than or equal to the maximum size, the above-described embodiment may be applied and/or performed.

For example, the above-described embodiment may be applied only when the size of a block is a predefined block size. The predefined block size may be 2x2, 4x4, 8x8, 16x16, 32x32, 64x64 or 128x128. The predefined block size may be (2*SIZE _X )x(2*SIZE _Y ). SIZE _X can be one of 1 or more integers. SIZE _Y can be one of 1 or more integers.

For example, the above-described embodiment may be applied only when the size of a block is greater than or equal to the minimum block size. The above-described embodiment can be applied only when the size of a block is larger than the minimum block size. Block minimum size can be 2x2, 4x4, 8x8, 16x16, 32x32, 64x64 or 128x128. Alternatively, the minimum block size may be (2*SIZE _{MIN_X} )x(2*SIZE _{MIN_Y} ). SIZE _{MIN_X} can be one of 1 or more integers. SIZE _{MIN_Y} can be one of 1 or more integers.

For example, the above-described embodiment may be applied only when the size of a block is less than or equal to the maximum block size. The above-described embodiment can be applied only when the size of a block is smaller than the maximum block size. The maximum block size can be 2x2, 4x4, 8x8, 16x16, 32x32, 64x64 or 128x128. Alternatively, the maximum block size may be (2*SIZE _{MAX_X} )x(2*SIZE _{MAX_Y} ). SIZE _{MAX_X} can be one of 1 or more integers. SIZE _{MAX_Y} can be one of 1 or more integers.

For example, the above-described embodiment may be applied only when the size of a block is greater than or equal to the minimum block size and less than or equal to the maximum block size. The foregoing embodiment can be applied only when the block size is greater than the minimum block size and less than or equal to the maximum block size. The above-described embodiment can be applied only when the size of a block is greater than or equal to the minimum block size and smaller than the maximum block size. The above-described embodiment can be applied only when the block size is larger than the minimum block size and smaller than the maximum block size.

In the above-described embodiments, the size of a block may mean a horizontal size of a block or a vertical size of a block. The size of a block can mean both the horizontal size of the block and the vertical size of the block. Also, the size of a block may mean the area of the block. Each of the area, minimum block size, and maximum block size may be one of integers greater than or equal to 1. Also, the size of a block may mean a result (or value) of a known formula using the horizontal size and vertical size of a block or a result (or value) of a formula in an embodiment.

Also, in the above embodiments, the first embodiment may be applied to the first size, and the second embodiment may be applied to the second size.

The above embodiments may be applied according to a temporal layer. A separate identifier may be signaled to identify a temporal layer to which the above embodiments are applicable, and the above embodiments may be applied to a temporal layer specified by the identifier. The identifier herein may be defined as the lowest layer and/or the highest layer to which the above embodiment is applicable, or may be defined to indicate a specific layer to which the above embodiment is applied. In addition, a fixed temporal layer to which the above embodiment is applied may be defined.

For example, the above embodiments may be applied only when the temporal layer of the target image is the lowest layer. For example, the above embodiments may be applied only when the temporal layer identifier of the target image is 1 or more. For example, the above embodiments may be applied only when the temporal layer of the target image is the highest layer.

A slice type or tile group type to which the above embodiments are applied may be defined, and the above embodiments may be applied according to the corresponding slice type or tile group type.

In the above-described embodiments, in applying a specified process to a specified subject, a specified condition may be required, and if it is described that the specified process is processed under a specified decision, the specified coding parameter If it is described that whether or not a specified condition is satisfied based on a specific coding parameter or a specified decision is made based on a specified coding parameter, it can be interpreted that the specified coding parameter can be replaced with another coding parameter. That is to say, a coding parameter that affects a specified condition or a specified decision may be regarded as exemplary only, and it will be understood that a combination of one or more other coding parameters in addition to the specified coding parameter plays a role of the above specified coding parameter. can

In the foregoing embodiments, the methods are described on the basis of a flow chart as a series of steps or units, but the present invention is not limited to the order of steps, and some steps may occur in a different order or concurrently with other steps as described above. can In addition, those skilled in the art will understand that the steps shown in the flow chart are not exclusive, that other steps may be included, or that one or more steps of the flow chart may be deleted without affecting the scope of the present invention. You will understand.

The embodiments described above include examples of various aspects. While not all possible combinations for representing the various aspects may be described, those skilled in the art will recognize that other combinations than those explicitly described are possible. Accordingly, it is intended that the present invention cover all other substitutions, modifications and variations falling within the scope of the following claims.

Embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded on a computer-readable recording medium. The computer readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the computer-readable recording medium may be specially designed and configured for the present invention, or may be known and usable to those skilled in the art of computer software.

A computer-readable recording medium may contain information used in embodiments according to the present invention. For example, a computer-readable recording medium may include a bitstream, and the bitstream may include information described in embodiments according to the present invention.

A bitstream may include computer-executable code and/or programs. Computer-executable codes and/or programs may include information described in the embodiments and may include syntax elements described in the embodiments. In other words, the information and syntax elements described in the actual examples may be regarded as computer-executable code in a bitstream, and may be regarded as at least part of a computer-executable code and/or program represented in a bitstream. The recording medium may include a non-transitory computer-readable medium.

Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floptical disks. media), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter or the like as well as machine language codes such as those produced by a compiler. The above hardware devices may be configured to act as one or more software modules to perform processing according to the present invention and vice versa.

In the above, the present invention has been described by specific details such as specific components and limited embodiments and drawings, but these are provided only to help a more general understanding of the present invention, and the present invention is not limited to the above embodiments. No, and those skilled in the art to which the present invention pertains may seek various modifications and variations from these descriptions.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments and should not be determined, and not only the claims to be described later, but also all modifications equivalent or equivalent to these claims fall within the scope of the spirit of the present invention. would be said to belong.

Claims (20)

determining prediction information to be used for encoding a target block;
generating encoded coding information by performing encoding on the coding information; and
Generating a bitstream including the encoded coding information
An image encoding method comprising a. According to claim 1,
The prediction information includes a motion information candidate list and final motion information. According to claim 1,
performing inter prediction on the target block;
An image encoding method further comprising a. According to claim 3,
The motion vector search method for the inter prediction is used,
The video encoding method according to claim 1 , wherein the motion vector search method is template matching. According to claim 4,
The inter prediction is bidirectional prediction,
An image encoding method in which a motion vector is fixed for one of the directions of the bi-directional prediction. According to claim 5,
An image encoding method in which motion search is performed for the other one of the directions. According to claim 3,
The motion vector search method for the inter prediction is used,
The motion vector search method is an image encoding method of matching both sides. obtaining a bitstream including encoded coding information;
generating coding information by decoding the encoded coding information; and
Determining prediction information to be used for decoding a target block
Image decoding method comprising a. According to claim 8,
The prediction information includes a motion information candidate list and final motion information. According to claim 8,
performing inter prediction on the target block using the prediction information;
Image decoding method further comprising. According to claim 10,
The motion vector search method for the inter prediction is used,
The motion vector search method is template matching. According to claim 11,
The inter prediction is bidirectional prediction,
An image decoding method in which a motion vector is fixed for one of the directions of the bi-directional prediction. According to claim 12,
An image decoding method in which a motion search is performed for the other one of the directions. According to claim 10,
The motion vector search method for the inter prediction is used,
The motion vector search method is a video decoding method of matching both sides. A computer-readable recording medium storing a bitstream for image decoding, wherein the bitstream comprises:
Coded Coding Information
including,
Decoding of the encoded coding information is performed to generate coding information;
A computer-readable recording medium on which prediction information to be used for decoding a target block is determined. According to claim 15,
The prediction information includes a motion information candidate list and final motion information. According to claim 15,
A computer readable recording medium in which inter prediction is performed on the target block using the prediction information. According to claim 17,
The motion vector search method for the inter prediction is used,
The motion vector search method is template matching. According to claim 18,
The inter prediction is bidirectional prediction,
A computer-readable recording medium in which a motion vector is fixed for one of the directions of the bi-directional prediction. According to claim 19,
A computer-readable recording medium in which motion search is performed for the other one of the directions.

Images