Method, apparatus, and recording medium for image encoding / decoding
By generating multiple prediction blocks and performing weighting and summing operations, the final prediction block is derived using the prediction information of surrounding blocks, the problem of insufficient image encoding/decoding accuracy in the prior art is solved, and the situation of not increasing signal transmission bits is realized to improve prediction accuracy.
Patent Information
- Application Number
- CN202380033938.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-05-10
- Filing Date
- 2023-04-11
- Publication Date
- 2024-11-22
AI Technical Summary
The existing image encoding/decoding technology is difficult to effectively improve the prediction accuracy of high-definition images, especially in high-resolution and high-resolution image processing. The existing technology cannot effectively use surrounding blocks to deduce prediction information, resulting in signal transmission bits. The number increases.
By generating multiple prediction blocks and performing weighting and summing operations, the final prediction block is derived using the prediction information of surrounding blocks, and the weight and calculation area of the prediction block are adaptively determined, thereby improving prediction accuracy.
Without increasing the number of signal transmission bits, the prediction accuracy of image encoding/decoding is improved, and the efficiency and quality of image processing are improved.
Smart Images

Figure CN119013993A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure generally relates to methods, devices, and storage media for image encoding / decoding. Background Art
[0002] With the continuous development of the information and communication industry, broadcast services supporting high definition (HD) resolution have become popular all over the world. Through this popularity, a large number of users have become accustomed to high-resolution and high-definition images and / or videos.
[0003] In order to meet the user's demand for high definition, many organizations have accelerated the development of next-generation imaging devices. User interest in UHD TV, which has a resolution more than four times that of full HD (FHD) TV, has increased as well as high-definition TV (HDTV) and FHD TV. As people's interest in it increases, image encoding / decoding technology for images with higher resolution and higher definition is currently required.
[0004] As the image compression technology, there are various technologies such as an inter-frame prediction technology, an intra-frame prediction technology, transformation, a quantization technology, a filtering technology, and an entropy coding technology.
[0005] The inter-frame prediction technique is a technique for predicting the value of a pixel included in the current picture using a picture before and / or after the current picture. The intra-frame prediction technique is a technique for predicting the value of a pixel included in the current picture using information about the pixel in the current picture. The transform and quantization technique may be a technique for compressing the energy of a residual signal. The entropy coding technique is a technique for assigning short codewords to frequently occurring values and assigning long codewords to less frequently occurring values.
[0006] By utilizing this image compression technology, data about images can be efficiently compressed, transmitted, and stored. Summary of the invention
[0007] Technical issues
[0008] The purpose of the present disclosure is to provide a method for obtaining multiple prediction blocks through multi-prediction and obtaining a final prediction block based thereon, and a device used for the method.
[0009] An object of the present disclosure is to provide a method for deriving prediction information for multi-prediction from surrounding blocks and an apparatus for performing the same.
[0010] An object of the present disclosure is to provide a method for adaptively determining at least one of a weighted sum operation application area or a weight of a plurality of prediction blocks and an apparatus for performing the same.
[0011] Technical Solution
[0012] The image decoding method according to an embodiment of the present disclosure includes: generating a first prediction block of a current block; generating a second prediction block of the current block or a subblock within the current block; and generating a final prediction block of the subblock or the current block by a weighted sum of the first prediction block and the second prediction block. In this case, the prediction information used to generate the first prediction block may be different from the prediction information used to generate the second prediction block.
[0013] The image encoding method according to an embodiment of the present disclosure includes: generating a first prediction block of a current block; generating a second prediction block of the current block or a subblock within the current block; and generating a final prediction block of the subblock or the current block by a weighted sum of the first prediction block and the second prediction block. In this case, the prediction information used to generate the first prediction block may be different from the prediction information used to generate the second prediction block.
[0014] In the image encoding / decoding method according to the embodiment of the present disclosure, the second prediction block may be acquired based on prediction information of a surrounding block located outside the current block and adjacent to the subblock.
[0015] In the image encoding / decoding method according to an embodiment of the present disclosure, when the difference between the maximum value and the minimum value of the prediction samples within the first prediction block is less than a threshold value, generation of the second prediction block may be allowed.
[0016] In the image encoding / decoding method according to an embodiment of the present disclosure, the number of main gradients may be determined from a gradient histogram generated based on the gradient of each prediction sample in the first prediction block, and when the number of main gradients is less than a threshold, generation of a second prediction block may be allowed.
[0017] In the image encoding / decoding method according to an embodiment of the present disclosure, a first prediction block may be generated based on a first intra prediction mode and a second prediction block may be generated based on a second intra prediction mode different from the first intra prediction mode.
[0018] In the image encoding / decoding method according to an embodiment of the present disclosure, the second intra prediction mode may be used to decode surrounding blocks adjacent to the subblock.
[0019] In the image encoding / decoding method according to the embodiment of the present disclosure, when a plurality of prediction blocks other than the first prediction block are additionally generated, a weighted sum operation may be sequentially performed on the plurality of prediction blocks.
[0020] In the image encoding / decoding method according to an embodiment of the present disclosure, the order in which the weighted sum operation is performed for the multiple prediction blocks can be determined based on the priority of the surrounding blocks, wherein the prediction information for each prediction block in the multiple prediction blocks is derived from the surrounding blocks.
[0021] In the image encoding / decoding method according to an embodiment of the present disclosure, the order of performing the weighted sum operation for the plurality of prediction blocks may be determined according to a prediction mode for each of the plurality of prediction blocks.
[0022] In the image encoding / decoding method according to an embodiment of the present disclosure, the weights applied to the first prediction block and the second prediction block can be determined by comparing a first matching cost with a second matching cost, wherein the first matching cost is obtained by applying first prediction information for the first prediction block to a surrounding template of the sub-block, and the second matching cost is obtained by applying second prediction information for the second prediction block to the template.
[0023] In the image encoding / decoding method according to the embodiment of the present disclosure, the weight may be determined by additionally considering a third matching cost obtained by performing a weighted sum on the first matching cost and the second matching cost.
[0024] In the image encoding / decoding method according to the embodiment of the present disclosure, the first matching cost and the second matching cost may be obtained by using the sub-sampling position within the template.
[0025] In the image encoding / decoding method according to the embodiment of the present disclosure, the additional prediction block may be obtained only for the sub-blocks adjacent to the boundary within the current block.
[0026] In the present disclosure, a recording medium storing a bit stream generated by an image encoding method may be provided.
[0027] Technical Effects
[0028] According to the present disclosure, prediction accuracy can be improved by obtaining a plurality of prediction blocks through multi-prediction and obtaining a final prediction block based thereon.
[0029] According to the present disclosure, by using surrounding blocks, prediction information for multi-prediction may be derived without increasing the number of signaled bits.
[0030] According to the present disclosure, prediction accuracy can be improved by adaptively determining at least one of weights of a plurality of prediction blocks or a weighted sum operation application area. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 is a block diagram showing a configuration of an embodiment of an encoding device to which the present disclosure is applied;
[0032] Figure 2 is a block diagram showing a configuration of an embodiment of a decoding device to which the present disclosure is applied;
[0033] Figure 3is a diagram schematically showing a partition structure of an image when the image is encoded and decoded;
[0034] Figure 4 is a diagram showing the form of a prediction unit (PU) that a coding unit (CU) may include;
[0035] Figure 5 is a diagram showing a form of a transform unit (TU) that may be included in a CU;
[0036] Figure 6 shows segmentation of blocks according to an example;
[0037] Figure 7 is a diagram for explaining an embodiment of an intra prediction process;
[0038] Figure 8 is a diagram showing reference samples used in an intra prediction process;
[0039] Fig. 9 is a diagram for explaining an embodiment of an inter-frame prediction process;
[0040] Fig.10 shows spatial candidates according to an embodiment;
[0041] Fig.11 shows the order in which motion information of spatial candidates is added to a merge list according to an embodiment;
[0042] Fig.12 shows a transform and quantization process according to an example;
[0043] Fig.13 shows a diagonal scan according to an example;
[0044] Fig.14 shows a horizontal scan according to an example;
[0045] Fig.15 shows vertical scanning according to an example;
[0046] Fig.16 is a configuration diagram of an encoding device according to an embodiment;
[0047] Fig.17 is a configuration diagram of a decoding device according to an embodiment;
[0048] Fig.18 An example of applying template matching is shown.
[0049] Fig.19 An example of a configuration template is shown.
[0050] Fig. 20 An example is shown where subsampling is applied to a template.
[0051] Fig.21 is a diagram for describing an example of applying template matching when an affine mode is applied to a current block.
[0052] Fig. 22 and Fig.23 Flowcharts of methods for obtaining a prediction block based on multi-prediction in an encoding device and a decoding device are respectively shown.
[0053] Fig.24 The position and size of each of the plurality of prediction blocks are shown.
[0054] Fig.25 Shows surrounding blocks adjacent to the current block.
[0055] Fig.26 is a diagram showing the boundaries of prediction blocks generated by multi-prediction.
[0056] Fig. 27 and Fig.28 The conditions for determining the strength of the deblocking filter are shown.
[0057] Fig.29 An example of deriving final prediction samples from boundary regions between prediction blocks is shown.
[0058] Fig.30 and Fig.31 Shows the positions of reference samples or surrounding blocks referenced to generate a prediction block.
[0059] Fig.32 The surrounding blocks referenced to derive motion information of the sub-block are shown.
[0060] Fig.33 and Fig.34 Shows the reference order of surrounding blocks.
[0061] Fig.35 is a diagram for describing an example of referring to motion information of surrounding blocks of a reference block.
[0062] Fig.36 Shows the subblock groups within the current block.
[0063] Fig.37 An example of determining a referenceable surrounding block according to the POC of a reference picture is shown.
[0064] Fig.38 is a diagram for describing an example of deriving an intra prediction mode per subblock.
[0065] Fig.39 is a diagram showing a weight map.
[0066] Fig.40 and Fig.41Syntax elements for multi-prediction including weights and offsets are shown.
[0067] Fig.42 The surrounding samples and samples within the current block used to calculate the similarity are shown.
[0068] Fig.43 An example is shown in which different weights are set for each row in a prediction block.
[0069] Fig.44 An example of selecting a prediction block for weighted sum operation according to prediction modes of surrounding blocks is shown.
[0070] Fig.45 and Fig.46 The execution rules of the weighted sum operation are shown.
[0071] Fig.47 is a diagram showing a region where weighted sum is performed. DETAILED DESCRIPTION
[0072] The present invention can be variously modified and can have various embodiments, and specific embodiments will be described in detail below with reference to the accompanying drawings. However, it should be understood that those embodiments are not intended to limit the present invention to a specific disclosed form, and they include all changes, equivalents or modifications included in the spirit and scope of the present invention.
[0073] The following detailed description of exemplary embodiments will be made with reference to the accompanying drawings showing specific embodiments. These embodiments are described so that a technician with ordinary knowledge in the technical field to which the present disclosure belongs can easily practice the embodiments. It should be noted that the various embodiments are different from each other, but do not need to be mutually exclusive. For example, without departing from the spirit and scope of the embodiments related to the embodiments, the specific shapes, structures and characteristics described herein can be implemented as other embodiments. In addition, it should be understood that the position or arrangement of each component in each disclosed embodiment can be changed without departing from the spirit and scope of the embodiments. Therefore, the attached detailed description is not intended to limit the scope of the present disclosure, and the scope of the exemplary embodiments is limited only by the attached claims and their equivalents, as long as they are appropriately described.
[0074] In the drawings, like reference numerals are used to designate the same or similar functions in various aspects. The shapes, sizes, etc. of components in the drawings may be exaggerated to make the description clear.
[0075] Terms such as "first" and "second" can be used to describe various components, but the components are not limited by the terms. These terms are only used to distinguish one component from another component. For example, without departing from the scope of this specification, a first component can be named a second component. Similarly, a second component can be named a first component. The term "and / or" can include a combination of multiple related description items or any one of multiple related description items.
[0076] It will be understood that when a component is referred to as being "connected" or "coupled" to another component, the two components may be directly connected or coupled to each other, or there may be an intermediate component between the two components. On the other hand, it will be understood that when a component is referred to as being "directly connected or coupled", there are no intermediate components between the two components.
[0077] In addition, in order to indicate different characteristic functions, the components described in the embodiments are shown independently, but this does not mean that each component is formed by separate hardware or software. That is, for ease of description, the components are arranged and included separately. For example, at least two of the components can be integrated into a single component. On the contrary, a component can be divided into multiple components. The embodiment of the integrated component or the embodiment of separating some parts is included in the scope of this specification as long as it does not depart from the essence of this specification.
[0078] The terms used in the embodiments are only used to describe specific embodiments and are not intended to limit the present invention. Unless otherwise specifically indicated in the context, singular expressions include plural expressions. In the embodiments, it should be understood that terms such as "including" or "having" are only intended to indicate the presence of features, numbers, steps, operations, components, parts, or a combination thereof, and are not intended to exclude the possibility that one or more other features, numbers, steps, operations, components, parts, or a combination thereof will exist or be added. That is, in the embodiments, the expression that describes a component "including" a specific component means that additional components may be included within the scope of the practice of the present invention or the technical spirit of the present invention, but does not exclude the presence of components other than the specific components.
[0079] In an embodiment, the term "at least one" may mean one of 1 or more numbers, such as 1, 2, 3, and 4. In an embodiment, the term "plurality" may mean one of 2 or more numbers, such as 2, 3, and 4.
[0080] Some components of the embodiment are not essential components for performing essential functions, but may be optional components for only improving performance. The embodiment may be implemented using only essential components for implementing the essence of the embodiment. For example, a structure including only essential constituent elements and not including optional constituent elements used only to improve performance is also included in the scope of the embodiment.
[0081] The embodiments will be described in detail below with reference to the accompanying drawings so that a person with ordinary knowledge in the technical field to which the embodiments belong can easily practice the embodiments. In the following description of the embodiments, detailed descriptions of known functions or configurations that are considered to obscure the subject matter of this specification will be omitted. In addition, throughout the drawings, the same reference numerals are used to represent the same components, and repeated descriptions of the same components will be omitted.
[0082] Hereinafter, "image" may refer to a single picture constituting a video, or may refer to the video itself. For example, "encoding and / or decoding of an image" may refer to "encoding and / or decoding of a video", and may also refer to "encoding and / or decoding of any one of the images constituting the video".
[0083] Hereinafter, the terms 'video' and 'moving picture' may be used to have the same meaning and may be used interchangeably with each other.
[0084] Hereinafter, the target image may be an encoding target image as a target to be encoded and / or a decoding target image as a target to be decoded. In addition, the target image may be an input image input to an encoding device or an input image input to a decoding device. And, the target image may be a current image, that is, a target to be currently encoded and / or decoded. For example, the terms "target image" and "current image" may be used to have the same meaning and may be used interchangeably with each other.
[0085] Hereinafter, the terms “image”, “picture”, “frame”, and “screen” may be used to have the same meaning and may be used interchangeably with each other.
[0086] Hereinafter, the target block may be an encoding target block (i.e., a target to be encoded) and / or a decoding target block (i.e., a target to be decoded). In addition, the target block may be a current block, i.e., a target to be encoded and / or decoded currently. Here, the terms "target block" and "current block" may be used to have the same meaning and may be used interchangeably with each other. The current block may represent an encoding target block that is an encoding target during encoding and / or a decoding target block that is a decoding target during decoding. In addition, the current block may be at least one of a coding block, a prediction block, a residual block, and a transform block.
[0087] Hereinafter, the terms "block" and "unit" may be used to have the same meaning and may be used interchangeably with each other. Alternatively, a "block" may refer to a specific unit.
[0088] Hereinafter, the terms "region" and "segment" may be used interchangeably with each other.
[0089] In the following embodiments, specific information, data, flags, indexes, elements, and attributes may have their respective values. The value "0" corresponding to each of the information, data, flags, indexes, elements, and attributes may indicate false, logically false, or a first predefined value. In other words, the value "0", false, logically false, and a first predefined value may be used interchangeably. The value "1" corresponding to each of the information, data, flags, indexes, elements, and attributes may indicate true, logically true, or a second predefined value. In other words, the value "1", true, logically true, and a second predefined value may be used interchangeably.
[0090] 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 of 0 or greater or an integer of 1 or greater. In other words, in an embodiment, each of the row, column, and index may be counted from 0 or may be counted from 1.
[0091] In an embodiment, the term “one or more” or the term “at least one” may mean the term “plurality.” The term “one or more” or the term “at least one” may be used interchangeably with “plurality.”
[0092] Next, terms to be used in the embodiments will be described.
[0093] Encoder: Encoder means a device for performing encoding. That is, encoder can mean an encoding device.
[0094] Decoder: A decoder refers to a device for performing decoding. That is, a decoder may mean a decoding device.
[0095] Unit: A unit may mean a unit of image encoding and decoding. The terms "unit" and "block" may be used to have the same meaning and may be used interchangeably with each other.
[0096] - A cell may be an M x N array of samples. Each of M and N may be a positive integer. A cell may generally mean an array of samples in a two-dimensional form.
[0097] -In encoding and decoding of an image, a "unit" may be a region generated by partitioning one image. In other words, a "unit" may be a region specified in one image. A single image may be partitioned into a plurality of units. Alternatively, one image may be partitioned into sub-parts, and when encoding or decoding is performed on the partitioned sub-parts, a unit may represent each partitioned sub-part.
[0098] - In encoding and decoding of an image, a predefined process may be performed on each unit according to the type of the unit.
[0099] - According to functions, the unit type may be classified into a macro unit, a coding unit (CU), a prediction unit (PU), a residual unit, a transform unit (TU), etc. Alternatively, according to functions, the unit may mean a block, a macro block, a coding tree unit, a coding tree block, a coding unit, a coding block, a prediction unit, a prediction block, a residual unit, a residual block, a transform unit, a transform block, etc. For example, a target unit as a target of encoding and / or decoding may be at least one of a CU, a PU, a residual unit, and a TU.
[0100] - The term "unit" may mean information including a luma component block, a chroma component block corresponding to the luma component block, and syntax elements for the respective blocks, so that the unit is designated to be distinguished from the block.
[0101] - The size and shape of the unit can be implemented in various ways. In addition, the unit can have any of a variety of sizes and shapes. In particular, the shape of the unit can include not only a square, but also a geometric figure that can be represented in two dimensions (2D), such as a rectangle, a trapezoid, a triangle, and a pentagon.
[0102] - In addition, the unit information may include one or more of a type of unit, a size of the unit, a depth of the unit, a coding order of the unit, and a decoding order of the unit, etc. For example, the type of the unit may indicate one of CU, PU, residual unit, and TU.
[0103] - A cell can be partitioned into sub-cells, each sub-cell having a size smaller than that of the associated cell.
[0104] Depth: Depth may mean the degree to which a unit is partitioned. In addition, the depth of a unit may indicate the level at which the corresponding unit exists when the unit is represented by a tree structure.
[0105] - The cell partition information may include a depth indicating the depth of the cell. The depth may indicate the number of times the cell is partitioned and / or the extent to which the cell is partitioned.
[0106] -In a tree structure, the root node can be considered to have the smallest depth and the leaf node can be considered to have the largest depth. The root node can be the highest (top) node. The leaf node can be the lowest node.
[0107] - A single unit may be hierarchically partitioned into a plurality of subunits while having depth information based on a tree structure. In other words, a unit and a subunit generated by partitioning a unit may correspond to a node and a subnode of a node, respectively. Each of the partitioned subunits may have a unit depth. Since the depth indicates the number of times a unit is partitioned and / or the degree to which a unit is partitioned, the partition information of a subunit may include information about the size of the subunit.
[0108] -In the tree structure, the top node may correspond to the initial node before partitioning. The top node may be referred to as a "root node". In addition, the root node may have a minimum depth value. Here, the top node may have a depth of level "0".
[0109] - A node with a depth of level "1" may represent a cell generated when the initial cell is partitioned once. A node with a depth of level "2" may represent a cell generated when the initial cell is partitioned twice.
[0110] - A leaf node with a depth of level "n" may represent a cell generated when the initial cell has been partitioned n times.
[0111] - A leaf node may be a bottom node, which cannot be partitioned further. The depth of a leaf node may be a maximum level. For example, the predefined value of the maximum level may be 3.
[0112] -QT depth may indicate the depth used for quad-partitioning. BT depth may indicate the depth used for binary-partitioning. TT depth may indicate the depth used for ternary-partitioning.
[0113] Sample: A sample can be the basic unit of a block. Depending on the bit depth (Bd), it can be from 0 to 2 Bd A value of -1 indicates a sample.
[0114] - A sample can be a pixel or a pixel value.
[0115] - Hereinafter, the terms "pixel" and "sample" may be used to have the same meaning and may be used interchangeably with each other.
[0116] Coding Tree Unit (CTU): A CTU may consist of a single luma component (Y) coding tree block and two chroma component (Cb, Cr) coding tree blocks related to the luma component coding tree block. In addition, CTU may mean information including the above blocks and syntax elements for each of the blocks.
[0117] - Each coding tree unit (CTU) may be partitioned using one or more partitioning methods such as quadtree (QT), binary tree (BT), and ternary tree (TT) to configure sub-units such as coding units, prediction units, and transform units. Quadtree may mean a quadtree. In addition, each coding tree unit may be partitioned using a multi-type tree (MTT) using one or more partitioning methods.
[0118] - "CTU" may be used as a term to designate a pixel block, which is a processing unit in image decoding and encoding processes, such as in the case of input image partitioning.
[0119] Coding Tree Block (CTB): “CTB” may be used as a term to designate any one of a Y coding tree block, a Cb coding tree block, and a Cr coding tree block.
[0120] Neighboring block: A neighboring block (or neighboring block) may mean a block adjacent to a target block. A neighboring block may mean a reconstructed neighboring block.
[0121] Hereinafter, the terms “neighboring block” and “adjacent block” may be used to have the same meaning and may be used interchangeably with each other.
[0122] The neighboring block may mean a reconstructed neighboring block.
[0123] Spatial neighboring blocks: Spatial neighboring blocks may be blocks that are spatially adjacent to the target block. Neighboring blocks may include spatial neighboring blocks.
[0124] - The target block and spatially neighboring blocks may be included in the target picture.
[0125] - The spatially neighboring block may mean a block whose boundary contacts the target block, or a block located within a predetermined distance from the target block.
[0126] - A spatially adjacent block may mean a block adjacent to a vertex of a target block. Here, a block adjacent to a vertex of a target block may mean 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.
[0127] Temporally neighboring blocks: Temporally neighboring blocks may be blocks that are temporally adjacent to the target block. Neighboring blocks may include temporally neighboring blocks.
[0128] - Temporally neighboring blocks may include co-located blocks (col blocks).
[0129] The col block may be a block in a previously reconstructed co-located picture (col picture). The position of the col block in the col picture may correspond to the position of the target block in the target picture. Alternatively, the position of the col block in the col picture may be equal to the position of the target block in the target picture. The col picture may be a picture included in a reference picture list.
[0130] - The temporal neighboring block may be a block that is temporally adjacent to the spatial neighboring block of the target block.
[0131] Prediction mode: The prediction mode may be information indicating a mode for intra prediction or a mode for inter prediction.
[0132] Prediction unit: A prediction unit may be a basic unit for prediction such as inter prediction, intra prediction, inter compensation, intra compensation, and motion compensation.
[0133] - A single prediction unit can be divided into multiple partitions or sub-prediction units with smaller sizes. Multiple partitions can also be basic units in the execution of prediction or compensation. Partitions generated by dividing a prediction unit can also be prediction units.
[0134] Prediction unit partition: A prediction unit partition may be a shape into which a prediction unit is divided.
[0135] Reconstructed neighboring unit: The reconstructed neighboring unit may be an already decoded and reconstructed unit that is neighboring to the target unit.
[0136] - The reconstructed neighboring cell may be a cell that is spatially adjacent to the target cell or temporally adjacent to the target cell.
[0137] The reconstructed spatially neighboring unit may be a unit that is included in the target picture and has been reconstructed through encoding and / or decoding.
[0138] -The reconstructed temporally neighboring unit may be a unit that is included in a reference image and has been reconstructed by encoding and / or decoding. The position of the reconstructed temporally neighboring unit in the reference image may be the same as the position of the target unit in the target picture, or may correspond to the position of the target unit in the target picture. Also, the reconstructed temporally neighboring unit may be a block that is adjacent to the corresponding block in the reference image. Here, the position of the corresponding block in the reference image may correspond to the position of the target block in the target image. Here, the fact that the positions of the blocks correspond to each other may mean that the positions of the blocks are the same as each other, may mean that one block is included in another block, or may mean that one block occupies a specific position in another block.
[0139] Sub-picture: A picture can be divided into one or more sub-pictures. A sub-picture can be composed of one or more tile rows and one or more tile columns.
[0140] - A sub-picture may be a region in a picture having a square shape or a rectangular (ie, non-square rectangular) shape. In addition, a sub-picture may include one or more CTUs.
[0141] - A sub-picture can be a rectangular area of one or more slices in a picture.
[0142] - A sub-picture may include one or more tiles, one or more partitions and / or one or more slices.
[0143] Tile: A tile may be an area in a picture having a square shape or a rectangular (ie, non-square rectangular) shape.
[0144] - A tile may include one or more CTUs.
[0145] - A parallel block can be partitioned into one or more partitions.
[0146] Block: A block may represent one or more CTU rows in a tile.
[0147] - A tile can be partitioned into one or more partitions. Each partition can include one or more CTU rows.
[0148] - Parallel blocks that are not partitioned into two parts can also represent chunking.
[0149] Slice: A slice may contain one or more tiles in a picture. Alternatively, a slice may include one or more partitions in a tile.
[0150] - A sub-picture may contain one or more slices that together cover a rectangular area of the picture. Thus, every sub-picture boundary is always also a slice boundary, and every vertical sub-picture boundary is always also a vertical tile boundary.
[0151] Parameter set: A parameter set may correspond to header information in the internal structure of a bitstream.
[0152] The parameter set may include at least one of a video parameter set (VPS), a sequence parameter set (SPS), a picture parameter set (PPS), an adaptation parameter set (APS), a decoding parameter set (DPS), etc.
[0153] - Information signaled by each parameter set can be applied to pictures that reference the corresponding parameter set. For example, information in a VPS can be applied to pictures that reference the VPS. Information in an SPS can be applied to pictures that reference the SPS. Information in a PPS can be applied to pictures that reference the PPS.
[0154] -Each parameter set can refer to a higher parameter set. For example, PPS can refer to SPS. SPS can refer to VPS.
[0155] - In addition, the parameter set may include a parallel block group, a slice header information, and a parallel block header information. A parallel block group may be a group including a plurality of parallel blocks. In addition, the meaning of "parallel block group" may be the same as the meaning of "slice".
[0156] Rate-distortion optimization: The encoding device may use rate-distortion optimization to provide high encoding efficiency by utilizing a combination of coding unit (CU) size, prediction mode, prediction unit (PU) size, motion information, and transform unit (TU) size.
[0157] -The rate-distortion optimization scheme can calculate the rate-distortion cost of the corresponding combination in order to select the best combination from the combination. The rate-distortion cost can be calculated using the equation "D+(λ*R)". Generally, the combination that can minimize the rate-distortion cost can be selected as the optimal combination in the rate-distortion optimization scheme.
[0158] -D may represent distortion. D may be the average of the squares of the differences between the original transform coefficients and the reconstructed transform coefficients in the transform unit (ie, the mean square error).
[0159] -R may represent rate, which may indicate the bit rate using relevant context information.
[0160] -λ represents the Lagrange multiplier. R may include not only encoding parameter information such as prediction mode, motion information, and coding block flag, but also bits generated due to encoding of transform coefficients.
[0161] - The encoding device may perform processes such as inter-frame prediction and / or intra-frame prediction, transformation, quantization, entropy coding, inverse quantization (dequantization) and / or inverse transformation in order to calculate accurate D and R. These processes may significantly increase the complexity of the encoding device.
[0162] - Bitstream: Bitstream may refer to a bit stream including encoded image information.
[0163] Parsing: Parsing may be the decision on the value of a syntax element made by performing entropy decoding on the bitstream. Alternatively, the term "parsing" may mean such entropy decoding itself.
[0164] Symbol: A symbol may be at least one of a syntax element, a coding parameter, and a transform coefficient of a coding target unit and / or a decoding target unit. In addition, a symbol may be a target of entropy coding or a result of entropy decoding.
[0165] Reference picture: A reference picture may be an image referenced by a unit in order to perform inter-frame prediction or motion compensation. Alternatively, a reference picture may be an image including a reference unit referenced by a target unit in order to perform inter-frame prediction or motion compensation.
[0166] Hereinafter, the terms 'reference picture' and 'reference image' may be used to have the same meaning and may be used interchangeably with each other.
[0167] Reference picture list: A reference picture list may be a list including one or more reference pictures used for inter prediction or motion compensation.
[0168] - The types of reference picture lists may include List Combination (LC), List 0 (L0), List 1 (L1), List 2 (L2), List 3 (L3), etc.
[0169] - For inter prediction, one or more reference picture lists may be used.
[0170] Inter prediction indicator: The inter prediction indicator may indicate the inter prediction direction of the target unit. The inter prediction may be one of unidirectional prediction and bidirectional prediction. Alternatively, the inter prediction indicator may indicate the number of reference pictures used to generate the prediction unit of the 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.
[0171] Prediction list utilization flag: The prediction list utilization flag may indicate whether to use at least one reference picture in a specific reference picture list to generate a prediction unit.
[0172] - The prediction list utilization flag may be used to derive the inter prediction indicator. Conversely, the prediction list utilization flag may be used to derive the prediction list utilization flag. For example, the case where the prediction list utilization flag indicates "0" as a first value may indicate that, for the target unit, the reference picture in the reference picture list is not used to generate the prediction block. The case where the prediction list utilization flag indicates "1" as a second value may indicate that, for the target unit, the reference picture list is used to generate the prediction unit.
[0173] Reference picture index: The reference picture index may be an index indicating a specific reference picture in a reference picture list.
[0174] Picture Order Count (POC): The POC value of a picture may indicate the order in which the corresponding picture is displayed.
[0175] Motion Vector (MV): A motion vector may be a 2D vector used for inter prediction or motion compensation. A motion vector may mean the offset between a target image and a reference image.
[0176] - For example, MV can be represented as (mv x ,mv y ) format. x can indicate the horizontal component, and mv y The vertical component may be indicated.
[0177] - Search range: The search range may be a 2D area where MV search is performed during inter prediction. For example, the size of the search range may be M×N. M and N may be respective positive integers.
[0178] Motion vector candidate: A motion vector candidate may be a block that is a prediction candidate when predicting a motion vector or a motion vector of a block that is a prediction candidate.
[0179] - The motion vector candidate may be included in the motion vector candidate list.
[0180] Motion vector candidate list: A motion vector candidate list may be a list configured using one or more motion vector candidates.
[0181] Motion vector candidate index: The motion vector candidate index may be an indicator for indicating a motion vector candidate in a motion vector candidate list. Alternatively, the motion vector candidate index may be an index of a motion vector predictor.
[0182] Motion information: The motion information may be information including at least one of a reference picture list, a reference image, a motion vector candidate, a motion vector candidate index, a merge candidate, and a merge index, as well as a motion vector, a reference picture index, and an inter prediction indicator.
[0183] Merge candidate list: A merge candidate list may be a list configured using one or more merge candidates.
[0184] Merge candidate: The merge candidate may be a spatial merge candidate, a temporal merge candidate, a combined merge candidate, a combined bi-predictive merge candidate, a history-based candidate, a candidate based on an average of two candidates, a zero merge candidate, etc. The merge candidate may include an inter prediction indicator, and may include motion information such as prediction type information, reference picture index for each list, motion vector, prediction list utilization flag, and an inter prediction indicator.
[0185] Merge index: The merge index may be an indicator for indicating a merge candidate in the merge candidate list.
[0186] - The merge index may indicate a reconstruction unit used for deriving a merge candidate between a reconstruction unit spatially adjacent to the target unit and a reconstruction unit temporally adjacent to the target unit.
[0187] - The merge index may indicate at least one of a plurality of pieces of motion information of a merge candidate.
[0188] Transform unit: A transform unit may be a basic unit of residual signal encoding and / or residual signal decoding (such as transform, inverse transform, quantization, inverse quantization, transform coefficient encoding, and transform coefficient decoding). A single transform unit may be partitioned into a plurality of sub-transform units having smaller sizes. Here, the transform may include one or more of a primary transform and a secondary transform, and the inverse transform may include one or more of a primary inverse transform and a secondary inverse transform.
[0189] Scaling: Scaling may refer to the process of multiplying a factor by the levels of transform coefficients.
[0190] - As a result of scaling of the transform coefficient levels, the transform coefficients may be generated. Scaling may also be referred to as "inverse quantization".
[0191] Quantization parameter (QP): A quantization parameter may be a value used to generate a transform coefficient level of a transform coefficient in quantization. Alternatively, a quantization parameter may also be a value used to generate a transform coefficient by scaling the transform coefficient level in inverse quantization. Alternatively, a quantization parameter may be a value mapped to a quantization step size.
[0192] Delta quantization parameter: The delta quantization parameter may mean the difference between the predicted quantization parameter and the quantization parameter of the target unit.
[0193] Scanning: Scanning may refer to a method for aligning the order of coefficients in a cell, block, or matrix. For example, a method for aligning a 2D array in the form of a one-dimensional (1D) array may be referred to as "scanning". Alternatively, a method for aligning a 1D array in the form of a 2D array may also be referred to as "scanning" or "inverse scanning".
[0194] Transform coefficient: The transform coefficient may be a coefficient value generated when the encoding device performs transformation. Alternatively, the transform coefficient may be a coefficient value generated when the decoding device performs at least one of entropy decoding and inverse quantization.
[0195] - Quantization levels or quantized transform coefficient levels generated by applying quantization to transform coefficients or residual signals may also be included in the meaning of the term "transform coefficient".
[0196] Quantization level: The quantization level may be a value generated when the encoding device performs quantization on a transform coefficient or a residual signal. Alternatively, the quantization level may be a value that is a target of inverse quantization when the decoding device performs inverse quantization.
[0197] -The quantized transform coefficient level as a result of transformation and quantization may also be included in the meaning of the quantization level.
[0198] Non-zero transform coefficient: A non-zero transform coefficient may be a transform coefficient having a value other than 0 or a transform coefficient level having a value other than 0. Alternatively, a non-zero transform coefficient may be a transform coefficient whose value is not zero or a transform coefficient level whose value is not zero.
[0199] Quantization matrix: A quantization matrix may be a matrix used in a quantization process or an inverse quantization process to improve the subjective image quality or the objective image quality of an image. A quantization matrix may also be referred to as a "scaling list".
[0200] Quantization matrix coefficient: A quantization matrix coefficient can be each element in the quantization matrix. A quantization matrix coefficient can also be referred to as a "matrix coefficient".
[0201] Default matrix: The default matrix may be a quantization matrix predefined by an encoding device and a decoding device.
[0202] Non-default matrix: A non-default matrix may be a quantization matrix that is not predefined by the encoding device and the decoding device. A non-default matrix may represent a quantization matrix that is signaled by a user from the encoding device to the decoding device.
[0203] Most Probable Mode (MPM): The MPM may represent an intra prediction mode having a high probability of being used for intra prediction of a target block.
[0204] The encoding device and the decoding device may determine one or more MPMs based on encoding parameters related to the target block and properties of an entity related to the target block.
[0205] The encoding device and the decoding device may determine one or more MPMs based on the intra prediction mode of the reference block. The reference block may include multiple reference blocks. The multiple reference blocks may include a spatially adjacent block adjacent to the left side of the target block and a spatially adjacent block adjacent to the top of the target block. In other words, depending on which intra prediction modes have been used for the reference block, one or more different MPMs may be determined.
[0206] - One or more MPMs may be determined in the same manner in the encoding device and the decoding device. That is, the encoding device and the decoding device may share the same MPM list including the one or more MPMs.
[0207] MPM list: The MPM list may be a list including one or more MPMs. The number of the one or more MPMs in the MPM list may be predefined.
[0208] MPM indicator: The MPM indicator may indicate an MPM to be used for intra prediction of a target block among one or more MPMs in an MPM list. For example, the MPM indicator may be an index of the MPM list.
[0209] - Since the MPM list is determined in the same manner in the encoding device and the decoding device, there is no need to transmit the MPM list itself from the encoding device to the decoding device.
[0210] - The MPM indicator may be signaled from the encoding apparatus to the decoding apparatus. When the MPM indicator is signaled, the decoding apparatus may determine an MPM to be used for intra prediction of a target block among the MPMs in the MPM list.
[0211] MPM usage indicator: The MPM usage indicator may indicate whether an MPM usage mode is to be used for prediction of a target block. The MPM usage mode may be a mode of determining an MPM to be used for intra prediction of a target block using an MPM list.
[0212] - An MPM usage indicator may be signaled from an encoding device to a decoding device.
[0213] Signaling: "Signaling" may mean that information is transmitted from an encoding device to a decoding device. Alternatively, "signaling" may mean that information is included in a bitstream or a recording medium by the encoding device. Information signaled by the encoding device may be used by a decoding device.
[0214] -The encoding device may generate encoding information by performing encoding on information to be transmitted with a signal. The encoding information may be sent from the encoding device to the decoding device. The decoding device may obtain the information by decoding the transmitted encoding information. Here, the encoding may be entropy encoding, and the decoding may be entropy decoding.
[0215] Selective Signaling: Information may be selectively signaled. Selective signaling of information may mean that an encoding device selectively includes the information in a bitstream or a recording medium (according to certain conditions). Selective signaling of information may mean that a decoding device selectively extracts the information from a bitstream (according to certain conditions).
[0216] Omission of signaling: Signaling for information may be omitted. Omission of signaling for information may mean that the encoding device (according to certain conditions) does not include the information in the bitstream or recording medium. Omission of signaling for information may mean that the decoding device (according to certain conditions) does not extract the information from the bitstream.
[0217] Statistic: A variable, encoding parameter, constant, etc. may have a value that can be calculated. A statistic may be a value generated by performing a calculation (operation) on the value of a specified target. For example, a statistic may indicate one or more of an average, a weighted average, a weighted sum, a minimum, a maximum, a modulus, a median, and an interpolated value of the values of a particular variable, a particular encoding parameter, a particular constant, etc.
[0218] Figure 1 is a block diagram showing the configuration of an embodiment of an encoding device to which the present disclosure is applied.
[0219] The encoding device 100 may be an encoder, a video encoding device, or an image encoding device. A video may include one or more images (pictures). The encoding device 100 may sequentially encode one or more images of a video.
[0220] Reference Figure 1 , the encoding device 100 includes an inter-frame prediction unit 110, an intra-frame prediction unit 120, a switch 115, a subtractor 125, a transform unit 130, a quantization unit 140, an entropy encoding unit 150, an inverse quantization (inverse quantization) unit 160, an inverse transform unit 170, an adder 175, a filter unit 180 and a reference picture buffer 190.
[0221] The encoding apparatus 100 may perform encoding on a target image using an intra mode and / or an inter mode. In other words, a prediction mode for a target block may be one of an intra mode and an inter mode.
[0222] Hereinafter, the terms “intra mode”, “intra prediction mode”, “intra-screen mode”, and “intra-screen prediction mode” may be used to have the same meaning and may be used interchangeably with each other.
[0223] Hereinafter, the terms “inter mode”, “inter prediction mode”, “inter-picture mode” and “inter-picture prediction mode” may be used to have the same meaning and may be used interchangeably with each other.
[0224] Hereinafter, the term "image" may indicate only a portion of an image, or may indicate a block. In addition, processing of an "image" may indicate sequential processing of a plurality of blocks.
[0225] In addition, the encoding device 100 can generate a bitstream including encoding information by encoding the target image, and can output and store the generated bitstream. The generated bitstream can be stored in a computer-readable storage medium and can be streamed through a wired and / or wireless transmission medium.
[0226] When the intra mode is used as the prediction mode, the switch 115 may switch to the intra mode. When the inter mode is used as the prediction mode, the switch 115 may switch to the inter mode.
[0227] The encoding apparatus 100 may generate a prediction block of the target block. In addition, after generating the prediction block, the encoding apparatus 100 may encode a residual block of the target block using a residual between the target block and the prediction block.
[0228] When the prediction mode is intra mode, the intra prediction unit 120 may use pixels of a previously encoded / decoded neighboring block adjacent to the target block as reference samples. The intra prediction unit 120 may perform spatial prediction on the target block using the reference samples, and may generate prediction samples of the target block via spatial prediction. The prediction samples may refer to samples in the prediction block.
[0229] The inter prediction unit 110 may include a motion prediction unit and a motion compensation unit.
[0230] When the prediction mode is inter mode, the motion prediction unit can search the reference image for the area that most closely matches the target block during the motion prediction process, and can derive a motion vector for the target block and the found area based on the found area. Here, the motion prediction unit can use the search range as the target area for the search.
[0231] The reference image may be stored in the reference picture buffer 190. More specifically, when encoding and / or decoding of the reference image has been processed, the encoded and / or decoded reference image may be stored in the reference picture buffer 190.
[0232] Since the decoded pictures are stored, the reference picture buffer 190 may be a decoded picture buffer (DPB).
[0233] The motion compensation unit may generate a prediction block of the target block by performing motion compensation using a motion vector. Here, the motion vector may be a two-dimensional (2D) vector for inter-frame prediction. In addition, the motion vector may indicate an offset between a target image and a reference image.
[0234] When the motion vector has a value other than an integer, the motion prediction unit and the motion compensation unit may generate a prediction block by applying an interpolation filter to a partial area of the reference image. In order to perform inter prediction or motion compensation, it may be determined which of the skip mode, merge mode, advanced motion vector prediction (AMVP) mode, and current picture reference mode corresponds to a method for predicting the motion of a PU included in a CU based on the CU and compensating for the motion, and inter prediction or motion compensation may be performed according to the mode.
[0235] The subtractor 125 may generate a residual block, which is the difference between the target block and the prediction block. The residual block may also be referred to as a "residual signal".
[0236] The residual signal may be the difference between the original signal and the prediction signal. Alternatively, the residual signal may be a signal generated by transforming or quantizing the difference between the original signal and the prediction signal or by transforming and quantizing the difference. The residual block may be a residual signal of a block unit.
[0237] The transform unit 130 may generate a transform coefficient by transforming the residual block, and may output the generated transform coefficient. Here, the transform coefficient may be a coefficient value generated by transforming the residual block.
[0238] The transform unit 130 may use one of a plurality of predefined transform methods when performing the transform.
[0239] The plurality of predefined transform methods may include discrete cosine transform (DCT), discrete sine transform (DST), Karhunen-Loeve transform (KLT), and the like.
[0240] The transform method for transforming the residual block may be determined based on at least one of the encoding parameters for the target block and / or the neighboring block. For example, the transform method may be determined based on at least one of the inter prediction mode for the PU, the intra prediction mode for the PU, the size of the TU, and the shape of the TU. Alternatively, transform information indicating the transform method may be transmitted from the encoding device 100 to the decoding device 200 using a signal.
[0241] When the transform skip mode is used, the transform unit 130 may omit transforming the residual block.
[0242] By applying quantization to the transform coefficients, quantized transform coefficient levels or quantization levels may be generated. Hereinafter, in an embodiment, each of the quantized transform coefficient levels and the quantization levels may also be referred to as a 'transform coefficient'.
[0243] The quantization unit 140 may generate a quantized transform coefficient level (ie, a quantization level or a quantization coefficient) by quantizing the transform coefficient according to the quantization parameter. The quantization unit 140 may output the generated quantized transform coefficient level. In this case, the quantization unit 140 may quantize the transform coefficient using a quantization matrix.
[0244] The entropy encoding unit 150 may generate a bitstream by performing entropy encoding based on probability distribution based on the value calculated by the quantization unit 140 and / or the encoding parameter value calculated in the encoding process. The entropy encoding unit 150 may output the generated bitstream.
[0245] The entropy encoding unit 150 may perform entropy encoding on information about pixels of an image and information required for decoding the image. For example, the information required for decoding the image may include syntax elements and the like.
[0246] When entropy coding is applied, fewer bits can be allocated to symbols that appear more frequently, and more bits can be allocated to symbols that appear less frequently. Since the symbols are represented by this allocation, the size of the bit string of the target symbol to be encoded can be reduced. Therefore, the compression performance of video coding can be improved by entropy coding.
[0247] In addition, for entropy coding, the entropy coding unit 150 may use a coding method such as exponential Golomb, context adaptive variable length coding (CAVLC) or context adaptive binary arithmetic coding (CABAC). For example, the entropy coding unit 150 may use a variable length coding / code (VLC) table to perform entropy coding. For example, the entropy coding unit 150 may derive a binarization method for a target symbol. In addition, the entropy coding unit 150 may derive a probability model of a target symbol / binary bit. The entropy coding unit 150 may use the derived binarization method, probability model and context model to perform arithmetic coding.
[0248] The entropy encoding unit 150 may transform coefficients in a 2D block form into a 1D vector form through a transform coefficient scanning method in order to encode quantized transform coefficient levels.
[0249] The coding parameters may be information required for coding and / or decoding. The coding parameters may include information encoded by the coding device 100 and transmitted from the coding device 100 to the decoding device, and may also include information that can be derived during the coding or decoding process. For example, the information transmitted to the decoding device may include syntax elements.
[0250] The coding parameters may include not only information (or flags or indexes) encoded by the coding device and transmitted by the coding device to the decoding device using a signal, such as syntax elements, but also information derived during the coding or decoding process. In addition, the coding parameters may include information required for encoding or decoding an image. For example, the coding parameters may include at least one value, combination or statistics of the following items: the size of the unit / block, the shape / form of the unit / block, the depth of the unit / block, the partition information of the unit / block, the partition structure of the unit / block, information indicating whether the unit / block is partitioned in a quadtree structure, information indicating whether the unit / block is partitioned in a binary tree structure, the partition direction of the binary tree structure (horizontal or vertical), the partition form of the binary tree structure (symmetric partitioning or asymmetric partitioning), information indicating whether the unit / block is partitioned in a ternary tree structure, the partition direction of the ternary tree structure (horizontal or vertical), the partition form of the ternary tree structure (symmetric partitioning or asymmetric partitioning, etc.). ), information indicating whether the unit / block is partitioned in a multi-type tree structure, a combination and direction of partitions of the multi-type tree structure (horizontal direction or vertical direction, etc.), a partition form of the multi-type tree structure (symmetric partition or asymmetric partition, etc.), a partition tree in the form of a multi-type tree (binary tree or ternary tree), a prediction type (intra-frame prediction or inter-frame prediction), an intra-frame prediction mode / direction, an intra-frame luminance prediction mode / direction, an intra-frame chrominance prediction mode / direction, an intra-frame partition information, an inter-frame partition information, a coding block partition flag, a prediction block partition flag, a transform block partition flag, a reference sample filtering method, a reference sample filter tap, a reference sample filter coefficient, a prediction block filtering method, a prediction block Filter taps, prediction block filter coefficients, prediction block boundary filtering method, prediction block boundary filter taps, prediction block boundary filter coefficients, inter-frame prediction mode, motion information, motion vector, motion vector difference, reference picture index, inter-frame prediction direction, inter-frame prediction indicator, prediction list utilization flag, reference picture list, reference image, POC, motion vector predictor, motion vector prediction index, motion vector prediction candidate, motion vector candidate list, information indicating whether to use merge mode, merge index, merge candidate, merge candidate list, information indicating whether to use skip mode, type of interpolation filter, taps of interpolation filter, filter coefficients of interpolation filter, size of a motion vector, precision of motion vector representation, transform type, transform size, information indicating whether a primary transform is used, information indicating whether an additional (secondary) transform is used, primary transform selection information (or primary transform index), secondary transform selection information (or secondary transform index), information indicating the presence or absence of a residual signal, coding block mode, coding block flag, quantization parameter, residual quantization parameter, quantization matrix, information about an in-loop filter, information indicating whether an in-loop filter is applied, coefficients of an in-loop filter, taps of an in-loop filter, shape / form of an in-loop filter, information indicating whether a deblocking filter is applied, coefficients of a deblocking filter,Taps of a deblocking filter, strength of a deblocking filter, shape / form of a deblocking filter, information indicating whether an adaptive sample offset is applied, a value of an adaptive sample offset, a category of an adaptive sample offset, a type of an adaptive sample offset, information indicating whether an adaptive intra-loop filter is applied, coefficients of an adaptive intra-loop filter, taps of an adaptive intra-loop filter, shape / form of an adaptive intra-loop filter, a binarization / debinarization method, a context model, a context model determination method, a context model update method, information indicating whether a normal mode is executed, information whether a bypass mode is executed, a valid coefficient flag, a last valid coefficient flag, a coding flag of a coefficient group, a position of a last valid coefficient, information indicating whether a coefficient value is greater than 1, information indicating whether a coefficient value is greater than 2, information indicating whether a coefficient value is greater than 3, remaining coefficient value information, sign information, reconstructed luminance samples, reconstructed chrominance samples, context binary bits, bypass binary bits The invention relates to a method for determining the encoding parameters of a coding parameter, ... ,
[0251] The first transform selection information may indicate a first transform applied to the target block.
[0252] The secondary transform selection information may indicate a secondary transform applied to the target block.
[0253] The residual signal may represent the difference between the original signal and the predicted signal. Alternatively, the residual signal may be a signal generated by transforming the difference between the original signal and the predicted signal. Alternatively, the residual signal may be a signal generated by transforming and quantizing the difference between the original signal and the predicted signal. The residual block may be the residual signal of the block.
[0254] Here, signaling information may mean that the encoding device 100 includes entropy encoding information generated by performing entropy encoding on a flag or an index in a bitstream, and the decoding device 200 obtains information by performing entropy decoding on the entropy encoding information extracted from the bitstream. Here, the information may include a flag, an index, etc.
[0255] A signal may mean information to be transmitted with a signal. Hereinafter, information for an image and a block may be referred to as a signal. In addition, hereinafter, the terms "information" and "signal" may be used to have the same meaning and may be used interchangeably with each other. For example, a specific signal may be a signal representing a specific block. An original signal may be a signal representing a target block. A prediction signal may be a signal representing a prediction block. A residual signal may be a signal representing a residual block.
[0256] The bitstream may include information based on a specific syntax. The encoding device 100 may generate a bitstream including information according to the specific syntax. The decoding device 200 may acquire information from the bitstream according to the specific syntax.
[0257] Since the encoding apparatus 100 performs encoding via inter-frame prediction, the encoded target image can be used as a reference image of another image to be processed subsequently. Therefore, the encoding apparatus 100 may reconstruct or decode the encoded target image and store the reconstructed or decoded image as a reference image in the reference picture buffer 190. For decoding, inverse quantization and inverse transformation of the encoded target image may be processed.
[0258] The quantization level may be inversely quantized by the inverse quantization unit 160, and may be inversely transformed by the inverse transform unit 170. The inverse quantization unit 160 may generate an inversely quantized coefficient by performing inverse quantization on the quantization level. The inverse transform unit 170 may generate an inversely quantized and inversely transformed coefficient by performing inverse transformation on the inversely quantized coefficient.
[0259] The inverse quantized and inverse transformed coefficients may be added to the prediction block by the adder 175. The inverse quantized and inverse transformed coefficients are added to the prediction block, and then a reconstructed block may be generated. Here, the inverse quantized and / or inverse transformed coefficients may represent coefficients on which one or more of inverse quantization and inverse transformation are performed, and may also represent a reconstructed residual block. Here, the reconstructed block may mean a restored block or a decoded block.
[0260] The reconstructed block may be filtered by the filter unit 180. The filter unit 180 may apply one or more of a deblocking filter, a sample adaptive offset (SAO) filter, an adaptive loop filter (ALF), and a non-local filter (NLF) to the reconstructed sample, the reconstructed block, or the reconstructed picture. The filter unit 180 may also be referred to as an "in-loop filter".
[0261] The deblocking filter can eliminate block distortion occurring at the boundary between blocks in the reconstructed picture. In order to determine whether to apply the deblocking filter, the number of rows or columns included in the block and including pixels for the following operation may be determined: whether to apply the deblocking filter to the target block based on the pixels.
[0262] When a deblocking filter is applied to a target block, the filter applied may be different according to the strength of the required deblocking filter. In other words, among different filters, a filter determined by considering the strength of the deblocking filter may be applied to the target block. When a deblocking filter is applied to a target block, one or more filters among a long tap filter, a strong filter, a weak filter, and a Gaussian filter may be applied to the target block according to the strength of the required deblocking filter.
[0263] In addition, when vertical filtering and horizontal filtering are performed on the target block, the horizontal filtering and the vertical filtering may be processed in parallel.
[0264] SAO may add a suitable offset to the value of a pixel to compensate for a coding error. SAO may perform correction of an offset using a difference between an original image and an image to which deblocking is applied on a pixel basis for an image to which deblocking is applied. In order to perform offset correction on an image, a method for dividing pixels included in an image into a certain number of regions, determining a region to which an offset is applied among the divided regions, and applying an offset to the determined region may be used, and a method for applying an offset in consideration of edge information of each pixel may also be used.
[0265] The ALF may perform filtering based on a value obtained by comparing the reconstructed image with the original image. After the pixels included in the image have been divided into a predetermined number of groups, the filter to be applied to each group may be determined, and filtering may be performed differently for each group. Information related to whether an adaptive loop filter is applied may be transmitted with a signal for each CU. Such information may be transmitted with a signal for a luminance signal. The shape and filter coefficients of the ALF to be applied to each block may be different for each block. Alternatively, an ALF having a fixed form may be applied to the block regardless of the characteristics of the block.
[0266] The non-local filter can perform filtering based on a reconstructed block similar to the target block. An area similar to the target block can be selected from the reconstructed picture, and the statistical characteristics of the selected similar area can be used to perform filtering of the target block. Information about whether to apply the non-local filter can be transmitted with a signal for the coding unit (CU). In addition, the shape and filter coefficients of the non-local filter to be applied to the block can be different according to the block.
[0267] The reconstructed block or the reconstructed image filtered by the filter unit 180 may be stored as a reference picture in the reference picture buffer 190. The reconstructed block filtered by the filter unit 180 may be a part of the reference picture. In other words, the reference picture may be a reconstructed picture composed of the reconstructed blocks filtered by the filter unit 180. The stored reference picture may be subsequently used for inter-frame prediction or motion compensation.
[0268] Figure 2 is a block diagram showing a configuration of an embodiment of a decoding device to which the present disclosure is applied.
[0269] The decoding device 200 may be a decoder, a video decoding device, or an image decoding device.
[0270] Reference Figure 2 , the decoding device 200 may include an entropy decoding unit 210, an inverse quantization (dequantization) unit 220, an inverse transform unit 230, an intra-frame prediction unit 240, an inter-frame prediction unit 250, a switch 245, an adder 255, a filter unit 260 and a reference picture buffer 270.
[0271] The decoding apparatus 200 may receive a bitstream output from the encoding apparatus 100. The decoding apparatus 200 may receive a bitstream stored in a computer-readable storage medium, and may receive a bitstream streamed through a wired / wireless transmission medium.
[0272] The decoding apparatus 200 may perform decoding on a bitstream in an intra mode and / or an inter mode. In addition, the decoding apparatus 200 may generate a reconstructed image or a decoded image through decoding, and may output the reconstructed image or the decoded image.
[0273] For example, switching to intra mode or inter mode based on the prediction mode for decoding may be performed by the switch 245. When the prediction mode for decoding is intra mode, the switch 245 may be operated to switch to intra mode. When the prediction mode for decoding is inter mode, the switch 245 may be operated to switch to inter mode.
[0274] The decoding device 200 can obtain a reconstructed residual block by decoding the input bit stream, and can generate a prediction block. When the reconstructed residual block and the prediction block are obtained, the decoding device 200 can generate a reconstructed block as a target to be decoded by adding the reconstructed residual block and the prediction block.
[0275] The entropy decoding unit 210 may generate symbols by performing entropy decoding on the bitstream based on the probability distribution of the bitstream. The generated symbols may include symbols in the form of quantized transform coefficient levels (i.e., quantization levels or quantization coefficients). Here, the entropy decoding method may be similar to the above-mentioned entropy encoding method. That is, the entropy decoding method may be an inverse process of the above-mentioned entropy encoding method.
[0276] The entropy decoding unit 210 may change a coefficient having a one-dimensional (1D) vector form into a 2D block shape through a transform coefficient scanning method in order to decode quantized transform coefficient levels.
[0277] For example, by scanning the block coefficients using an upper right diagonal scan, the coefficients of the block can be changed into a 2D block shape. Optionally, which of the upper right diagonal scan, vertical scan, and horizontal scan to use can be determined according to the size of the corresponding block and / or the intra prediction mode.
[0278] The quantized coefficients may be inversely quantized by the inverse quantization unit 220. The inverse quantization unit 220 may generate an inversely quantized coefficient by performing inverse quantization on the quantized coefficients. In addition, the inversely quantized coefficients may be inversely transformed by the inverse transform unit 230. The inverse transform unit 230 may generate a reconstructed residual block by performing inverse transformation on the inversely quantized coefficients. As a result of performing inverse quantization and inverse transformation on the quantized coefficients, a reconstructed residual block may be generated. Here, the inverse quantization unit 220 may apply a quantization matrix to the quantized coefficients when generating the reconstructed residual block.
[0279] When the intra mode is used, the intra prediction unit 240 may generate a predicted block by performing spatial prediction using pixel values of a previously decoded neighboring block adjacent to the target block with respect to the target block.
[0280] The inter prediction unit 250 may include a motion compensation unit. Alternatively, the inter prediction unit 250 may be designated as a "motion compensation unit."
[0281] When the inter mode is used, the motion compensation unit may generate a prediction block by performing motion compensation using a motion vector and a reference image stored in the reference picture buffer 270 with respect to the target block.
[0282] The motion compensation unit may apply an interpolation filter to a partial area of a reference image when a motion vector has a value other than an integer, and may generate a prediction block using the reference image to which the interpolation filter is applied. In order to perform motion compensation, the motion compensation unit may determine which of a skip mode, a merge mode, an advanced motion vector prediction (AMVP) mode, and a current picture reference mode corresponds to a motion compensation method for a PU included in the CU based on the CU, and may perform motion compensation according to the determined mode.
[0283] The reconstructed residual block and the prediction block may be added to each other by the adder 255. The adder 255 may generate a reconstructed block by adding the reconstructed residual block to the prediction block.
[0284] The reconstructed block may be filtered by the filter unit 260. The filter unit 260 may apply at least one of a deblocking filter, an SAO filter, an ALF, and an NLF to the reconstructed block or the reconstructed image. The reconstructed image may be a picture including the reconstructed block.
[0285] The filter unit may output a reconstructed image.
[0286] The reconstructed image and / or the reconstructed block filtered by the filter unit 260 may be stored as a reference picture in the reference picture buffer 270. The reconstructed block filtered by the filter unit 260 may be a part of the reference picture. In other words, the reference picture may be an image composed of the reconstructed blocks filtered by the filter unit 260. The stored reference picture may be subsequently used for inter-frame prediction or motion compensation.
[0287] Figure 3 is a diagram schematically showing a partition structure of an image when the image is encoded and decoded.
[0288] Figure 3 An example where a single unit is partitioned into a plurality of sub-units may be schematically shown.
[0289] In order to efficiently partition an image, coding units (CUs) may be used in encoding and decoding. The term "unit" may be used to collectively specify 1) a block containing image samples and 2) a syntax element. For example, "a partition of a unit" may mean "a partition of blocks corresponding to a unit".
[0290] CU can be used as a basic unit for image encoding / decoding. CU can be used as a unit to which one mode selected from intra mode and inter mode is applied in image encoding / decoding. In other words, in image encoding / decoding, it can be determined which one of the intra mode and inter mode is applied to each CU.
[0291] In addition, a CU may be a basic unit in prediction, transformation, quantization, inverse transformation, inverse quantization, and encoding / decoding of a transform coefficient.
[0292] Reference Figure 3 , the image 300 may be sequentially partitioned into units corresponding to a maximum coding unit (LCU), and a partition structure may be determined for each LCU. Here, LCU may be used to have the same meaning as a coding tree unit (CTU).
[0293] The partition of a unit may mean the partition of a block corresponding to the unit. The block partition information may include depth information about the depth of the unit. The depth information may indicate the number of times the unit is partitioned and / or the degree to which the unit is partitioned. A single unit may be hierarchically partitioned into a plurality of sub-units with depth information based on a tree structure.
[0294] Each of the sub-units of the partition may have depth information. The depth information may be information indicating the size of the CU. The depth information may be stored for each CU.
[0295] Each CU may have depth information. When a CU is partitioned, a CU generated by partitioning may have a depth increased by 1 from the depth of the partitioned CU.
[0296] The partition structure may mean the distribution of coding units (CUs) for efficiently encoding an image in the LCU 310. Such a distribution may be determined according to whether a single CU is to be partitioned into a plurality of CUs. The number of CUs generated by partitioning may be a positive integer of 2 or more, including 2, 3, 4, 8, 16, etc.
[0297] According to the number of CUs generated by partitioning, the horizontal size and vertical size of each of the CUs generated by partitioning may be smaller than the horizontal size and vertical size of the CU before partitioning. For example, the horizontal size and vertical size of each of the CUs generated by partitioning may be half of the horizontal size and vertical size of the CU before partitioning.
[0298] The CU of each partition may be recursively partitioned into four CUs in the same manner. Through recursive partitioning, at least one of the horizontal size and the vertical size of the CU of each partition may be reduced compared to at least one of the horizontal size and the vertical size of the CU before partitioning.
[0299] Partitioning of the CU may be performed recursively up to a predefined depth or a predefined size.
[0300] For example, the depth of the CU may have a value ranging from 0 to 3. According to the depth of the CU, the size of the CU may range from a size of 64×64 to a size of 8×8.
[0301] For example, the depth of the LCU 310 may be 0, and the depth of the minimum coding unit (SCU) may be a predefined maximum depth. Here, as described above, the LCU may be a CU having a maximum coding unit size, and the SCU may be a CU having a minimum coding unit size.
[0302] Partitioning may start from the LCU 310, and the depth of the CU may increase by one whenever the horizontal and / or vertical size of the CU is reduced by partitioning.
[0303] For example, for each depth, a non-partitioned CU may have a size of 2N×2N. In addition, in the case where the CU is partitioned, a CU of size 2N×2N may be partitioned into four CUs, each of which has a size of N×N. Whenever the depth increases by 1, the value of N may be halved.
[0304] Reference Figure 3 , an LCU with a depth of 0 may have 64×64 pixels or 64×64 blocks. 0 may be the minimum depth. An SCU with a depth of 3 may have 8×8 pixels or 8×8 blocks. 3 may be the maximum depth. Here, a CU with 64×64 blocks as an LCU may be represented by a depth of 0. A CU with 32×32 blocks may be represented by a depth of 1. A CU with 16×16 blocks may be represented by a depth of 2. A CU with 8×8 blocks as an SCU may be represented by a depth of 3.
[0305] Information about whether the corresponding CU is partitioned may be represented by partition information of the CU. The partition information may be 1-bit information. All CUs except the SCU may include partition information. For example, the value of the partition information of a non-partitioned CU may be a first value. The value of the partition information of a partitioned CU may be a second value. When the partition information indicates whether the CU is partitioned, the first value may be "0" and the second value may be "1".
[0306] For example, when a single CU is partitioned into four CUs, the horizontal size and vertical size of each of the four CUs generated by partitioning may be half of the horizontal size and vertical size of the CU before partitioning. When a CU with a size of 32×32 is partitioned into four CUs, the size of each of the four partitioned CUs may be 16×16. When a single CU is partitioned into four CUs, it can be considered that the CU has been partitioned in a quadtree structure. In other words, it can be considered that quadtree partitioning has been applied to the CU.
[0307] For example, when a single CU is partitioned into two CUs, the horizontal size or vertical size of each of the two CUs generated by partitioning may be half of the horizontal size or vertical size of the CU before partitioning. When a CU with a size of 32×32 is partitioned vertically into two CUs, the size of each of the two CUs after partitioning may be 16×32. When a CU with a size of 32×32 is partitioned horizontally into two CUs, the size of each of the two CUs after partitioning may be 32×16. When a single CU is partitioned into two CUs, it can be considered that the CU has been partitioned in a binary tree structure. In other words, it can be considered that binary tree partitioning has been applied to the CU.
[0308] For example, when a single CU is partitioned (or split) into three CUs, the original CU before partitioning is partitioned so that its horizontal size or vertical size is divided at a ratio of 1:2:1, thereby enabling the generation of three sub-CUs. For example, when a CU with a size of 16×32 is partitioned horizontally into three sub-CUs, the three sub-CUs generated by the partitioning may have sizes of 16×8, 16×16, and 16×8, respectively, in the direction from top to bottom. For example, when a CU with a size of 32×32 is partitioned vertically into three sub-CUs, the three sub-CUs generated by the partitioning may have sizes of 8×32, 16×32, and 8×32, respectively, in the direction from left to right. When a single CU is partitioned into three CUs, the CU may be considered to be partitioned in a ternary tree form. In other words, it may be considered that ternary tree partitioning has been applied to the CU.
[0309] Both quadtree partitioning and binary tree partitioning are applied to Figure 3 LCU 310.
[0310] In the encoding apparatus 100, a coding tree unit (CTU) having a size of 64×64 may be partitioned into a plurality of smaller CUs by a recursive quadtree structure. A single CU may be partitioned into four CUs having the same size. Each CU may be recursively partitioned and may have a quadtree structure.
[0311] Through recursive partitioning of CUs, the best partitioning method that results in the minimum rate-distortion penalty can be selected.
[0312] Figure 3 The coding tree unit (CTU) 320 in FIG. 1 is an example of a total CTU to which quadtree partitioning, binary tree partitioning, and ternary tree partitioning are applied.
[0313] As described above, in order to partition a CTU, at least one of quadtree partitioning, binary tree partitioning, and ternary tree partitioning may be applied to the CTU. The partitioning may be applied based on a specific priority.
[0314] For example, quadtree partitioning may be preferentially applied to CTUs. A CU that cannot be further partitioned in the form of a quadtree may correspond to a leaf node of a quadtree. A CU corresponding to a leaf node of a quadtree 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 quadtree may be partitioned in the form of a binary tree or a ternary tree, or may not be further partitioned. In this case, each CU generated by applying a binary tree partitioning or a ternary tree partitioning to a CU corresponding to a leaf node of a quadtree is prevented from being subjected to quadtree partitioning again, thereby effectively performing partitioning of a block and / or signaling of block partition information.
[0315] The four partition information may be used to signal the partition of the CU corresponding to each node of the quadtree. The four partition information having a first value (e.g., "1") may indicate that the corresponding CU is partitioned in a quadtree form. The four partition information having a second value (e.g., "0") may indicate that the corresponding CU is not partitioned in a quadtree form. The four partition information may be a flag having a specific length (e.g., 1 bit).
[0316] There may be no priority between binary tree partitioning and ternary tree partitioning. That is, the CU corresponding to the leaf node of the quadtree may be partitioned in a binary tree form or a ternary tree form. In addition, the CU generated by binary tree partitioning or ternary tree partitioning may be further partitioned in a binary tree form or a ternary tree form, or may not be further partitioned.
[0317] Partitioning performed when there is no priority between binary tree partitioning and ternary tree partitioning may be referred to as "multi-type tree partitioning". That is, the CU corresponding to the leaf node of the quadtree may be the root node of the multi-type tree. The partition of the CU corresponding to each node of the multi-type tree may be signaled using at least one of information indicating whether the CU is partitioned in a multi-type tree, partition direction information, and partition tree information. For the partition of the CU corresponding to each node of the multi-type tree, information indicating whether partitioning in a multi-type tree is performed, partition direction information, and partition tree information may be sequentially signaled.
[0318] For example, information indicating whether a CU is partitioned in a multi-type tree and having a first value (e.g., "1") may indicate that the corresponding CU is partitioned in a multi-type tree. Information indicating whether a CU is partitioned in a multi-type tree and having a second value (e.g., "0") may indicate that the corresponding CU is not partitioned in a multi-type tree.
[0319] When a CU corresponding to each node of the multi-type tree is partitioned in the multi-type tree form, the corresponding CU may further include partition direction information.
[0320] The partition direction information may indicate the partition direction of the multi-type tree partition. The partition direction information having a first value (e.g., "1") may indicate that the corresponding CU is partitioned in the vertical direction. The partition direction information having a second value (e.g., "0") may indicate that the corresponding CU is partitioned in the horizontal direction.
[0321] When a CU corresponding to each node of a multi-type tree is partitioned in a multi-type tree form, the corresponding CU may further include partition tree information. The partition tree information may indicate a tree used for multi-type tree partitioning.
[0322] For example, partition tree information having a first value (eg, "1") may indicate that the corresponding CU is partitioned in a binary tree form. Partition tree information having a second value (eg, "0") may indicate that the corresponding CU is partitioned in a ternary tree form.
[0323] Here, each of the above-mentioned information indicating whether partitioning with a multi-type tree is performed, the partition tree information, and the partition direction information may be a flag having a specific length (eg, 1 bit).
[0324] At least one of the above-mentioned four partition information, the information indicating whether to perform partitioning with a multi-type tree, the partition direction information, and the partition tree information may be entropy encoded and / or entropy decoded. In order to perform entropy encoding / decoding of such information, information of a neighboring CU adjacent to the target CU may be used.
[0325] For example, it can be considered that the partition form (i.e., partition / non-partition, partition tree and / or partition direction) of the left CU and / or the upper CU and the partition form of the target CU will be similar to each other with a high probability. Therefore, based on the information of the neighboring CU, context information for entropy encoding and / or entropy decoding of the information of the target CU can be derived. Here, the information of the neighboring CU may include at least one of the following items: 1) four partition information of the neighboring CU, 2) information indicating whether the neighboring CU is partitioned in a multi-type tree, 3) partition direction information of the neighboring CU, and 4) partition tree information of the neighboring CU.
[0326] In another embodiment, among the binary tree partitioning and the ternary tree partitioning, the binary tree partitioning may be performed preferentially. That is, the binary tree partitioning may be applied first, and then the CU corresponding to the leaf node of the binary tree may be set as the root node of the ternary tree. In this case, the quadtree partitioning or the binary tree partitioning may not be performed on the CU corresponding to the node of the ternary tree.
[0327] A CU that is no longer further partitioned by quadtree partitioning, binary tree partitioning, and / or ternary tree partitioning may be a unit for coding, prediction, and / or transformation. That is, the CU may not be further partitioned for prediction and / or transformation. Therefore, a partition structure for partitioning a CU into a prediction unit (PU) and / or a transformation unit (TU), its partition information, and the like may not be present in the bitstream.
[0328] However, when the size of the CU as a unit of partition is larger than the size of the maximum transform block, the CU may be recursively partitioned until the size of the CU becomes smaller than or equal to the size of the maximum transform block. For example, when the size of the CU is 64×64 and the size of the maximum transform block is 32×32, the CU may be partitioned into four 32×32 blocks in order to perform transform. For example, when the size of the CU is 32×64 and the size of the maximum transform block is 32×32, the CU may be partitioned into two 32×32 blocks.
[0329] In this case, information indicating whether the CU is partitioned for transformation may not be separately signaled. Without signaling, whether the CU is partitioned may be determined via a comparison between 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 larger than the horizontal size of the largest transform block, the CU may be bisected vertically. In addition, when the vertical size of the CU is larger than the vertical size of the largest transform block, the CU may be bisected horizontally.
[0330] Information about the maximum size and / or minimum size of the CU and information about the maximum size and / or minimum size of the transform block may be signaled or determined at a level higher than the level of the CU. For example, the higher level may be a sequence level, a picture level, a tile level, a tile group level, or a slice level. For example, the minimum size of the CU may be set to 4×4. For example, the maximum size of the transform block may be set to 64×64. For example, the maximum size of the transform block may be set to 4×4.
[0331] Information about the minimum size of the CU corresponding to the leaf node of the quadtree (i.e., the minimum size of the quadtree) and / or information about the maximum depth of the path from the root node of the multi-type tree to the leaf node (i.e., the maximum depth of the multi-type tree) may be signaled or determined at a higher level than the level of the CU corresponding to the leaf node of the quadtree. For example, the higher level may be a sequence level, a picture level, a slice level, a tile group level, or a tile level. Information about the minimum size of the quadtree and / or information about the maximum depth of the multi-type tree may be signaled or determined separately at each level in the intra-slice level and the inter-slice level.
[0332] Information about the difference between the size of the CTU and the maximum size of the transform block may be signaled or determined at a level higher than the level of the CU. For example, the higher level may be a sequence level, a picture level, a slice level, a parallel block group level, or a parallel block level. Information about the maximum size of the CU corresponding to each node of the binary tree (i.e., the maximum size of the binary tree) may be determined based on the size and difference information of the CTU. The maximum size of the CU corresponding to each node of the ternary tree (i.e., the maximum size of the ternary tree) may have different values depending on the type of the slice. For example, the maximum size of the ternary tree at the intra-slice level may be 32×32. For example, the maximum size of the ternary tree at the inter-slice level may be 128×128. For example, the minimum size of the CU corresponding to each node of the binary tree (i.e., the minimum size of the binary tree) and / or the minimum size of the CU corresponding to each node of the ternary tree (i.e., the minimum size of the ternary tree) may be set to the minimum size of the CU.
[0333] In another example, the maximum size of the binary tree and / or the maximum size of the ternary tree may be signaled or determined at the slice level. In addition, the minimum size of the binary tree and / or the minimum size of the ternary tree may be signaled or determined at the slice level.
[0334] Based on the above-mentioned various block sizes and depths, quad partition information, information indicating whether partitioning with a multi-type tree is performed, partition tree information, and / or partition direction information may or may not exist in a bitstream.
[0335] For example, when the size of the CU is not greater than the minimum size of the quadtree, the CU may not include quad partition information, and the quad partition information of the CU may be inferred as a second value.
[0336] For example, when the size (horizontal size and vertical size) of the CU corresponding to each node of the multi-type tree is greater than the maximum size (horizontal size and vertical size) of the binary tree and / or the maximum size (horizontal size and vertical size) of the ternary tree, the CU may not be partitioned in a binary tree form and / or a ternary tree form. With this determination, information indicating whether partitioning in a multi-type tree is performed may not be signaled but may be inferred as a second value.
[0337] Optionally, when the size (horizontal size and vertical size) of the CU corresponding to each node of the multi-type tree is equal to the minimum size (horizontal size and vertical size) of the binary tree, or when the size (horizontal size and vertical size) of the CU is equal to twice the minimum size (horizontal size and vertical size) of the ternary tree, the CU may not be partitioned in a binary tree form and / or a ternary tree form. With this determination method, information indicating whether partitioning in a multi-type tree is performed may not be transmitted by a signal, but may be inferred as a second value. The reason is that when the CU is partitioned in a binary tree form and / or a ternary tree form, a CU smaller than the minimum size of the binary tree and / or the minimum size of the ternary tree is generated.
[0338] Optionally, binary tree partitioning or ternary tree partitioning may be limited based on the size of a virtual pipeline data unit (ie, the size of a pipeline buffer). For example, when a CU is partitioned into sub-CUs that do not fit into the size of a pipeline buffer by binary tree partitioning or ternary tree partitioning, binary tree partitioning or ternary tree partitioning may be limited. The size of a pipeline buffer may be equal to the maximum size of a transform block (eg, 64×64).
[0339] For example, when the size of the pipeline buffer is 64×64, the following partitions may be restricted.
[0340] - Ternary tree partitioning of N×M CUs (where N and / or M is 128)
[0341] - 128×N CU horizontal binary tree partitioning (where N<=64)
[0342] -N×128 CU vertical binary tree partition (where N<=64)
[0343] Optionally, when the depth of the CU corresponding to each node of the multi-type tree is equal to the maximum depth of the multi-type tree, the CU may not be partitioned in a binary tree form and / or a ternary tree form. With this determination, information indicating whether partitioning in a multi-type tree is performed may not be transmitted by a signal, but may be inferred as a second value.
[0344] Alternatively, information indicating whether partitioning in a multi-type tree is performed may be transmitted by signal only when at least one of vertical binary tree partitioning, horizontal binary tree partitioning, vertical ternary tree partitioning, and horizontal ternary tree partitioning is possible for a CU corresponding to each node of a multi-type tree. Otherwise, the CU may not be partitioned in a binary tree form and / or a ternary tree form. With this determination, information indicating whether partitioning in a multi-type tree is performed may not be transmitted by signal, but may be inferred as a second value.
[0345] Optionally, for the CU corresponding to each node of the multi-type tree, partition direction information may be signaled only when both vertical binary tree partitioning and horizontal binary tree partitioning are possible or only when both vertical ternary tree partitioning and horizontal ternary tree partitioning are possible. Otherwise, the partition direction information may not be signaled but may be inferred as a value indicating the direction in which the CU may be partitioned.
[0346] Optionally, for the CU corresponding to each node of the multi-type tree, partition tree information may be signaled only when both vertical binary tree partitioning and vertical ternary tree partitioning are possible or only when both horizontal binary tree partitioning and horizontal ternary tree partitioning are possible. Otherwise, the partition tree information may not be signaled but may be inferred as a value indicating a tree of partitions that may be applied to the CU.
[0347] Figure 4 is a diagram illustrating a form of prediction units that a coding unit may include.
[0348] When a CU that is not further partitioned among CUs partitioned from an LCU can be divided into one or more prediction units (PUs), such division is also called 'partitioning'.
[0349] PU can be a basic unit for prediction. PU can be encoded and decoded in any one of skip mode, inter mode and intra mode. PU can be partitioned into various shapes according to each mode. For example, Figure 1 The target block described and the above reference Figure 2 The described target blocks may each be a PU.
[0350] A CU may not be partitioned into PUs. When a CU is not partitioned into PUs, the size of the CU and the size of the PU may be equal to each other.
[0351] In skip mode, partitions may not exist in a CU.In skip mode, a 2Nx2N mode 410 may be supported without partitions, wherein in the 2Nx2N mode 410, the sizes of the PU and the CU are the same as each other.
[0352] In inter mode, there may be eight types of partition shapes in a CU. For example, in inter mode, 2N×2N mode 410, 2N×N mode 415, N×2N mode 420, N×N mode 425, 2N×nU mode 430, 2N×nD mode 435, nL×2N mode 440, and nR×2N mode 445 may be supported.
[0353] In the intra mode, a 2N×2N mode 410 and an N×N mode 425 may be supported.
[0354] In 2N×2N mode 410, a PU having a 2N×2N size may be encoded. A PU having a 2N×2N size may mean a PU having the same size as that of a CU. For example, a PU having a 2N×2N size may have a size of 64×64, 32×32, 16×16, or 8×8.
[0355] In N×N mode 425 , a PU having an N×N size may be encoded.
[0356] For example, in intra prediction, when the size of the PU is 8×8, the PUs of four partitions may be encoded. The size of the PU of each partition may be 4×4.
[0357] When a PU is encoded in intra mode, any one of multiple intra prediction modes may be used to encode the PU. For example, HEVC technology may provide 35 intra prediction modes, and a PU may be encoded in any one of the 35 intra prediction modes.
[0358] Which of the 2Nx2N mode 410 and the NxN mode 425 is to be used to encode the PU may be determined based on the rate-distortion cost.
[0359] The encoding device 100 may perform an encoding operation on a PU of size 2N×2N. Here, the encoding operation may be an operation of encoding the PU according to each of a plurality of intra prediction modes that may be used by the encoding device 100. Through the encoding operation, an optimal intra prediction mode for a PU of size 2N×2N may be derived. The optimal intra prediction mode may be an intra prediction mode that incurs a minimum rate distortion cost when encoding a PU of size 2N×2N among a plurality of intra prediction modes that may be used by the encoding device 100.
[0360] In addition, the encoding device 100 may sequentially perform encoding operations on the respective PUs obtained from the N×N partitions. Here, the encoding operation may be an operation of encoding the PU according to each of a plurality of intra prediction modes that can be used by the encoding device 100. Through the encoding operation, an optimal intra prediction mode for a PU of size N×N may be derived. The optimal intra prediction mode may be an intra prediction mode that incurs a minimum rate distortion cost when encoding a PU of size N×N among a plurality of intra prediction modes that can be used by the encoding device 100.
[0361] The encoding device 100 may determine which one of the PU having a size of 2N×2N and the PU having a size of N×N is to be encoded based on a comparison of the rate-distortion cost of the PU having a size of 2N×2N and the rate-distortion cost of the PU having a size of N×N.
[0362] A single CU may be partitioned into one or more PUs, and a PU may be partitioned into multiple PUs.
[0363] For example, when a single PU is partitioned into four PUs, the horizontal size and vertical size of each of the four PUs generated by partitioning may be half the horizontal size and vertical size of the PU before partitioning. When a PU having a size of 32×32 is partitioned into four PUs, the size of each of the four partitioned PUs may be 16×16. When a single PU is partitioned into four PUs, it may be considered that the PU has been partitioned in a quadtree structure.
[0364] For example, when a single PU is partitioned into two PUs, the horizontal size or vertical size of each of the two PUs generated by partitioning may be half the horizontal size or vertical size of the PU before partitioning. When a PU with a size of 32×32 is partitioned vertically into two PUs, the size of each of the two partitioned PUs may be 16×32. When a PU with a size of 32×32 is partitioned horizontally into two PUs, the size of each of the two partitioned PUs may be 32×16. When a single PU is partitioned into two PUs, it may be considered that the PU has been partitioned in a binary tree structure.
[0365] Figure 5 is a diagram illustrating a form of a transformation unit that may be included in a coding unit.
[0366] A transform unit (TU) may have a basic unit for processes such as transformation, quantization, inverse transformation, inverse quantization, entropy encoding, and entropy decoding in a CU.
[0367] A TU may have a square shape or a rectangular shape. The shape of a TU may be determined based on the size and / or shape of a CU.
[0368] In a CU partitioned from an LCU, a CU that is not further partitioned into a CU may be partitioned into one or more TUs. Here, the partition structure of the TU may be a quadtree structure. For example, Figure 5 As shown, a single CU 510 can be partitioned once or multiple times according to a quadtree structure. With this partitioning, a single CU 510 can be composed of TUs of various sizes.
[0369] It can be considered that when a single CU is partitioned two or more times, the CU is recursively partitioned. Through partitioning, a single CU can be composed of transform units (TUs) of various sizes.
[0370] Alternatively, a single CU may be partitioned into one or more TUs based on the number of vertical and / or horizontal lines that partition the CU.
[0371] The CU may be partitioned into symmetric TUs or asymmetric TUs. In order to partition into asymmetric TUs, information about the size and / or shape of each TU may be sent from the encoding device 100 to the decoding device 200. Alternatively, the size and / or shape of each TU may be derived from the information about the size and / or shape of the CU.
[0372] A CU may not be partitioned into TUs. When a CU is not partitioned into TUs, the size of the CU and the size of the TU may be equal to each other.
[0373] A single CU may be partitioned into one or more TUs, and a TU may be partitioned into multiple TUs.
[0374] For example, when a single TU is partitioned into four TUs, the horizontal size and vertical size of each of the four TUs generated by partitioning may be half the horizontal size and vertical size of the TU before partitioning. When a TU having a size of 32×32 is partitioned into four TUs, the size of each of the four partitioned TUs may be 16×16. When a single TU is partitioned into four TUs, the TU may be considered to have been partitioned in a quadtree structure.
[0375] For example, when a single TU is partitioned into two TUs, the horizontal size or vertical size of each of the two TUs generated by partitioning may be half the horizontal size or vertical size of the TU before partitioning. When a TU with a size of 32×32 is partitioned vertically into two TUs, the size of each of the two partitioned TUs may be 16×32. When a TU with a size of 32×32 is partitioned horizontally into two TUs, the size of each of the two partitioned TUs may be 32×16. When a single TU is partitioned into two TUs, the TU may be considered to have been partitioned in a binary tree structure.
[0376] Can be different from Figure 5The CU is partitioned in the manner shown.
[0377] For example, a single CU may be partitioned into three CUs. The horizontal sizes or vertical sizes of the three CUs generated by the partitioning may be 1 / 4, 1 / 2, and 1 / 4 of the horizontal size or vertical size of the original CU before the partitioning, respectively.
[0378] For example, when a CU having a size of 32×32 is vertically partitioned into three CUs, the sizes of the three CUs generated from the partitioning may be 8×32, 16×32, and 8×32, respectively. In this way, when a single CU is partitioned into three CUs, it can be considered that the CU is partitioned in the form of a ternary tree.
[0379] One of the exemplary partitioning forms (i.e., quadtree partitioning, binary tree partitioning, and ternary tree partitioning) may be applied to the partitioning of the CU, and multiple partitioning schemes may be combined and used together for the partitioning of the CU. Here, the case where multiple partitioning schemes are combined and used together may be referred to as "complex tree format partitioning".
[0380] Figure 6 The segmentation of blocks according to an example is shown.
[0381] During video encoding and / or decoding, the target block can be segmented, such as Figure 6 For example, the target block may be a CU.
[0382] For the partitioning of the target block, an indicator indicating the partitioning information may be signaled from the encoding apparatus 100 to the decoding apparatus 200. The partitioning information may be information indicating how the target block is partitioned.
[0383] The split information may be one or more of a split flag (hereinafter referred to as "split_flag"), a quad flag (hereinafter referred to as "QB_flag"), a quadtree flag (hereinafter referred to as "quadtree_flag"), a binary tree flag (hereinafter referred to as "binarytree_flag") and a binary type flag (hereinafter referred to as "Btype_flag").
[0384] "split_flag" may be a flag indicating whether a block is split. For example, a split_flag value of 1 may indicate that the corresponding block is split. A split_flag value of 0 may indicate that the corresponding block is not split.
[0385] "QB_flag" may be a flag indicating which of the quadtree form and the binary tree form corresponds to the shape in which the block is partitioned. For example, a QB_flag value of 0 may indicate that the block is partitioned in a quadtree form. A QB_flag value of 1 may indicate that the block is partitioned in a binary tree form. Alternatively, a QB_flag value of 0 may indicate that the block is partitioned in a binary tree form. A QB_flag value of 1 may indicate that the block is partitioned in a quadtree form.
[0386] "quadtree_flag" may be a flag indicating whether the block is partitioned in a quadtree form. For example, a quadtree_flag value of 1 may indicate that the block is partitioned in a quadtree form. A quadtree_flag value of 0 may indicate that the block is not partitioned in a quadtree form.
[0387] "binarytree_flag" may be a flag indicating whether the block is split in a binary tree form. For example, a binarytree_flag value of 1 may indicate that the block is split in a binary tree form. A binarytree_flag value of 0 may indicate that the block is not split in a binary tree form.
[0388] "Btype_flag" may be a flag indicating which of the vertical partitioning and the horizontal partitioning corresponds to the partitioning direction when the block is partitioned in the form of a binary tree. For example, a Btype_flag value of 0 may indicate that the block is partitioned in the horizontal direction. A Btype_flag value of 1 may indicate that the block is partitioned in the vertical direction. Alternatively, a Btype_flag value of 0 may indicate that the block is partitioned in the vertical direction. A Btype_flag value of 1 may indicate that the block is partitioned in the horizontal direction.
[0389] For example, as shown in Table 1 below, it can be derived by signaling at least one of quadtree_flag, binarytree_flag, and Btype_flag. Figure 6 The segmentation information of the blocks in .
[0390] [Table 1]
[0391]
[0392] For example, as shown in Table 2 below, it can be derived by signaling at least one of split_flag, QB_flag, and Btype_flag. Figure 6 The segmentation information of the blocks in .
[0393] [Table 2]
[0394]
[0395] Depending on the size and / or shape of the block, the partitioning method may be limited to a quadtree or a binary tree. When this limitation is applied, split_flag may be a flag indicating whether the block is partitioned in a quadtree form or a flag indicating whether the block is partitioned in a binary tree form. The size and shape of the block may be derived from the depth information of the block, and the depth information may be transmitted from the encoding device 100 to the decoding device 200 in a signal.
[0396] When the size of the block falls within a certain range, only segmentation in the form of a quadtree may be possible. For example, the certain range may be defined by at least one of a maximum block size and a minimum block size, wherein only segmentation in the form of a quadtree is possible under the maximum block size and the minimum block size.
[0397] Information indicating a maximum block size and a minimum block size for which only quadtree partitioning can be performed may be signaled through a bitstream from the encoding apparatus 100 to the decoding apparatus 200. In addition, the information may be signaled for at least one of units such as a video, a sequence, a picture, a parameter, a tile group, and a slice (or segment).
[0398] Alternatively, the maximum block size and / or the minimum block size may be a fixed size predefined by the encoding device 100 and the decoding device 200. For example, when the size of the block is above 64×64 and below 256×256, only segmentation in the form of a quadtree is possible. In this case, split_flag may be a flag indicating whether segmentation in the form of a quadtree is performed.
[0399] When the size of the block is larger than the maximum size of the transform block, only partitioning in a quadtree form is possible. Here, the sub-block generated by the partitioning may be at least one of a CU and a TU.
[0400] In this case, split_flag may be a flag indicating whether the CU is partitioned in a quadtree form.
[0401] When the size of the block falls within a specific range, segmentation may be possible only in the form of a binary tree or a ternary tree. For example, the specific range may be defined by at least one of a maximum block size and a minimum block size, wherein segmentation may be possible only in the form of a binary tree or a ternary tree under the maximum block size and the minimum block size.
[0402] Information indicating a maximum block size and / or a minimum block size that can be partitioned only in a binary tree form or in a ternary tree form may be signaled through a bitstream from the encoding apparatus 100 to the decoding apparatus 200. In addition, the information may be signaled for at least one of units such as a sequence, a picture, and a slice (or segment).
[0403] Alternatively, the maximum block size and / or the minimum block size may be a fixed size predefined by the encoding device 100 and the decoding device 200. For example, when the size of the block is above 8×8 and below 16×16, only splitting in a binary tree form may be possible. In this case, split_flag may be a flag indicating whether to perform splitting in a binary tree form or a ternary tree form.
[0404] The above description about partitioning in the form of a quadtree may be equally applied to the form of a binary tree and / or a ternary tree.
[0405] The partitioning of a block may be limited by a previous partitioning. For example, when a block is partitioned in a specific binary tree form and then a plurality of sub-blocks are generated from the partitioning, each sub-block may be partitioned only in a specific tree form. Here, the specific tree form may be at least one of a binary tree form, a ternary tree form, and a quadtree form.
[0406] When the horizontal size or the vertical size of the partition block is a size that cannot be further divided, the above indicator may not be signaled.
[0407] Figure 7 is a diagram for explaining an embodiment of an intra prediction process.
[0408] Figure 7 The arrows extending radially from the center of the graph in indicate the prediction direction of the intra prediction mode. In addition, the numbers appearing near the arrows indicate examples of mode values assigned to the intra prediction mode or the prediction direction of the intra prediction mode.
[0409] exist Figure 7 In FIG. 1 , the number 0 may represent the planar mode as a non-directional intra prediction mode, and the number 1 may represent the DC mode as a non-directional intra prediction mode.
[0410] Intra-frame encoding and / or decoding may be performed using reference samples of neighboring blocks of the target block. The neighboring blocks may be reconstructed neighboring blocks. The reference samples may refer to neighboring samples.
[0411] For example, intra encoding and / or decoding may be performed using a value of a reference sample included in a reconstructed neighboring block or an encoding parameter of the reconstructed neighboring block.
[0412] The encoding device 100 and / or the decoding device 200 may generate a prediction block by performing intra prediction on the target block based on information about samples in the target image. When performing intra prediction, the encoding device 100 and / or the decoding device 200 may generate a prediction block of the target block by performing intra prediction based on information about samples in the target image. When performing intra prediction, 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.
[0413] The prediction block may be a block generated as a result of performing intra prediction. The prediction block may correspond to at least one of a CU, a PU, and a TU.
[0414] The unit of the prediction block may have a size corresponding to at least one of a CU, a PU, and a TU. The prediction block may have a square shape having a size of 2N×2N or N×N. The size of N×N may include a size of 4×4, 8×8, 16×16, 32×32, 64×64, etc.
[0415] Alternatively, the prediction block may be a square block having a size of 2×2, 4×4, 8×8, 16×16, 32×32, 64×64, etc. or a rectangular block having a size of 2×8, 4×8, 2×16, 4×16, 8×16, etc.
[0416] Intra-prediction may be performed considering the intra-prediction mode of the target block. The number of intra-prediction modes that the target block may have may be a predefined fixed value, and may be a value determined differently according to the properties of the prediction block. For example, the properties of the prediction block may include the size of the prediction block, the type of the prediction block, etc. In addition, the properties of the prediction block may indicate encoding parameters of the prediction block.
[0417] For example, the number of intra prediction modes may be fixed to N regardless of the size of the prediction block. Alternatively, the number of intra prediction modes may be 3, 5, 9, 17, 34, 35, 36, 65, 67, or 95, for example.
[0418] The intra prediction mode can be a non-directional mode or a directional mode.
[0419] For example, intra prediction modes may include Figure 7 The numbers 0 to 66 shown in correspond to the two non-directional modes and the 65 directional modes.
[0420] For example, in the case of using a specific intra prediction method, the intra prediction mode may include Figure 7 The numbers -14 to 80 shown in the figure correspond to two non-directional modes and 93 directional modes.
[0421] The two non-directional modes may include a DC mode and a planar mode.
[0422] The directional mode may be a prediction mode with a specific direction or a specific angle. The directional mode may also be referred to as an "angle mode".
[0423] The intra prediction mode may be represented by at least one of a mode number, a mode value, a mode angle, and a mode direction. In other words, the terms "(mode) number of the intra prediction mode", "(mode) value of the intra prediction mode", "(mode) angle of the intra prediction mode", and "(mode) direction of the intra prediction mode" may be used to have the same meaning and may be used interchangeably.
[0424] The number of intra prediction modes may be M. The value of M may be 1 or more. In other words, the number of intra prediction modes may be M, which includes the number of non-directional modes and the number of directional modes.
[0425] The number of intra prediction modes may be fixed to M regardless of the size of the block and / or the color components. For example, the number of intra prediction modes may be fixed to any one of 35 and 67 regardless of the size of the block.
[0426] Alternatively, the number of intra prediction modes may differ according to the shape, size and / or type of color component of the block.
[0427] For example, in Figure 7 In FIG. 5 , the directional prediction mode indicated by the dotted line can be applied only to the prediction of non-square blocks.
[0428] For example, the larger the size of the block, the larger the number of intra prediction modes. Alternatively, the larger the size of the block, the smaller the number of intra prediction modes. When the size of the block is 4×4 or 8×8, the number of intra prediction modes may be 67. When the size of the block is 16×16, the number of intra prediction modes may be 35. When the size of the block is 32×32, the number of intra prediction modes may be 19. When the size of the block is 64×64, the number of intra prediction modes may be 7.
[0429] For example, the number of intra prediction modes may differ depending on whether the color component is a luminance signal or a chrominance signal. Alternatively, the number of intra prediction modes corresponding to a luminance component block may be greater than the number of intra prediction modes corresponding to a chrominance component block.
[0430] For example, in a vertical mode having a mode value of 50, prediction may be performed in a vertical direction based on pixel values of reference samples. For example, in a horizontal mode having a mode value of 18, prediction may be performed in a horizontal direction based on pixel values of reference samples.
[0431] Even in a directional mode other than the above-described modes, the encoding apparatus 100 and the decoding apparatus 200 may perform intra prediction on a target unit using a reference sample according to an angle corresponding to the directional mode.
[0432] An intra-frame prediction mode located to the right relative to the vertical mode may be referred to as a "vertical-right mode". An intra-frame prediction mode located below the horizontal mode may be referred to as a "horizontal-below mode". For example, in Figure 7 , the intra prediction mode having one of the mode values 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, and 66 may be the vertical-right mode. The intra prediction mode having one of the mode values 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, and 17 may be the horizontal-down mode.
[0433] The non-directional mode may include a DC mode and a planar mode. For example, the value of the DC mode may be 1. The value of the planar mode may be 0.
[0434] The directional mode may include an angular mode. Among the plurality of intra prediction modes, the remaining modes except the DC mode and the planar mode may be directional modes.
[0435] When the intra prediction mode is the DC mode, the prediction block may be generated based on the average value of the pixel values of the plurality of reference pixels. For example, the value of the pixel of the prediction block may be determined based on the average value of the pixel values of the plurality of reference pixels.
[0436] The number of intra prediction modes and the mode values of each intra prediction mode are merely exemplary and may be defined differently according to embodiments, implementations and / or requirements.
[0437] In order to perform intra prediction on a target block, a step of checking whether a sample included in a reconstructed neighboring block can be used as a reference sample of the target block may be performed. When there is a sample that cannot be used as a reference sample of the target block among the samples in the neighboring block, a value generated by copying and / or interpolating at least one sample value among the samples included in the reconstructed neighboring block may replace the sample value of the sample that cannot be used as a reference sample. When the value generated by copying and / or interpolating replaces the sample value of the existing sample, the sample may be used as a reference sample of the target block.
[0438] When intra prediction is used, a filter may be applied to at least one of a reference sample and a prediction sample based on at least one of an intra prediction mode and a size of a target block.
[0439] The type of filter to be applied to at least one of the reference sample and the prediction sample may be different according to at least one of the intra prediction mode of the target block, the size of the target block, and the shape of the target block. The type of filter may be classified according to one or more of the length of the filter tap, the value of the filter coefficient, and the filter strength. The length of the filter tap may mean the number of the filter tap. In addition, the number of the filter tap may mean the length of the filter.
[0440] When the intra prediction mode is the planar mode, the sample value of the predicted target block can be generated according to the position of the predicted target sample in the prediction block when generating the prediction block of the predicted target block, using the weighted sum of the upper reference sample of the target block, the left reference sample of the target block, the upper right reference sample of the target block, and the lower left reference sample of the target block.
[0441] When the intra prediction mode is the DC mode, when generating the prediction block of the target block, the average value of the reference sample above the target block and the reference sample on the left side of the target block may be used. Furthermore, filtering using the value of the reference sample may be performed on a specific row or a specific column in the target block. The specific row may be one or more upper rows adjacent to the reference sample. The specific column may be one or more left columns adjacent to the reference sample.
[0442] When the intra prediction mode is a directional mode, the 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.
[0443] In order to generate the above-mentioned prediction samples, real number-based interpolation may be performed.
[0444] The intra prediction mode of the target block may be predicted from intra prediction modes of neighboring blocks adjacent to the target block, and information used for the prediction may be entropy encoded / decoded.
[0445] For example, when the intra prediction modes of the target block and the neighboring block are identical to each other, a predefined flag may be used to signal that the intra prediction modes of the target block and the neighboring block are identical.
[0446] For example, an indicator indicating an intra prediction mode that is the same as the intra prediction mode of the target block among the intra prediction modes of a plurality of neighboring blocks may be signaled.
[0447] When intra prediction modes of a target block and a neighboring block are different from each other, information about the intra prediction mode of the target block may be encoded and / or decoded using entropy encoding and / or decoding.
[0448] Figure 8 is a diagram showing reference samples used in the intra prediction process.
[0449] Reconstructed reference samples used for intra prediction of the target block may include a lower left reference sample, a left reference sample, an upper left reference sample, an upper reference sample, and an upper right reference sample.
[0450] For example, a left reference sample may mean a reconstructed reference pixel adjacent to the left side of a target block. An upper reference sample may mean a reconstructed reference pixel adjacent to the top of a target block. An upper left corner reference sample may mean a reconstructed reference pixel located at the upper left corner of a target block. A lower left reference sample may mean a reference sample located below a left sample line among samples located on the same line as a left sample line composed of left reference samples. An upper right reference sample may mean a reference sample located on the right side of an upper sample line among samples located on the same line as an upper sample line composed of upper reference samples.
[0451] When the size of the target block is N×N, the number of lower-left reference samples, left-side reference samples, upper reference samples, and upper-right reference samples may each be N.
[0452] By performing intra prediction on the target block, a prediction block may be generated. Generating the prediction block may include determining values of pixels in the prediction block. The target block and the prediction block may be of equal size.
[0453] The reference sample used for intra prediction of the target block may vary according to the intra prediction mode of the target block. The direction of the intra prediction mode may represent the dependency between the reference sample and the pixels of the prediction block. For example, the value of a specified reference sample may be used as the value of one or more specified pixels in the prediction block. In this case, the specified reference sample and one or more specified pixels in the prediction block may be samples and pixels located on a straight line in the direction of the intra prediction mode. In other words, the value of the specified reference sample may be copied as the value of a pixel located in a direction opposite to the direction of the intra prediction mode. Alternatively, the value of a pixel in the prediction block may be the value of a reference sample located in the direction of the intra prediction mode relative to the position of the pixel.
[0454] In an example, when the intra prediction mode of the target block is a vertical mode, the upper reference sample can be used for intra prediction. When the intra prediction mode is a vertical mode, the value of a pixel in the prediction block can be the value of a reference sample located vertically above the position of the pixel. Therefore, the upper reference sample adjacent to the top of the target block can be used for intra prediction. In addition, the value of a pixel in a row of the prediction block can be the same as the value of the above reference sample.
[0455] In an example, when the intra prediction mode of the target block is a horizontal mode, the left reference sample can be used for intra prediction. When the intra prediction mode is a horizontal mode, the value of a pixel in the prediction block can be the value of a reference sample horizontally located to the left of the position of the pixel. Therefore, the left reference sample adjacent to the left side of the target block can be used for intra prediction. In addition, the value of a pixel in a column of the prediction block can be the same as the value of the left reference sample.
[0456] In an example, when the mode value of the intra prediction mode of the current block is 34, at least some of the left reference samples, the upper left reference samples, and at least some of the upper reference samples may be used for intra prediction. When the mode value of the intra prediction mode is 34, the value of a pixel in the prediction block may be the value of a reference sample diagonally located at the upper left corner of the pixel.
[0457] Also, in the case of an intra prediction mode whose mode value is a value within a range of 52 to 66, at least a portion of an upper right reference sample may be used for intra prediction.
[0458] Furthermore, in the case of an intra prediction mode whose mode value is a value within a range of 2 to 17, at least a portion of the lower left reference sample may be used for intra prediction.
[0459] Also, in the case of an intra prediction mode whose mode value is a value within a range of 19 to 49, a top left corner reference sample may be used for intra prediction.
[0460] The number of reference samples used to determine the pixel value of a pixel in a prediction block may be 1 or 2 or more.
[0461] As described above, the pixel value of the pixel in the prediction block can be determined according to the position of the pixel and the position of the reference sample indicated by the direction of the intra-frame prediction mode. When the position of the pixel and the position of the reference sample indicated by the direction of the intra-frame prediction mode are integer positions, the value of one reference sample indicated by the integer position can be used to determine the pixel value of the pixel in the prediction block.
[0462] When the position of a pixel and the position of a reference sample indicated by the direction of an intra-frame prediction mode are not integer positions, an interpolated reference sample based on two reference samples closest to the position of the reference sample may be generated. The value of the interpolated reference sample may be used to determine the pixel value of a pixel in a prediction block. In other words, when the position of a pixel in a prediction block and the position of a reference sample indicated by the direction of an intra-frame prediction mode indicate a position between two reference samples, an interpolated value based on the values of the two samples may be generated.
[0463] The prediction block generated via prediction may not be the same as the original target block. In other words, there may be a prediction error as the difference between the target block and the prediction block, and there may also be a prediction error between the pixels of the target block and the pixels of the prediction block.
[0464] Hereinafter, the terms “difference”, “error” and “residual” may be used to have the same meaning and may be used interchangeably with each other.
[0465] For example, in the case of directional intra prediction, the longer the distance between the pixels of the prediction block and the reference samples, the greater the prediction error that may occur. Such a prediction error may lead to discontinuity between the generated prediction block and the neighboring blocks.
[0466] In order to reduce the prediction error, filtering for the prediction block can be used. The filtering can be configured to adaptively apply the filter to the area in the prediction block that is considered to have a large prediction error. For example, the area that is considered to have a large prediction error can be the boundary of the prediction block. In addition, the area in the prediction block that is considered to have a large prediction error can be different according to the intra-frame prediction mode, and the characteristics of the filter can also be different according to the intra-frame prediction mode.
[0467] like Figure 8 As shown, for intra prediction of the target block, at least one of reference lines 0 to 3 may be used.
[0468] Figure 8 Each reference line in may indicate a reference sample line including one or more reference samples. When the number of the reference line is smaller, it may indicate that the line of the reference sample is closer to the target block.
[0469] The samples in segments A and F can be obtained by filling in using samples in segments B and E that are closest to the target block, rather than from reconstructed neighboring blocks.
[0470] Index information indicating a reference sample line to be used for intra prediction of a target block may be transmitted by a signal. The index information may indicate a reference sample line to be used for intra prediction of a target block among a plurality of reference sample lines. For example, the index information may have a value corresponding to any one of 0 to 3.
[0471] When the top boundary of the target block is the boundary of the CTU, only the reference sample line 0 may be available. Therefore, in this case, the index information may not be signaled. When an additional reference sample line other than the reference sample line 0 is used, filtering of the prediction block to be described later may not be performed.
[0472] In case of inter-color intra prediction, a prediction block of a target block of a second color component may be generated based on a corresponding reconstructed block of a first color component.
[0473] For example, the first color component may be a luma component and the second color component may be a chroma component.
[0474] To perform inter-color intra prediction, parameters for a linear model between a first color component and a second color component may be derived based on the template.
[0475] The template may include a reference sample above the target block (upper reference sample) and / or a reference sample to the left of the target block (left reference sample), and may include an upper reference sample and / or a left reference sample of the reconstructed block of the first color component corresponding to the reference samples.
[0476] For example, the parameters of the linear model can be derived using the following items: 1) the value of the sample with the maximum value of the first color component among the samples in the template, 2) the value of the sample of the second color component corresponding to the sample of the first color component, 3) the value of the sample with the minimum value of the first color component 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.
[0477] When the parameters of the linear model are derived, a prediction block of the target block may be generated by applying the corresponding reconstructed block to the linear model.
[0478] Depending on the image format, subsampling may be performed on samples adjacent to a reconstructed block of a first color component and a corresponding reconstructed block of the first color component. For example, when one sample of a second color component corresponds to four samples of a first color component, one corresponding sample may be calculated by performing subsampling on four samples of the first color component. When subsampling is performed, derivation of parameters of a linear model and inter-color intra prediction may be performed based on the corresponding samples of the subsampling.
[0479] Information on whether to perform inter-color intra prediction and / or the range of a template may be signaled in an intra prediction mode.
[0480] The target block may be partitioned into two or four sub-blocks in a horizontal direction and / or a vertical direction.
[0481] The sub-blocks generated by partitioning can be sequentially reconstructed. That is, when intra prediction is performed on each sub-block, a sub-prediction block of the sub-block can be generated. In addition, when inverse quantization (inverse quantization) and / or inverse transformation is performed on each sub-block, a sub-residual block of the corresponding sub-block can be generated. The reconstructed sub-block can be generated by adding the sub-prediction block to the sub-residual block. The reconstructed sub-block can be used as a reference sample for intra prediction of a sub-block with the next priority.
[0482] A subblock may be a block including a certain number of samples (e.g., 16) or more. For example, when the target block is an 8×4 block or a 4×8 block, the target block may be partitioned into two subblocks. In addition, when the target block is a 4×4 block, the target block cannot be partitioned into subblocks. When the target block has another size, the target block may be partitioned into four subblocks.
[0483] Information about whether to perform intra prediction based on such a subblock and / or information about a partition direction (horizontal direction or vertical direction) may be signaled.
[0484] Such subblock-based intra prediction may be restricted such that it is performed only when using reference sample line 0. When subblock-based intra prediction is performed, filtering of a prediction block to be described below may not be performed.
[0485] A final prediction block may be generated by performing filtering on a prediction block generated through intra prediction.
[0486] The filtering may be performed by applying a specific weight to a filtering target sample, a left reference sample, an upper reference sample, and / or an upper-left reference sample, wherein the filtering target sample is a target to be filtered.
[0487] The weight and / or reference sample (eg, range of reference samples, position of reference samples, etc.) used for filtering may be determined based on at least one of a block size, an intra prediction mode, and a position of a filtering target sample in a prediction block.
[0488] For example, filtering may be performed only in specific intra prediction modes (eg, DC mode, planar mode, vertical mode, horizontal mode, diagonal mode, and / or adjacent diagonal mode).
[0489] The adjacent diagonal pattern may be a pattern having a number obtained by adding k to the number of the diagonal pattern, and may be a pattern having a number obtained by subtracting k from the number of the diagonal pattern. In other words, the number of the adjacent diagonal pattern may be the sum of the number of the diagonal pattern and k, or may be the difference between the number of the diagonal pattern and k. For example, k may be a positive integer of 8 or less.
[0490] The intra prediction mode of the target block may be derived using intra prediction modes of neighboring blocks existing near the target block, and the intra prediction mode thus derived may be entropy encoded and / or entropy decoded.
[0491] For example, when the intra prediction mode of the target block is the same as that of the neighboring block, information indicating that the intra prediction mode of the target block is the same as that of the neighboring block may be signaled using specific flag information.
[0492] In addition, for example, indicator information of a neighboring block having the same intra prediction mode as the intra prediction mode of the target block among the intra prediction modes of the plurality of neighboring blocks may be signaled.
[0493] For example, when the intra prediction mode of the target block is different from the intra prediction mode of the neighboring block, information about the intra prediction mode of the target block may be entropy encoded and / or decoded by performing entropy encoding and / or entropy decoding based on the intra prediction mode of the neighboring block.
[0494] Fig. 9 is a diagram for explaining an embodiment of an inter-frame prediction process.
[0495] Fig. 9 The rectangle shown in can represent an image (or picture). Fig. 9 In the example, an arrow may indicate a prediction direction. An arrow pointing from a first picture to a second picture means that the second picture refers to the first picture. That is, each picture may be encoded and / or decoded according to the prediction direction.
[0496] According to the encoding type, images can be classified into intra-frame pictures (I pictures), unidirectional prediction pictures or prediction coded pictures (P pictures), and bidirectional prediction pictures or bidirectional prediction coded pictures (B pictures). Each picture can be encoded and / or decoded according to its encoding type.
[0497] When the target image to be encoded is an I picture, the target image can be encoded using data contained in the image itself without referring to inter-frame prediction of other images. For example, the I picture can be encoded only via intra-frame prediction.
[0498] When the target image is a P picture, the target image may be encoded via inter prediction using a reference picture existing in one direction. Here, the one direction may be a forward direction or a backward direction.
[0499] When the target image is a B picture, the image may be encoded via inter-frame prediction using reference pictures existing in both directions, or may be encoded via inter-frame prediction using reference pictures existing in one of the forward direction and the backward direction. Here, the two directions may be the forward direction and the backward direction.
[0500] A P picture and a B picture encoded and / or decoded using a reference picture may be regarded as an image in which inter-frame prediction is used.
[0501] Hereinafter, inter prediction in the inter mode according to an embodiment will be described in detail.
[0502] Inter-frame prediction or motion compensation may be performed using a reference image and motion information.
[0503] In the inter mode, the encoding apparatus 100 may perform inter prediction and / or motion compensation on the target block. The decoding apparatus 200 may perform inter prediction and / or motion compensation corresponding to the inter prediction and / or motion compensation performed by the encoding apparatus 100 on the target block.
[0504] During inter prediction, motion information of a target block may be separately derived by the encoding apparatus 100 and the decoding apparatus 200. The motion information may be derived using motion information of a reconstructed neighboring block, motion information of a col block, and / or motion information of a block adjacent to the col block.
[0505] For example, the encoding apparatus 100 or the decoding apparatus 200 may perform prediction and / or motion compensation by using motion information of the spatial candidate and / or the temporal candidate as motion information of the target block. The target block may mean a PU and / or a PU partition.
[0506] The spatial candidates may be reconstructed blocks that are spatially adjacent to the target block.
[0507] The temporal candidate may be a reconstructed block in a previously reconstructed co-located picture (col picture) corresponding to the target block.
[0508] In inter-frame prediction, the encoding device 100 and the decoding device 200 can improve encoding efficiency and decoding efficiency by using motion information of spatial candidates and / or temporal candidates. The motion information of the spatial candidate may be referred to as "spatial motion information". The motion information of the temporal candidate may be referred to as "temporal motion information".
[0509] Hereinafter, the motion information of the spatial candidate may be the motion information of the PU including the spatial candidate. The motion information of the temporal candidate may be the motion information of the PU including the temporal candidate. The motion information of the candidate block may be the motion information of the PU including the candidate block.
[0510] Inter prediction may be performed using reference pictures.
[0511] The reference picture may be at least one of a picture before the target picture and a picture after the target picture. The reference picture may be an image used for prediction of the target block.
[0512] In inter prediction, a region in a reference picture may be specified by using a reference picture index (or refIdx) indicating a reference picture, a motion vector (to be described later), etc. Here, the region specified in the reference picture may indicate a reference block.
[0513] Inter prediction can select a reference picture, and can also select a reference block corresponding to the target block from the reference picture. In addition, inter prediction can use the selected reference block to generate a prediction block of the target block.
[0514] Motion information may be derived by each of the encoding apparatus 100 and the decoding apparatus 200 during inter prediction.
[0515] A spatial candidate may be a block that satisfies the following items: 1) exists in the target picture, 2) has been previously reconstructed by encoding and / or decoding, and 3) is adjacent to the target block or is located at the corner of the target block. Here, a "block located at the corner of the target block" may be a block that is adjacent to the target block in the horizontal direction in the vertical direction, or a block that is adjacent to the target block in the vertical direction in the horizontal direction. In addition, a "block located at the corner of the target block" may have the same meaning as a "block adjacent to the corner of the target block". The meaning of a "block located at the corner of the target block" may be included in the meaning of a "block adjacent to the target block".
[0516] For example, the spatial candidate may be a reconstructed block located to the left of the target block, a reconstructed block located above the target block, a reconstructed block located at the lower left corner of the target block, a reconstructed block located at the upper right corner of the target block, or a reconstructed block located at the upper left corner of the target block.
[0517] Each of the encoding apparatus 100 and the decoding apparatus 200 may identify a block existing at a position spatially corresponding to the target block in the col picture. The position of the target block in the target picture and the position of the identified block in the col picture may correspond to each other.
[0518] Each of the encoding device 100 and the decoding device 200 may determine the col block existing at a predefined relative position of the identified block as a temporal candidate. The predefined relative position may be a position existing inside and / or outside the identified block.
[0519] For example, the col block may include a first col block and a second col block. When the coordinates of the identified block are (xP, yP) and the size of the identified block is represented by (nPSW, nPSH), the first col block may be a block located at the coordinates (xP+nPSW, yP+nPSH). The second col block may be a block located at the coordinates (xP+(nPSW>>1), yP+(nPSH>>1)). When the first col block is not available, the second col block may be selectively used.
[0520] The motion vector of the target block may be determined based on the motion vector of the col block. Each of the encoding device 100 and the decoding device 200 may scale the motion vector of the col block. The scaled motion vector of the col block may be used as the motion vector of the target block. In addition, the motion vector of the motion information of the temporal candidate stored in the list may be a scaled motion vector.
[0521] The ratio of the motion vector of the target block to the motion vector of the col block may be the same as the ratio of the first temporal distance to the second temporal distance. The first temporal distance may be the distance between the reference picture and the target picture of the target block. The second temporal distance may be the distance between the reference picture and the col picture of the col block.
[0522] The scheme for deriving motion information may change according to the inter prediction mode of the target block. For example, as an inter prediction mode applied to inter prediction, there may be an advanced motion vector prediction (AMVP) mode, a merge mode, a skip mode, a merge mode with a motion vector difference, a sub-block merge mode, a triangle partition mode, an inter-intra combined prediction mode, an affine inter mode, a current picture reference mode, etc. The merge mode may also be referred to as a "motion merge mode". Below, each mode is described in detail.
[0523] 1) AMVP model
[0524] When the AMVP mode is used, the encoding device 100 may search for a similar block in a neighboring area of the target block. The encoding device 100 may obtain a prediction block by performing prediction on the target block using motion information of the found similar block. The encoding device 100 may encode a residual block, which is the difference between the target block and the prediction block.
[0525] 1-1) Creation of a list of motion vector prediction candidates
[0526] When the AMVP mode is used as the prediction mode, each of the encoding device 100 and the decoding device 200 may create a prediction motion vector candidate list using a motion vector of a spatial candidate, a motion vector of a temporal candidate, and a zero vector. The prediction motion vector candidate list may include one or more prediction motion vector candidates. At least one of the motion vector of the spatial candidate, the motion vector of the temporal candidate, and the zero vector may be determined and used as a prediction motion vector candidate.
[0527] Hereinafter, the terms "prediction motion vector (candidate)" and "motion vector (candidate)" may be used to have the same meaning and may be used interchangeably with each other.
[0528] Hereinafter, the terms “prediction motion vector candidate” and “AMVP candidate” may be used to have the same meaning and may be used interchangeably with each other.
[0529] Hereinafter, the terms “prediction motion vector candidate list” and “AMVP candidate list” may be used to have the same meaning and may be used interchangeably with each other.
[0530] The spatial candidate may include a reconstructed spatial neighboring block. In other words, the motion vector of the reconstructed neighboring block may be referred to as a "spatial prediction motion vector candidate".
[0531] The temporal candidates may include the col block and blocks adjacent to the col block. In other words, the motion vector of the col block or the motion vector of the block adjacent to the col block may be referred to as a "temporal prediction motion vector candidate".
[0532] The zero vector may be the (0,0) motion vector.
[0533] The predicted motion vector candidate may be a motion vector predictor for predicting a motion vector. In addition, in the encoding apparatus 100, each predicted motion vector candidate may be an initial search position of a motion vector.
[0534] 1-2) Searching for motion vector using the predicted motion vector candidate list
[0535] The encoding device 100 may determine a motion vector to be used for encoding a target block within a search range using the predicted motion vector candidate list. In addition, the encoding device 100 may determine a predicted motion vector candidate to be used as a predicted motion vector of a target block among the predicted motion vector candidates present in the predicted motion vector candidate list.
[0536] The motion vector to be used for encoding the target block may be a motion vector that can be encoded at a minimum cost.
[0537] Also, the encoding apparatus 100 may determine whether to encode the target block using the AMVP mode.
[0538] 1-3) Transmission of inter-frame prediction information
[0539] The encoding apparatus 100 may generate a bitstream including inter prediction information required for inter prediction, and the decoding apparatus 200 may perform inter prediction on a target block using the inter prediction information of the bitstream.
[0540] The inter prediction information may include: 1) mode information indicating whether the AMVP mode is used, 2) a prediction motion vector index, 3) a motion vector difference (MVD), 4) a reference direction, and 5) a reference picture index.
[0541] Hereinafter, the terms 'prediction motion vector index' and 'AMVP index' may be used to have the same meaning and may be used interchangeably with each other.
[0542] In addition, the inter-frame prediction information may include a residual signal.
[0543] When the mode information indicates that the AMVP mode is used, the decoding apparatus 200 may acquire a prediction motion vector index, an MVD, a reference direction, and a reference picture index from a bitstream through entropy decoding.
[0544] The predicted motion vector index may indicate a predicted motion vector candidate to be used for prediction of the target block among the predicted motion vector candidates included in the predicted motion vector candidate list.
[0545] 1-4) Inter-frame prediction in AMVP mode using inter-frame prediction information
[0546] The decoding apparatus 200 may derive a prediction motion vector candidate using the prediction motion vector candidate list, and may determine motion information of a target block based on the derived prediction motion vector candidate.
[0547] The decoding device 200 may determine a motion vector candidate for the target block among the predicted motion vector candidates included in the predicted motion vector candidate list using the predicted motion vector index. The decoding device 200 may select a predicted motion vector candidate indicated by the predicted motion vector index from among the predicted motion vector candidates included in the predicted motion vector candidate list as the predicted motion vector of the target block.
[0548] The encoding device 100 may generate an entropy-coded predicted motion vector index by applying entropy encoding to the predicted motion vector index, and may generate a bitstream including the entropy-coded predicted motion vector index. The entropy-coded predicted motion vector index may be signaled from the encoding device 100 to the decoding device 200 through the bitstream. The decoding device 200 may extract the entropy-coded predicted motion vector index from the bitstream, and may obtain the predicted motion vector index by applying entropy decoding to the entropy-coded predicted motion vector index.
[0549] The motion vector to be actually used for inter prediction of the target block may not match the predicted motion vector. In order to indicate the difference between the motion vector to be actually used for inter prediction of the target block and the predicted motion vector, MVD may be used. The encoding apparatus 100 may derive a predicted motion vector similar to the motion vector to be actually used for inter prediction of the target block so as to use as small an MVD as possible.
[0550] The motion vector difference (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 may generate an entropy-encoded MVD by applying entropy encoding to the MVD. The encoding apparatus 100 may generate a bitstream including the entropy-encoded MVD.
[0551] The MVD may be transmitted from the encoding apparatus 100 to the decoding apparatus 200 through a bitstream. The decoding apparatus 200 may extract the entropy-encoded MVD from the bitstream, and may acquire the MVD by applying entropy decoding to the entropy-encoded MVD.
[0552] The decoding device 200 may derive the motion vector of the target block by summing the MVD and the predicted motion vector. In other words, the motion vector of the target block derived by the decoding device 200 may be the sum of the MVD and the motion vector candidate.
[0553] In addition, the encoding device 100 may generate entropy-encoded MVD resolution information by applying entropy encoding to the calculated MVD resolution information, and may generate a bitstream including the entropy-encoded MVD resolution information. The decoding device 200 may extract the entropy-encoded MVD resolution information from the bitstream, and may obtain the MVD resolution information by applying entropy decoding to the entropy-encoded MVD resolution information. The decoding device 200 may use the MVD resolution information to adjust the resolution of the MVD.
[0554] In addition, the encoding apparatus 100 may calculate the MVD based on the affine model. The decoding apparatus 200 may derive the affine controlled motion vector of the target block through the sum of the MVD and the affine controlled motion vector candidate, and may derive the motion vector of the subblock using the affine controlled motion vector.
[0555] The reference direction may indicate a list of reference pictures to be used for prediction of the target block. For example, the reference direction may indicate one of reference picture list L0 and reference picture list L1.
[0556] The reference direction only indicates a reference picture list to be used for prediction of a target block and may not mean that the direction of the reference picture is limited to the forward direction or the backward direction. In other words, each of the reference picture lists L0 and L1 may include pictures in the forward direction and / or the backward direction.
[0557] The reference direction being unidirectional may mean using a single reference picture list. The reference direction being bidirectional may mean using two reference picture lists. In other words, the reference direction may indicate one of a case where only reference picture list L0 is used, a case where only reference picture list L1 is used, and a case where two reference picture lists are used.
[0558] The reference picture index may indicate a reference picture used to predict a target block among reference pictures present in the reference picture list. The encoding device 100 may generate an entropy-coded reference picture index by applying entropy coding to the reference picture index, and may generate a bitstream including the entropy-coded reference picture index. The entropy-coded reference picture index may be signaled from the encoding device 100 to the decoding device 200 through a bitstream. The decoding device 200 may extract the entropy-coded reference picture index from the bitstream, and may obtain the reference picture index by applying entropy decoding to the entropy-coded reference picture index.
[0559] When two reference picture lists are used to predict a target block, a single reference picture index and a single motion vector may be used for each of the reference picture lists. In addition, when two reference picture lists are used to predict a target block, two prediction blocks may be specified for the target block. For example, the (final) prediction block of the target block may be generated using an average or weighted sum of the two prediction blocks of the target block.
[0560] The motion vector of the target block may be derived by predicting the motion vector index, MVD, reference direction, and reference picture index.
[0561] The decoding apparatus 200 may generate a prediction block of the target block based on the derived motion vector and the reference picture index. For example, the prediction block may be a reference block indicated by the derived motion vector in the reference picture indicated by the reference picture index.
[0562] Since the predicted motion vector index and the MVD are encoded without encoding the motion vector itself of the target block, the number of bits transmitted from the encoding apparatus 100 to the decoding apparatus 200 can be reduced, and encoding efficiency can be improved.
[0563] For the target block, the motion information of the reconstructed neighboring blocks may be used. In a specific inter-frame prediction mode, the encoding device 100 may not separately encode the actual motion information of the target block. Instead of encoding the motion information of the target block, additional information that can be used to derive the motion information of the target block using the motion information of the reconstructed neighboring blocks may be encoded. Since the additional information is encoded, the number of bits sent to the decoding device 200 may be reduced, and the encoding efficiency may be improved.
[0564] For example, as an inter-frame prediction mode in which the motion information of the target block is not directly encoded, there may be a skip mode and / or a merge mode. Here, each of the encoding device 100 and the decoding device 200 may use an identifier and / or an index indicating a unit whose motion information among the reconstructed neighboring units is to be used as the motion information of the target unit.
[0565] 2) Merge mode
[0566] As a scheme for deriving motion information of a target block, there is merging. The term "merge" may mean the merging of motions of multiple blocks. "Merge" may mean that the motion information of one block is also applied to other blocks. In other words, the merge mode may be a mode for deriving the motion information of a target block from the motion information of neighboring blocks.
[0567] When the merge mode is used, the encoding device 100 may use the motion information of the spatial candidate and / or the motion information of the temporal candidate to predict the motion information of the target block. The spatial candidate may include a reconstructed spatial neighboring block that is spatially adjacent to the target block. The spatial neighboring blocks may include left neighboring blocks and upper neighboring blocks. The temporal candidate may include a col block. The terms "spatial candidate" and "spatial merge candidate" may be used to have the same meaning and may be used interchangeably with each other. The terms "temporal candidate" and "temporal merge candidate" may be used to have the same meaning and may be used interchangeably with each other.
[0568] The encoding apparatus 100 may obtain a prediction block through prediction. The encoding apparatus 100 may encode a residual block, which is a difference between a target block and a prediction block.
[0569] 2-1) Creation of Merge Candidate List
[0570] When the merge mode is used, each of the encoding device 100 and the decoding device 200 may use the motion information of the spatial candidate and / or the motion information of the temporal candidate to create a merge candidate list. The motion information may include 1) a motion vector, 2) a reference picture index, and 3) a reference direction. The reference direction may be unidirectional or bidirectional. The reference direction may mean an inter-frame prediction indicator.
[0571] The merge candidate list may include a merge candidate. The merge candidate may be motion information. In other words, the merge candidate list may be a list in which a plurality of pieces of motion information are stored.
[0572] The merge candidate may be a plurality of pieces of motion information of a temporal candidate and / or a spatial candidate. In other words, the merge candidate list may include motion information of a temporal candidate and / or a spatial candidate, etc.
[0573] Furthermore, the merge candidate list may include a new merge candidate generated by combining merge candidates already present in the merge candidate list. In other words, the merge candidate list may include new motion information generated by combining a plurality of pieces of motion information previously present in the merge candidate list.
[0574] In addition, the merge candidate list may include a history-based merge candidate. The history-based merge candidate may be motion information of a block encoded and / or decoded before the target block.
[0575] Furthermore, the merge candidate list may include a merge candidate based on an average of two merge candidates.
[0576] The merge candidate may be a specific mode for deriving inter-frame prediction information. The merge candidate may be information indicating a specific mode for deriving inter-frame prediction information. The inter-frame prediction information of the target block may be derived according to the specific mode indicated by the merge candidate. In addition, the specific mode may include a process for deriving a series of inter-frame prediction information. The specific mode may be an inter-frame prediction information derivation mode or a motion information derivation mode.
[0577] Inter prediction information of the 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.
[0578] For example, the motion information derivation mode in the merge candidate list may be at least one of 1) a motion information derivation mode for a sub-block unit and 2) an affine motion information derivation mode.
[0579] In addition, the merge candidate list may include motion information of a zero vector. A zero vector may also be referred to as a "zero merge candidate."
[0580] In other words, the multiple motion information in the merge candidate list can be at least one of the following items: 1) motion information of spatial candidates, 2) motion information of temporal candidates, 3) motion information generated by combining multiple motion information previously existing in the merge candidate list, and 4) zero vector.
[0581] The motion information may include 1) a motion vector, 2) a reference picture index, and 3) a reference direction. The reference direction may also be referred to as an "inter prediction indicator". The reference direction may be unidirectional or bidirectional. A unidirectional reference direction may indicate L0 prediction or L1 prediction.
[0582] A merge candidate list may be created before performing prediction in merge mode.
[0583] The number of merge candidates in the merge candidate list may be predefined. Each of the encoding device 100 and the decoding device 200 may add merge candidates to the merge candidate list according to a predefined scheme and a predefined priority so that the merge candidate list has a predefined number of merge candidates. The merge candidate list of the encoding device 100 and the merge candidate list of the decoding device 200 may be made identical to each other using a predefined scheme and a predefined priority.
[0584] The merge may be applied on a CU or PU basis. When the merge is performed on a CU basis or a PU basis, the encoding device 100 may transmit a bitstream including predefined information to the decoding device 200. For example, the predefined information may include 1) information indicating whether to perform the merge on each block partition, and 2) information on a block to be merged with among blocks that are spatial candidates and / or temporal candidates of the target block.
[0585] 2-2) Searching for motion vectors using the merge candidate list
[0586] The encoding device 100 may determine a merge candidate to be used for encoding the target block. For example, the encoding device 100 may perform prediction on the target block using a merge candidate in the merge candidate list, and may generate a residual block of the merge candidate. The encoding device 100 may encode the target block using a merge candidate that causes the minimum cost in encoding the prediction and residual block.
[0587] Also, the encoding apparatus 100 may determine whether to encode the target block using the merge mode.
[0588] 2-3) Transmission of inter-frame prediction information
[0589] The encoding device 100 may generate a bitstream including inter-frame prediction information required for inter-frame prediction. The encoding device 100 may generate entropy-coded inter-frame prediction information by performing entropy encoding on the inter-frame prediction information, and may transmit the bitstream including the entropy-coded inter-frame prediction information to the decoding device 200. Through the bitstream, the entropy-coded inter-frame prediction information may be signaled by the encoding device 100 to the decoding device 200. The decoding device 200 may extract the entropy-coded inter-frame prediction information from the bitstream, and may obtain the inter-frame prediction information by applying entropy decoding to the entropy-coded inter-frame prediction information.
[0590] The decoding apparatus 200 may perform inter prediction on the target block using inter prediction information of the bitstream.
[0591] The inter prediction information may include 1) mode information indicating whether a merge mode is used, 2) a merge index, and 3) correction information.
[0592] In addition, the inter-frame prediction information may include a residual signal.
[0593] Only when the mode information indicates that the merge mode is used, the decoding device 200 can obtain the merge index from the bitstream.
[0594] The mode information may be a merge flag. The unit of the mode information may be a block. The information about the block may include the mode information, and the mode information may indicate whether the merge mode is applied to the block.
[0595] The merge index may indicate a merge candidate to be used for prediction of the target block among the merge candidates included in the merge candidate list. Alternatively, the merge index may indicate a block with which the target block is to be merged among neighboring blocks spatially or temporally adjacent to the target block.
[0596] The encoding apparatus 100 may select a merge candidate having the highest encoding performance among the merge candidates included in the merge candidate list, and set a value of the merge index to indicate the selected merge candidate.
[0597] The correction information may be information for correcting a motion vector. The encoding apparatus 100 may generate the correction information. The decoding apparatus 200 may correct a motion vector of a merge candidate selected by a merge index based on the correction information.
[0598] 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 that corrects a motion vector based on signaled correction information may be referred to as a 'merge mode with motion vector difference'.
[0599] 2-4) Inter-frame prediction using merge mode of inter-frame prediction information
[0600] The decoding apparatus 200 may perform prediction on the target block using a merge candidate indicated by a merge index among merge candidates included in the merge candidate list.
[0601] 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.
[0602] 3) Skip mode
[0603] The skip mode may be a mode in which the motion information of the spatial candidate or the motion information of the temporal candidate is applied to the target block without change. In addition, the skip mode may be a mode in which the residual signal is not used. In other words, when the skip mode is used, the reconstructed block may be the same as the predicted block.
[0604] The difference between the merge mode and the skip mode is whether the residual signal is transmitted or used. That is, the skip mode may be similar to the merge mode except that the residual signal is not transmitted or used.
[0605] When the skip mode is used, the encoding device 100 may transmit information about a block whose motion information in a block as a spatial candidate or a temporal candidate is to be used as motion information of a target block to the decoding device 200 through a bitstream. The encoding device 100 may generate entropy-coded information by performing entropy coding on the information, and may signal the entropy-coded information to the decoding device 200 through a bitstream. The decoding device 200 may extract the entropy-coded information from the bitstream, and may obtain information by applying entropy decoding to the entropy-coded information.
[0606] In addition, when the skip mode is used, the encoding device 100 may not transmit other syntax information, such as MVD, to the decoding device 200. For example, when the skip mode is used, the encoding device 100 may not signal a syntax element related to at least one of the MVD, the coded block flag, and the transform coefficient level to the decoding device 200.
[0607] 3-1) Creation of Merge Candidate List
[0608] The skip mode may also use a merge candidate list. In other words, the merge candidate list may be used in both the merge mode and the skip mode. In this regard, the merge candidate list may also be referred to as a "skip candidate list" or a "merge / skip candidate list".
[0609] Alternatively, the skip mode may use an additional candidate list different from the candidate list of the merge mode. In this case, in the following description, the merge candidate list and the merge candidate may be replaced with the skip candidate list and the skip candidate, respectively.
[0610] The merge candidate list may be created before performing prediction in skip mode.
[0611] 3-2) Searching for motion vectors using the merge candidate list
[0612] The encoding device 100 may determine a merge candidate to be used for encoding the target block. For example, the encoding device 100 may perform prediction on the target block using a merge candidate in the merge candidate list. The encoding device 100 may encode the target block using a merge candidate that causes the minimum cost in prediction.
[0613] Also, the encoding apparatus 100 may determine whether to encode the target block using the skip mode.
[0614] 3-3) Transmission of inter-frame prediction information
[0615] 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 the inter prediction information in the bitstream.
[0616] The inter prediction information may include 1) mode information indicating whether a skip mode is used, and 2) a skip index.
[0617] The skip index can be the same as the merge index described above.
[0618] When the skip mode is used, the target block may be encoded without using the residual signal. The inter prediction information may not include the residual signal. Alternatively, the bitstream may not include the residual signal.
[0619] Only when the mode information indicates that the skip mode is used, the decoding device 200 can obtain the skip index from the bitstream. As described above, the merge index and the skip index can be the same as each other. Only when the mode information indicates that the merge mode or the skip mode is used, the decoding device 200 can obtain the skip index from the bitstream.
[0620] The skip index may indicate a merge candidate to be used for prediction of a target block among merge candidates included in the merge candidate list.
[0621] 3-4) Inter-frame prediction in skip mode using inter-frame prediction information
[0622] The decoding apparatus 200 may perform prediction on the target block using a merge candidate indicated by a skip index among merge candidates included in the merge candidate list.
[0623] The motion vector of the target block may be specified by the motion vector of the merge candidate indicated by the skip index, the reference picture index, and the reference direction.
[0624] 4) Current picture reference mode
[0625] The current picture reference mode may indicate a prediction mode using a previously reconstructed region in a target picture to which the target block belongs.
[0626] A motion vector for specifying a previously reconstructed region may be used.A reference picture index of a target block may be used to determine whether the target block has been encoded in a current picture reference mode.
[0627] A flag or index indicating whether the target block is a block encoded in the current picture reference mode may be signaled by the encoding apparatus 100 to the decoding apparatus 200. Alternatively, whether the target block is a block encoded in the current picture reference mode may be inferred by a reference picture index of the target block.
[0628] When the target block is encoded in the current picture reference mode, the target picture may exist at a fixed position or an arbitrary position in the reference picture list of the target block.
[0629] For example, the fixed position may be a position where the value of the reference picture index is 0 or a last position.
[0630] When the target picture exists at an arbitrary position in the reference picture list, an additional reference picture index indicating such arbitrary position may be signaled by the encoding apparatus 100 to the decoding apparatus 200 .
[0631] 5) Sub-block merging mode
[0632] The sub-block merge mode may be a mode in which motion information is derived from sub-blocks of a CU.
[0633] When the subblock merge mode is applied, the subblock merge candidate list may be generated using the motion information of the co-located subblock (col subblock) of the target subblock in the reference image (ie, subblock-based temporal merge candidates) and / or affine control point motion vector merge candidates.
[0634] 6) Triangle partitioning mode
[0635] In the triangular partition mode, the target block can be partitioned in the diagonal direction, and sub-target blocks generated by the partitioning can be generated. For each sub-target block, motion information of the corresponding sub-target block can be derived, and the derived motion information can be used to derive the prediction samples of each sub-target block. The prediction samples of the target block can be derived by the weighted sum of the prediction samples of the sub-target blocks generated by the partitioning.
[0636] 7) Combined inter-intra prediction mode
[0637] The combined inter-intra prediction mode may be a mode of deriving a prediction sample of a target block using a weighted sum of a prediction sample generated via inter prediction and a prediction sample generated via intra prediction.
[0638] In the above-mentioned mode, the decoding device 200 may autonomously correct the derived motion information. For example, the decoding device 200 may search for motion information having a minimum sum of absolute differences (SAD) in a specific area based on a reference block indicated by the derived motion information, and may derive the found motion information as the corrected motion information.
[0639] In the above-described mode, the decoding apparatus 200 may compensate for the prediction sample derived via the inter-frame prediction using the optical flow.
[0640] In the above-mentioned AMVP mode, merge mode, skip mode, etc., motion information to be used for prediction of a target block may be specified among a plurality of pieces of motion information in a list using index information of the list.
[0641] In order to improve encoding efficiency, the encoding apparatus 100 may signal only the index of the element that causes the minimum cost in inter prediction of the target block among the elements in the list. The encoding apparatus 100 may encode the index and may signal the encoded index.
[0642] Therefore, the above lists (i.e., the predicted motion vector candidate list and the merge candidate list) must be able to be derived by the encoding device 100 and the decoding device 200 using the same scheme based on the same data. Here, the same data may include a reconstructed picture and a reconstructed block. In addition, in order to specify an element using an index, the order of the elements in the list must be fixed.
[0643] Fig.10 Spatial candidates according to an embodiment are shown.
[0644] exist Fig.10 , the positions of the spatial candidates are shown.
[0645] The large block in the center of the figure may represent the target block, and the five small blocks may represent spatial candidates.
[0646] The coordinates of the target block may be (xP, yP), and the size of the target block may be represented by (nPSW, nPSH).
[0647] The spatial candidate A0 may be a block adjacent to the lower left corner of the target block. A0 may be a block occupying pixels located at coordinates ((xP-1, yP+nPSH).
[0648] Spatial candidate A1 may be a block adjacent to the left side of the target block. A1 may be the bottommost block among the blocks adjacent to the left side of the target block. Alternatively, A1 may be a block adjacent to the top of A0. A1 may be a block occupying pixels located at coordinates (xP-1, yP+nPSH-1).
[0649] The spatial candidate B0 may be a block adjacent to the upper right corner of the target block. B0 may be a block occupying pixels located at coordinates (xP+nPSW, yP-1).
[0650] Spatial candidate B1 may be a block adjacent to the top of the target block. B1 may be the rightmost block among the blocks adjacent to the top of the target block. Alternatively, B1 may be a block adjacent to the left of B0. B1 may be a block occupying pixels located at coordinates (xP+nPSW-1, yP-1).
[0651] The spatial candidate B2 may be a block adjacent to the upper left corner of the target block. B2 may be a block occupying pixels located at coordinates (xP-1, yP-1).
[0652] Determination of availability of spatial and temporal candidates
[0653] 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.
[0654] Hereinafter, the candidate blocks may include spatial candidates and temporal candidates.
[0655] For example, the determination may be performed by sequentially applying the following steps 1) to 4) .
[0656] Step 1) When the PU including the candidate block is outside the boundary of the picture, the availability of the candidate block may be set to “false.” The expression “availability is set to false” may have the same meaning as “set to unavailable.”
[0657] Step 2) When the PU including the candidate block is outside the boundary of the slice, the availability of the candidate block may be set to “false.” When the target block and the candidate block are located in different slices, the availability of the candidate block may be set to “false.”
[0658] Step 3) When the PU including the candidate block is outside the boundary of the tile, the availability of the candidate block may be set to “false”. When the target block and the candidate block are located in different tiles, the availability of the candidate block may be set to “false”.
[0659] Step 4) When the prediction mode of the PU including the candidate block is an intra prediction mode, the availability of the candidate block may be set to “false”. When the PU including the candidate block does not use inter prediction, the availability of the candidate block may be set to “false”.
[0660] Fig.11 An order in which motion information of spatial candidates is added to a merge list according to an embodiment is shown.
[0661] like Fig.11 As shown in , when multiple motion information of spatial candidates are added to the merge list, the order of A1, B1, B0, A0 and B2 can be used. That is, multiple motion information of available spatial candidates can be added to the merge list in the order of A1, B1, B0, A0 and B2.
[0662] Methods for deriving merge lists in merge mode and skip mode
[0663] 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 sent from the encoding device 100 to the decoding device 200. The slice header of the slice may include N. In other words, the maximum number of merge candidates in the merge list for the target block of the slice may be set by the slice header. For example, the value of N may be substantially 5.
[0664] A plurality of pieces of motion information (ie, merge candidates) may be added to the merge list in the order of the following steps 1) to 4).
[0665] Step 1) In the spatial candidates, available spatial candidates can be added to the merge list. Multiple motion information of available spatial candidates can be Fig.11 The order shown is added to the merge list. Here, when the motion information of the available space candidate overlaps with other motion information already in the merge list, the motion information of the available space candidate may not be added to the merge list. The operation of checking whether the corresponding motion information overlaps with other motion information existing in the list may be referred to as "overlap check".
[0666] The maximum number of added motion information pieces may be N.
[0667] Step 2)When the number of motion information in the merge list is less than N and the temporal candidate is available, the motion information of the temporal candidate may be added to the merge list. Here, when the motion information of the available temporal candidate overlaps with other motion information already in the merge list, the motion information of the available temporal candidate may not be added to the merge list.
[0668] Step 3) When the number of pieces of motion information in the merge list is less than N and the type of the target slice is 'B', combined motion information generated by combined bidirectional prediction (bi-prediction) may be added to the merge list.
[0669] The target slice may be a slice including the target block.
[0670] The combined motion information may be a combination of L0 motion information and L1 motion information. The L0 motion information may be motion information that refers only to the reference picture list L0. The L1 motion information may be motion information that refers only to the reference picture list L1.
[0671] In the merge list, there may be one or more pieces of L0 motion information. In addition, in the merge list, there may be one or more pieces of L1 motion information.
[0672] The combined motion information may include one or more pieces of combined motion information. When generating the combined motion information, one or more pieces of L0 motion information and one or more pieces of L1 motion information may be predefined to be used for the generated L0 motion information and L1 motion information. The one or more pieces of combined motion information may be generated in a predefined order via a combined bidirectional prediction using a pair of different motion information in a merge list. One of the pair of different motion information may be L0 motion information, and the other of the pair may be L1 motion information.
[0673] For example, the combined motion information added with the highest priority may be a combination of L0 motion information with merge index 0 and L1 motion information with merge index 1. When the motion information with merge index 0 is not L0 motion information or when the motion information with merge index 1 is not L1 motion information, the combined motion information may be neither generated nor added. Next, the combined motion information added with the next priority may be a combination of L0 motion information with merge index 1 and L1 motion information with merge index 0. The subsequent detailed combination may conform to other combinations in the field of video encoding / decoding.
[0674] Here, when the combined motion information overlaps with other motion information already present in the merge list, the combined motion information may not be added to the merge list.
[0675] Step 4)When the number of pieces of motion information in the merge list is less than N, the motion information of the zero vector may be added to the merge list.
[0676] The zero-vector motion information may be motion information in which the motion vector is a zero vector.
[0677] The number of pieces of zero vector motion information may be one or more. The reference picture indexes of one or more pieces of zero vector motion information may be different from each other. For example, the 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.
[0678] The number of pieces of zero-vector motion information may be the same as the number of reference pictures in the reference picture list.
[0679] The reference direction of the zero-vector motion information may be bidirectional. Both of the motion vectors may be zero vectors. The number of pieces 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. Optionally, 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 reference direction that is unidirectional may be used for a reference picture index that may be applied only to a single reference picture list.
[0680] The encoding device 100 and / or the decoding device 200 may sequentially add zero-vector motion information to the merge list while changing the reference picture index.
[0681] When 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.
[0682] The order of the above steps 1) to 4) is merely exemplary and may be changed. In addition, some of the above steps may be omitted according to predefined conditions.
[0683] Method for deriving a candidate list of motion vector predictions in AMVP mode
[0684] The maximum number of motion vector predictor candidates in the motion vector predictor candidate list may be predefined. The predefined maximum number is indicated by N. For example, the predefined maximum number may be 2.
[0685] A plurality of pieces of motion information (ie, prediction motion vector candidates) may be added to the prediction motion vector candidate list in the order of the following steps 1) to 3).
[0686] Step 1) An available spatial candidate among the spatial candidates may be added to the prediction motion vector candidate list. The spatial candidate may include a first spatial candidate and a second spatial candidate.
[0687] The first spatial candidate may be one of A0, A1, scaled A0, and scaled A1. The second spatial candidate may be one of B0, B1, B2, scaled B0, scaled B1, and scaled B2.
[0688] Multiple pieces of motion information of available spatial candidates may be added to the prediction motion vector candidate list in the order of the first spatial candidate and the second spatial candidate. In this case, when the motion information of the available spatial candidate overlaps with other motion information already in the prediction motion vector candidate list, the motion information of the available spatial candidate may not be added to the prediction motion vector candidate list. In other words, when the value of N is 2, if the motion information of the second spatial candidate is the same as the motion information of the first spatial candidate, the motion information of the second spatial candidate may not be added to the prediction motion vector candidate list.
[0689] The maximum number of added motion information pieces may be N.
[0690] Step 2) When the number of pieces of motion information in the prediction motion vector candidate list is less than N and the temporal candidate is available, the motion information of the temporal candidate may be added to the prediction motion vector candidate list. In this case, when the motion information of the available temporal candidate overlaps with other motion information already in the prediction motion vector candidate list, the motion information may not be added to the prediction motion vector candidate list.
[0691] Step 3) When the number of pieces of motion information in the prediction motion vector candidate list is less than N, zero-vector motion information may be added to the prediction motion vector candidate list.
[0692] The zero-vector motion information may include one or more pieces of zero-vector motion information. The reference picture indexes of the one or more pieces of zero-vector motion information may be different from each other.
[0693] The encoding apparatus 100 and / or the decoding apparatus 200 may sequentially add a plurality of pieces of zero-vector motion information to the prediction motion vector candidate list while changing the reference picture index.
[0694] When the zero-vector motion information overlaps with other motion information already present in the prediction motion vector candidate list, the zero-vector motion information may not be added to the prediction motion vector candidate list.
[0695] The above description of the zero-vector motion information made in conjunction with the merge list can also be applied to the zero-vector motion information. The repeated description thereof will be omitted.
[0696] The order of the above steps 1) to 3) is merely exemplary and may be changed. In addition, some steps may be omitted according to predefined conditions.
[0697] Fig.12 Transformation and quantization processes according to examples are shown.
[0698] like Fig.12 As shown, the quantization level may be generated by performing a transform and / or quantization process on the residual signal.
[0699] The residual signal may be generated as a difference between the original block and the predicted block. Here, the predicted block may be a block generated via intra prediction or inter prediction.
[0700] The residual signal may be transformed into a signal in the frequency domain through a transform process as part of the quantization process.
[0701] The transform kernel used for the transform may include various DCT kernels such as discrete cosine transform (DCT) type 2 (DCT-II) and discrete sine transform (DST) kernel.
[0702] These transform kernels may perform a separable transform or a two-dimensional (2D) non-separable transform on the residual signal. The separable transform may be a transform indicating that a one-dimensional (1D) transform is performed on the residual signal in each of horizontal and vertical directions.
[0703] The DCT type and DST type adaptively used for 1D transform may include DCT-V, DCT-VIII, DST-I, and DST-VII in addition to DCT-II, as shown in each of the following Tables 3 and 4.
[0704] [Table 3]
[0705] Transformation Sets Transformation Candidates 0 DST-VII, DCT-VIII 1 DST-VII, DST-I 2 DST-VII, DCT-V
[0706] [Table 4]
[0707] Transformation Sets Transformation Candidates 0 DST-VII, DCT-VIII, DST-I 1 DST-VII, DST-I, DCT-VIII 2 DST-VII, DCT-V, DST-I
[0708] As shown in Tables 3 and 4, when deriving a DCT type or a DST type to be used for transformation, a transform set may be used. Each transform set may include multiple transform candidates. Each transform candidate may be a DCT type or a DST type.
[0709] Table 5 below shows an example of a transform set to be applied to a horizontal direction and a transform set to be applied to a vertical direction according to an intra prediction mode.
[0710] [Table 5]
[0711]
[0712] In Table 5, the numbers of vertical transform sets and horizontal transform sets in the horizontal direction to be applied to the residual signal according to the intra prediction mode of the target block are indicated.
[0713] As illustrated in Table 5, the transform set to be applied to the horizontal direction and the vertical direction may be predefined according to the intra prediction mode of the target block. The encoding device 100 may perform transform and inverse transform on the residual signal using the transform included in the transform set corresponding to the intra prediction mode of the target block. In addition, the decoding device 200 may perform inverse transform on the residual signal using the transform included in the transform set corresponding to the intra prediction mode of the target block.
[0714] In the transform and inverse transform, a transform set to be applied to the residual signal may be determined as illustrated in Tables 3, 4, and 5, and the transform set to be applied to the residual signal may not be signaled. The transform indication information may be signaled from the encoding apparatus 100 to the decoding apparatus 200. The transform indication information may be information indicating which one of a plurality of transform candidates included in the transform set to be applied to the residual signal is used.
[0715] For example, when the size of the target block is 64×64 or smaller, a transform set each having three transforms may be configured according to an intra prediction mode. An optimal transform method may be selected from a total of nine multi-transform methods generated by a combination of three transforms in the horizontal direction and three transforms in the vertical direction. Through such an optimal transform method, a residual signal may be encoded and / or decoded, and thus encoding efficiency may be improved.
[0716] Here, information indicating which of the transforms belonging to each transform set has been used for at least one of the vertical transform and the horizontal transform may be entropy encoded and / or decoded. Here, truncated unary binarization may be used to encode and / or decode such information.
[0717] As described above, methods using various transforms may be applied to a residual signal generated via intra prediction or inter prediction.
[0718] The transform may include at least one of a primary transform and a secondary transform. 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.
[0719] The first transform may be referred to as a “primary transform.” Furthermore, the first transform may also be referred to as an “adaptive multi-transform (AMT) scheme.” AMT may mean that, as described above, different transforms are applied to corresponding 1D directions (ie, vertical and horizontal directions).
[0720] The secondary transform may be a transform for improving the energy concentration on the transform coefficients generated by the primary transform. Similar to the primary transform, the secondary transform may be a separable transform or a non-separable transform. Such a non-separable transform may be a non-separable secondary transform (NSST).
[0721] The first transformation may be performed using at least one of a plurality of predefined transformation methods, for example, the plurality of predefined transformation methods may include discrete cosine transform (DCT), discrete sine transform (DST), Karhunen-Loeve transform (KLT), and the like.
[0722] Furthermore, the first transform may be a transform having various transform types according to a kernel function defining a discrete cosine transform (DCT) or a discrete sine transform (DST).
[0723] For example, the transform type may be determined based on at least one of the following items: 1) a prediction mode of the target block (e.g., one of intra-frame prediction and inter-frame prediction), 2) a size of the target block, 3) a shape of the target block, 4) an intra-frame prediction mode of the target block, 5) a component of the target block (e.g., one of a luminance component and a chrominance component), and 6) a partition type applied to the target block (e.g., one of a quadtree, a binary tree, and a ternary tree).
[0724] For example, according to the transform kernel presented in Table 6 below, the first transform may include transforms such as DCT-2, DCT-5, DCT-7, DST-7, DST-1, DST-8, and DCT-8. In Table 6 below, various transform types and transform kernel functions for multiple transform selection (MTS) are illustrated.
[0725] MTS may refer to selecting a combination of one or more DCT and / or DST kernels to transform the residual signal in the horizontal and / or vertical direction.
[0726] [Table 6]
[0727]
[0728] In Table 6, i and j may be integer values equal to or greater than 0 and less than or equal to N-1.
[0729] A secondary transform may be performed on transform coefficients generated by performing the primary transform.
[0730] As in the first transform, the transform set may also be defined in the secondary transform. The method for deriving and / or determining the above transform set may be applied not only to the first transform but also to the secondary transform.
[0731] The primary transformation and the secondary transformation can be determined for a specific goal.
[0732] For example, the primary transform and the secondary transform may be applied to a signal component corresponding to one or more of a luminance (luma) component and a chrominance (chroma) component. Whether to apply the primary transform and / or the secondary transform may be determined based on at least one of the encoding parameters for the target block and / or the neighboring block. For example, whether to apply the primary transform and / or the secondary transform may be determined based on the size and / or shape of the target block.
[0733] In the encoding apparatus 100 and the decoding apparatus 200 , transform information indicating a transform method to be used for a target may be derived by using the designation information.
[0734] For example, the transform information may include a transform index to be used for the primary transform and / or the secondary transform. Alternatively, the transform information may indicate that the primary transform and / or the secondary transform is not to be used.
[0735] For example, when the target of the primary transform and the secondary transform is a target block, the transform method to be applied to the primary transform and / or the secondary transform indicated by the transform information may be determined according to at least one of the encoding parameters of the target block and / or a block adjacent to the target block.
[0736] Alternatively, transform information indicating a transform method for a specific target may be transmitted from the encoding apparatus 100 to the decoding apparatus 200 .
[0737] For example, for a single CU, whether to use a primary transform, an index indicating the primary transform, whether to use a secondary transform, and an index indicating the secondary transform may be derived as transform information by the decoding device 200. Alternatively, for a single CU, transform information indicating whether to use a primary transform, an index indicating the primary transform, whether to use a secondary transform, and an index indicating the secondary transform may be transmitted by a signal.
[0738] The quantized transform coefficient (ie, the quantization level) may be generated by performing quantization on a result generated by performing the primary transform and / or the secondary transform, or by performing quantization on a residual signal.
[0739] Fig.13 A diagonal scan according to an example is shown.
[0740] Fig.14 A horizontal scan according to an example is shown.
[0741] Fig.15 A vertical scan according to an example is shown.
[0742] The quantized transform coefficient may be scanned via at least one of (upper right) diagonal scanning, vertical scanning, and horizontal scanning according to at least one of an intra prediction mode, a block size, and a block shape. The block may be a transform unit (TU).
[0743] Each scan may be initiated at a specific starting point and may be terminated at a specific ending point.
[0744] For example, by using Fig.13 The coefficients of the block can be scanned by diagonal scanning, and the quantized transform coefficients can be changed into 1D vector form. Optionally, the quantized transform coefficients can be used according to the size of the block and / or the intra prediction mode. Fig.14 Horizontal scan or Fig.15 The image is scanned vertically instead of diagonally.
[0745] Vertical scanning may be an operation of scanning 2D block coefficients in a column direction. Horizontal scanning may be an operation of scanning 2D block coefficients in a row direction.
[0746] In other words, which of diagonal scanning, vertical scanning, and horizontal scanning to use may be determined according to the size of a block and / or an inter prediction mode.
[0747] like Fig.13 , 14 As shown in FIG. 15 , the quantized transform coefficients may be scanned in a diagonal direction, a horizontal direction, or a vertical direction.
[0748] The quantized transform coefficients may be represented by a block shape. Each block may include multiple sub-blocks. Each sub-block may be defined according to a minimum block size or a minimum block shape.
[0749] In scanning, a scanning order according to a type or direction of scanning may be mainly applied to a subblock. In addition, a scanning sequence according to a scanning direction may be applied to a quantized transform coefficient in each subblock.
[0750] For example, Fig.13 , Fig.14 and Fig.15 As shown, when the size of the target block is 8×8, the quantized transform coefficients can be generated by first transforming, second transforming and quantizing the residual signal of the target block. Therefore, one of the three types of scanning sequences can be applied to the four 4×4 sub-blocks, and the quantized transform coefficients can also be scanned for each 4×4 sub-block according to the scanning sequence.
[0751] The encoding apparatus 100 may generate entropy-coded quantized transform coefficients by performing entropy encoding on the scanned quantized transform coefficients, and may generate a bitstream including the entropy-coded quantized transform coefficients.
[0752] The decoding device 200 may extract entropy-encoded quantized transform coefficients from the bitstream, and may generate quantized transform coefficients by performing entropy decoding on the entropy-encoded quantized transform coefficients. The quantized transform coefficients may be aligned in the form of 2D blocks via inverse scanning. Here, as a method of inverse scanning, at least one of upper right diagonal scanning, vertical scanning, and horizontal scanning may be performed.
[0753] In the decoding device 200, inverse quantization may be performed on the quantized transform coefficients. A secondary inverse transform may be performed on the result generated by performing the inverse quantization according to whether the secondary inverse transform is performed. In addition, a primary inverse transform may be performed on the result generated by performing the secondary inverse transform according to whether the primary inverse transform is to be performed. A reconstructed residual signal may be generated by performing the primary inverse transform on the result generated by performing the secondary inverse transform.
[0754] For luma components reconstructed via intra prediction or inter prediction, inverse mapping with dynamic range may be performed prior to in-loop filtering.
[0755] The dynamic range may be divided into 16 equal segments and the mapping function for each segment may be signaled.Such a mapping function may be signaled at a slice level or a tile group level.
[0756] An inverse mapping function for performing inverse mapping may be derived based on the mapping function.
[0757] In-loop filtering, storage of reference pictures, and motion compensation may be performed in the inverse mapping region.
[0758] The prediction block generated via inter prediction can be changed to a mapping region by mapping using a mapping function, and the changed prediction block can be used to generate a reconstructed block. However, since intra prediction is performed in the mapping region, the prediction block generated via intra prediction can be used to generate a reconstructed block without mapping and / or inverse mapping.
[0759] For example, when the target block is a residual block of a chroma component, the residual block may be changed into an inverse mapping area by scaling the chroma component of the mapping area.
[0760] Whether scaling is available can be signaled at the slice level or at the parallel block group level.
[0761] For example, scaling may be applied only if a mapping is available for luma components and the partitions of the luma components and the partitions of the chroma components follow the same tree structure.
[0762] Scaling may be performed based on an average value of values of samples in a luma prediction block corresponding to a chroma prediction block. Here, when a target block uses inter prediction, the luma prediction block may mean a mapped luma prediction block.
[0763] The value required for scaling may be derived by referencing the lookup table using the index of the segment to which the average value of the sample values of the luma prediction block belongs.
[0764] The residual block may be changed to the inverse mapping region by scaling the residual block using the finally derived value. Thereafter, for the block of the chrominance component, reconstruction, intra prediction, inter prediction, in-loop filtering, and storage of reference pictures may be performed in the inverse mapping region.
[0765] For example, information indicating whether mapping and / or inverse mapping of luma components and chroma components is available may be signaled through a sequence parameter set.
[0766] A prediction block of a target block may be generated based on a block vector. The block vector may indicate a displacement between the target block and a reference block. The reference block may be a block in a target image.
[0767] In this manner, a prediction mode that generates a prediction block by referring to a target image may be referred to as an 'intra-block copy (IBC) mode'.
[0768] The IBC mode may be applied to a CU of a specific size. For example, the IBC mode may be applied to an M×NCU. Here, M and N may be less than or equal to 64.
[0769] The IBC mode may include a skip mode, a merge mode, an AMVP mode, etc. In the case of the skip mode or the merge mode, a merge candidate list may be configured, and a merge index may be signaled, and thus a single merge candidate may be specified among the merge candidates present in the merge candidate list. The block vector of the specified merge candidate may be used as the block vector of the target block.
[0770] In the case of AMVP mode, a differential block vector may be transmitted with a signal. In addition, a prediction block vector may be derived from the left neighboring block and the upper neighboring block of the target block. In addition, an index indicating which neighboring block will be used may be transmitted with a signal.
[0771] The prediction block in the IBC mode may be included in the target CTU or the left CTU and may be limited to blocks within the previously reconstructed region. For example, the value of the block vector may be limited so that the prediction block of the target block is located in a specific region. The specific region may be a region defined by three 64×64 blocks encoded and / or decoded before the 64×64 block including the target block. The value of the block vector is limited in this way, and thus the memory consumption and device complexity caused by the implementation of the IBC mode may be reduced.
[0772] Fig.16 is a configuration diagram of an encoding device according to an embodiment.
[0773] The encoding device 1600 may correspond to the above-described encoding device 100 .
[0774] The encoding device 1600 may include a processing unit 1610, a memory 1630, a user interface (UI) input device 1650, a UI output device 1660, and a storage device 1640, which communicate with each other through a bus 1690. The encoding device 1600 may also include a communication unit 1620 coupled to a network 1699.
[0775] The processing unit 1610 may be a central processing unit (CPU) or a semiconductor device for executing processing instructions stored in the memory 1630 or the storage device 1640. The processing unit 1610 may be at least one hardware processor.
[0776] The processing unit 1610 may generate and process a signal, data, or information input to, output from, or used in the encoding device 1600, and may perform checks, comparisons, determinations, etc. related to the signal, data, or information. In other words, in an embodiment, generation and processing of data or information and checks, comparisons, and determinations related to data or information may be performed by the processing unit 1610.
[0777] The processing unit 1610 may include an inter-frame prediction unit 110, an intra-frame prediction unit 120, a switch 115, a subtractor 125, a transform unit 130, a quantization unit 140, an entropy encoding unit 150, an inverse quantization unit 160, an inverse transform unit 170, an adder 175, a filter unit 180 and a reference picture buffer 190.
[0778] At least some of the inter prediction unit 110, the intra prediction unit 120, the switch 115, the subtractor 125, the transform unit 130, the quantization unit 140, the entropy encoding unit 150, the inverse quantization unit 160, the inverse transform unit 170, the adder 175, the filter unit 180, and the reference picture buffer 190 may be program modules and may communicate with an external device or system. The program module may be included in the encoding apparatus 1600 in the form of an operating system, an application program module, or other program modules.
[0779] The program modules may be physically stored in various types of known storage devices. In addition, at least some of the program modules may also be stored in a remote storage device that can communicate with the encoding device 1600.
[0780] The program modules may include, but are not limited to, routines, subroutines, programs, objects, components, and data structures for performing functions or operations according to the embodiments or for implementing abstract data types according to the embodiments.
[0781] The program module may be implemented using instructions or codes executed by at least one processor of the encoding device 1600 .
[0782] The processing unit 1610 can execute instructions or codes in the inter-frame prediction unit 110, the intra-frame prediction unit 120, the switch 115, the subtractor 125, the transform unit 130, the quantization unit 140, the entropy encoding unit 150, the inverse quantization unit 160, the inverse transform unit 170, the adder 175, the filter unit 180 and the reference picture buffer 190.
[0783] The storage unit may represent the memory 1630 and / or the storage device 1640. Each of the memory 1630 and the storage device 1640 may be any of various types of volatile or non-volatile storage media. For example, the memory 1630 may include at least one of a read-only memory (ROM) 1631 and a random access memory (RAM) 1632.
[0784] 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 the storage unit.
[0785] For example, the storage unit may store pictures, blocks, lists, motion information, inter-frame prediction information, bitstreams, and the like.
[0786] The encoding device 1600 may be implemented in a computer system including a computer-readable storage medium.
[0787] The storage medium may store at least one module required for the operation of the encoding device 1600. The memory 1630 may store at least one module, and may be configured such that the at least one module is executed by the processing unit 1610.
[0788] Functions related to communication of data or information of the encoding device 1600 may be performed through the communication unit 1620 .
[0789] For example, the communication unit 1620 may transmit the bitstream to the decoding device 1600 which will be described later.
[0790] Fig.17 is a configuration diagram of a decoding device according to an embodiment.
[0791] The decoding device 1700 may correspond to the above-mentioned decoding device 200 .
[0792] The decoding device 1700 may include a processing unit 1710, a memory 1730, a user interface (UI) input device 1750, a UI output device 1760, and a storage device 1740, which communicate with each other through a bus 1790. The decoding device 1700 may also include a communication unit 1720 coupled to a network 1799.
[0793] The processing unit 1710 may be a central processing unit (CPU) or a semiconductor device for executing processing instructions stored in the memory 1730 or the storage device 1740. The processing unit 1710 may be at least one hardware processor.
[0794] The processing unit 1710 may generate and process a signal, data, or information input to, output from, or used in the decoding device 1700, and may perform checks, comparisons, determinations, etc. related to the signal, data, or information. In other words, in an embodiment, generation and processing of data or information and checks, comparisons, and determinations related to data or information may be performed by the processing unit 1710.
[0795] The processing unit 1710 may include an entropy decoding unit 210 , an inverse quantization unit 220 , an inverse transform unit 230 , an intra prediction unit 240 , an inter prediction unit 250 , a switch 245 , an adder 255 , a filter unit 260 , and a reference picture buffer 270 .
[0796] At least some of the entropy decoding unit 210, the inverse quantization unit 220, the inverse transform unit 230, the intra prediction unit 240, the inter prediction unit 250, the adder 255, the switch 245, the filter unit 260, and the reference picture buffer 270 of the decoding device 200 may be program modules and may communicate with an external device or system. The program module may be included in the decoding device 1700 in the form of an operating system, an application program module, or other program modules.
[0797] The program module can be physically stored in various types of well-known storage devices. In addition, at least some of the program modules can also be stored in a remote storage device that can communicate with the decoding device 1700.
[0798] The program modules may include, but are not limited to, routines, subroutines, programs, objects, components, and data structures for performing functions or operations according to the embodiments or for implementing abstract data types according to the embodiments.
[0799] The program module may be implemented using instructions or codes executed by at least one processor of the decoding device 1700 .
[0800] The processing unit 1710 may execute instructions or codes in the entropy decoding unit 210 , the inverse quantization unit 220 , the inverse transform unit 230 , the intra prediction unit 240 , the inter prediction unit 250 , the switch 245 , the adder 255 , the filter unit 260 , and the reference picture buffer 270 .
[0801] The storage unit may represent the memory 1730 and / or the storage device 1740. Each of the memory 1730 and the storage device 1740 may be any of various types of volatile or non-volatile storage media. For example, the memory 1730 may include at least one of the ROM 1731 and the RAM 1732.
[0802] 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.
[0803] For example, the storage unit may store pictures, blocks, lists, motion information, inter-frame prediction information, bitstreams, and the like.
[0804] The decoding device 1700 may be implemented in a computer system including a computer-readable storage medium.
[0805] The storage medium may store at least one module required for the operation of the decoding device 1700. The memory 1730 may store at least one module, and may be configured such that the at least one module is executed by the processing unit 1710.
[0806] Functions related to communication of data or information of the decoding device 1700 may be performed through the communication unit 1720 .
[0807] For example, the communication unit 1720 may receive a bitstream from the encoding device 1700 .
[0808] 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, with respect to a function related to prediction, the processing unit may represent the switch 115 and / or the switch 245. With respect to a 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. With respect to a 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. With respect to a 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. With respect to a function related to quantization, the processing unit may represent the quantization unit 140 and the inverse quantization unit 160, and may indicate the inverse quantization unit 220. Regarding functions related to entropy encoding and / or entropy decoding, the processing unit may indicate the entropy encoding unit 150 and / or the entropy decoding unit 210. Regarding functions related to filtering, the processing unit may indicate the filter unit 180 and / or the filter unit 260. Regarding functions related to reference pictures, the processing unit may indicate the reference picture buffer 190 and / or the reference picture buffer 270.
[0809] In the present specification, "specific information is predefined" means that the corresponding specific information in the encoding device and the decoding device is determined as the same value / method by the same rules and / or methods in the encoding device and the decoding device. The specific information may be one of encoding parameters and / or block information, a search mode and / or order, and whether a specific mode is used, but is not limited thereto. Entropy encoding / decoding may be performed in units of at least one of a sequence level, a picture level, a parallel block level, a parallel block group level, a slice level, a coding tree unit (CTU) level, a coding unit (CU) level, and a prediction unit (PU), but the unit of execution is not limited thereto.
[0810] In this specification, motion information (MI_L0, MI_L1) refers to motion information having motion information MI_L0 in the L0 direction and motion information MI_L1 in the L1 direction.
[0811] In this specification, 'motion information is corrected' means an intermediate search step and / or a final search step for performing a method of deriving decoder-side motion information from corresponding motion information to derive corrected motion information.
[0812] In this specification, a surrounding block refers to a block that is temporally or spatially close to a current block.
[0813] "Temporally close" means that it belongs to a picture whose picture order count (POC) interval differs by less than POC_THRES from the current picture and / or the co-located pictures of the current picture.
[0814] POC_THRES may be a positive number of 1 or more. POC_THRES may be predefined. Optionally, information about POC_THRES may be encoded / decoded.
[0815] "Spatially close" means that the spatial distance from the predefined position of the current block differs by less than SPATIAL_THRES.
[0816] For example, the predefined position of the current block may be one of the upper left, center, upper right, lower left, and lower right of the current block, but is not limited thereto. SPATIAL_THRES may be a positive number of 1 or more. Optionally, SPATIAL_THRES may be predefined. Information about SPATIAL_THRES may be encoded / decoded.
[0817] In this specification, "matching cost" may refer to a calculation result value of a cost function between templates in a method for deriving decoder-side motion information. For example, the matching cost may be at least one of a bilateral matching cost or a template matching cost, but is not limited thereto.
[0818] In this specification, "matching cost for specific motion information" can be one of the following matching costs: the matching cost between templates determined from the corresponding specific motion information, or the matching cost between templates determined based on motion information derived by the method for deriving decoder-side motion information from the corresponding motion information.
[0819] In the present specification, a 'motion vector' may be one of a motion vector in inter prediction or a block vector (BV) in intra block copy (IBC).
[0820] In this specification, a screen may have the same meaning as an image.
[0821] In this specification, index has the same meaning as index.
[0822] In this specification, bi-prediction, bi-directional prediction, inter-frame bi-prediction and bi-directional inter-frame prediction have the same meaning.
[0823] In this specification, surrounding samples refer to samples and / or pixels that are temporally or spatially close to the current block. Samples can be predicted samples or reconstructed samples. Pixels can be predicted pixels or reconstructed pixels.
[0824] When inter prediction is performed for the current block, the motion vector resolution may be adaptively determined. When adaptive motion vector resolution (AMVR) for adaptively determining the motion vector resolution is used, the resolution of the motion vector difference may be adjusted in units of blocks.
[0825] By adjusting the resolution of the motion vector difference, the encoding / decoding efficiency can be improved. The adjusted resolution may be one of 16-pel, 8-pel, 4-pel, full-pel, half-pel, or quarter-pel, but is not limited thereto. "Pel" refers to a unit of pixel. For example, when the adjusted resolution of the current block is 4-pel, each component of the motion vector difference may have a multiple value of 4 pixels.
[0826] Candidates that can be selected for motion vector difference resolution may be predefined in the encoding device and the decoding device. In this case, index information specifying one of the plurality of candidates may be encoded and signaled.
[0827] The motion vector of the current block may be derived by referring to a pre-reconstructed area adjacent to the current block. Specifically, the pre-reconstructed area adjacent to the current block may be set as a current template, and a reference template most similar to the current template may be searched in a search area. As above, deriving the motion vector of the current block by using a template may be referred to as template matching.
[0828] Fig.18 An example of applying template matching is shown.
[0829] When template matching is applied, a motion vector of the current block may be derived through a comparison operation between a current template of the current block and a reference template of a reference block.
[0830] In template matching, each of the current template and the reference template may be configured by including at least one of the following subsets: a subset of pixels belonging to N lines adjacent to the left side of the current block and the reference block, or a subset of pixels belonging to M lines adjacent to the top (above) of the current block and the reference block. Here, N and M are integers including 0, and may be 0, 1, 2, 3, 4, or an integer greater than 4. N and M may be set to the same value. For example, N and M may be predefined in the encoding device and the decoding device. Optionally, N and M may be determined differently according to at least one of the form of the current block or whether a block adjacent to the current block is encoded by inter-frame prediction.
[0831] Fig.19 An example of a configuration template is shown.
[0832] exist Fig.19 , an example of configuring a current template adjacent to a current block is shown, but a reference template adjacent to a reference block may also be configured in the same manner as the shown example.
[0833] As in Fig.19 In the examples shown in (a) and (b) of FIG. 1 , the current template may be configured with pixels included in N columns on the left side of the current block and pixels included in M columns at the top of the current block. In this case, as in Fig.19 In the example shown in (a), pixels whose x-axis and y-axis coordinates are both smaller than the value of the upper left pixel of the current block can be set to be not included in the template area. Fig.19 In the example shown in (a), pixels whose x-axis and y-axis coordinates are both smaller than the value of the upper left pixel of the current block may be set to be included in the template area.
[0834] Optionally, as in Fig.19 In the example shown in (c), the current template may be configured with only the left pixel of the current block, or as in Fig.19 In the example shown in (d), the current template may be configured with only the top pixels of the current block.
[0835] exist Fig.19 In FIG. 4 , N and M are shown to be 4 respectively.
[0836] However, the configurations of the current template and the reference template are not limited to the described examples.
[0837] The current template and the reference template with the lowest cost in the search area are searched, and the distance between the current block and the reference block is set as a motion vector. However, the method for deriving the motion vector is not limited thereto. Here, the cost between the current template and the reference template (i.e., template matching cost) may be calculated by a comparison operation between the current template and the reference template. Specifically, the sum of absolute differences (SAD) between the current template and the reference template may be set as the cost between the current template and the reference template.
[0838] The reference block may refer to a block indicated by an initial motion vector or a motion vector derived through template matching.
[0839] When template matching is applied, the search step may be performed by applying at least one search mode according to at least one of an inter prediction mode and a motion vector resolution applied to the current block. As an example, the search step for the current block may be performed according to at least one search mode according to the inter prediction mode and the resolution applied to the current block, in the order from top to bottom in Table 7.
[0840]
Table 7
[0841]
[0842] In Table 7, ALT_IF refers to the index of the adaptive interpolation filter. The adaptive interpolation filter is used to calculate the pixel value at the sample position of a specific resolution, and different interpolation filters can be selected according to the index. For example, the specific resolution may be half-pel, but is not limited thereto. The interpolation filter selected by the index may be a 6-tap interpolation filter or an 8-tap interpolation filter, but is not limited thereto. When performing a comparison operation between the current template and the reference template, sub-sampling may be performed in a predefined manner for each template. By performing a comparison operation between sub-sampled templates, the computational complexity may be reduced.
[0843] Fig. 20 An example is shown where subsampling is applied to a template.
[0844] Fig. 20 An example is shown in which subsampling is applied to the current template around the current block, but subsampling may also be performed in the same manner as the reference template.
[0845] As in Fig. 20 In the example shown in (a) of FIG. 1 , subsampling can be performed at intervals of 2 pixels in the vertical direction. As a result, as in Fig. 20 In the example shown in (a), only pixels belonging to odd-numbered rows can be used for template matching.
[0846] Alternatively, contrary to what is shown, subsampling may be performed to use only pixels belonging to even-numbered rows for template matching.
[0847] Although not shown, sub-sampling may also be performed at intervals of 2 pixels in the horizontal direction.
[0848] Optionally, as in Fig. 20 In the example shown in (b) of FIG. 1 , subsampling can be performed at intervals of 2 pixels in the vertical direction and the horizontal direction. As a result, as in Fig. 20 In the example shown in (b) of , only pixels at positions where the absolute value of the x-axis coordinate is an even number and the absolute value of the y-axis coordinate is an odd number can be used for template matching.
[0849] Even when the affine mode is applied, template matching can be applied.
[0850] Fig.21 is a diagram for describing an example of applying template matching when an affine mode is applied to a current block.
[0851] exist Fig.21 In , CPMV refers to the control point motion vector.
[0852] When the affine mode is applied, template matching can be performed in units of sub-templates. As an example, in the case of a sub-template of a predefined size (e.g., Fig.21After performing template matching on each of L0-L3 and A0-A3 in the current block, a sub-reference template with a minimum sum of costs of the sub-templates can be determined. Then, based on the sub-reference template, a motion vector of each sub-block in the current block can be derived or a control point motion vector of the current block can be derived.
[0853] Fig. 22 and Fig.23 Flowcharts of methods for obtaining a prediction block based on multi-prediction in an encoding device and a decoding device are respectively shown.
[0854] When multi-prediction is applied, multiple prediction blocks can be obtained. In this case, the first prediction block among the multiple prediction blocks can be obtained based on the same size and position as the current block. On the other hand, the second prediction block can be obtained based on a size or position different from the current block.
[0855] Fig.24 The position and size of each of the plurality of prediction blocks are shown.
[0856] exist Fig.24 In the example shown, it is shown that the first prediction block has the same position and size as the current block.
[0857] It is shown that the second prediction block and the third prediction block based on the upper left sample have the same position as the current block, but their sizes are smaller than the size of the current block.
[0858] It is shown that the fourth prediction block has a different position and size from the current block.
[0859] As described above, when multi-prediction is applied, at least one of the plurality of prediction blocks may have the same position and / or size as the current block, while the other prediction blocks may have positions and / or sizes different from the current block. In other words, any one of the plurality of prediction blocks may have a different size or a different position than the other prediction blocks.
[0860] Different from what is shown, all of the multiple prediction blocks may be obtained based on the position and / or size of the current block. Alternatively, all of the multiple prediction modes may be obtained based on the position and / or size of a sub-block within the current block.
[0861] Optionally, each of the plurality of prediction blocks may have a different position and / or size. As an example, a first prediction block among the plurality of prediction blocks may be generated to correspond to the position and size of a first subblock in the current block, and a second prediction block among the plurality of prediction blocks may be generated to correspond to the position and size of a second subblock in the current block.
[0862] Here, the subblock is a block partitioned from the current block and may have a size smaller than the current block. When the current block is partitioned into a plurality of subblocks, each of the subblocks may use an adjacent subblock as a neighboring block.
[0863] As an example, the size of the sub-block may be N×M. Here, N and M may be positive integers. For example, the sub-block size may be 4×4 or 8×8.
[0864] The information showing the subblock size may be entropy encoded / decoded in sequence units, in picture units, in slice units, in tile units, or in block units. Here, the unit of the block may be at least one of a CTU unit, a CU unit, or a PU unit.
[0865] Optionally, the size of the sub-block may be adaptively determined according to the size of the current block. For example, when the number of samples included in the current block is equal to or less than K, the size of the sub-block may be set to a size of 4×4. On the other hand, when the number of samples included in the current block is greater than K, the size of the sub-block may be set to 8×8. Here, K is a positive integer, for example, may be 256.
[0866] Optionally, the size of the sub-block may be predefined in the encoding device and the decoding device.
[0867] The sub-block may be a square or a rectangle. The shape of the sub-block may be determined according to the shape of the current block. As an example, when the current block is a rectangle, the sub-block may also be a rectangle.
[0868] The information about the shape of the subblock may be entropy encoded / decoded in sequence units, in picture units, in slice units, in tile units, or in block units.
[0869] Optionally, the shape of the sub-block may be predefined in the encoding device and the decoding device.
[0870] Hereinafter, for ease of description, it is assumed that the first prediction block among the plurality of prediction blocks is obtained based on the position and size of the current block. In addition, it is assumed that the remaining prediction blocks other than the first prediction block are obtained based on the position and size of the current block or based on the position and size of the subblock within the current block.
[0871] When it is assumed that, among a plurality of prediction blocks, the remaining prediction blocks except the first prediction block are obtained in units of subblocks, the plurality of prediction blocks may exist only for a portion of the plurality of subblocks in the current block.
[0872] As an example, for a first subblock of a current block, a first prediction block obtained by performing prediction on the current block and a second prediction block obtained by performing prediction on the first subblock may overlap. On the other hand, for a second subblock, there may be only a first prediction block obtained by performing prediction on the current block. In this case, for the first subblock, a final prediction block is obtained by a weighted sum operation of the first prediction block and the second prediction block, and for the second subblock, a subblock at a corresponding position in the first prediction block may be set as the final prediction block.
[0873] In addition, the number and / or position of the sub-blocks for generating the plurality of prediction blocks may be predefined in the encoding device and the decoding device. As an example, the plurality of prediction blocks may be generated only for the sub-blocks of the current block adjacent to a specific boundary. Here, the specific boundary may include at least one of the top boundary, the left boundary, the right boundary, or the bottom boundary of the current block.
[0874] Based on the above, multi-prediction is described in detail.
[0875] refer to Fig. 22 , multiple prediction blocks can be generated for the current block or sub-block S2210.
[0876] The final prediction block of the current block or sub-block may be obtained by weighted sum of multiple prediction blocks S2220.
[0877] In the encoding device, at least one of prediction information for at least one of the plurality of prediction blocks or information for weighted sum operation of the plurality of prediction blocks may be encoded and signaled S2230.
[0878] In the decoder, at least one of prediction information or information for weighted sum operation may be decoded from a bitstream S2310.
[0879] In the decoder, a plurality of prediction blocks may be generated for the current block or sub-block S2320.
[0880] And, a final prediction block of the current block or subblock may be obtained by performing a weighted sum on a plurality of prediction blocks S2330.
[0881] After determining whether to perform multi-prediction, a plurality of prediction blocks may be generated according to the determination or a single prediction block may be generated.
[0882] Optionally, for the current block, after obtaining the first prediction block, it may be determined whether to perform multi-prediction based on the characteristics of the first prediction block or prediction information used to obtain the first prediction block. When it is determined to perform multi-prediction, at least one prediction block may be additionally generated. On the other hand, when it is determined not to perform multi-prediction, the first prediction block may be determined as the final prediction block of the current block or sub-block.
[0883] In addition, whether to perform multi-prediction may be determined in units of sub-blocks. In this case, it may be determined whether to perform multi-prediction only on part of the sub-blocks.
[0884] Determines whether to apply multi-forecast
[0885] Information indicating whether multi-prediction is applied may be encoded and signaled. As an example, multipred_flag may be encoded and signaled, where multipred_flag is an indicator indicating whether multi-prediction is applied to the current block. multipred_flag may be encoded and signaled at the level of a picture, a slice, or a parallel block. Alternatively, multipred_flag may be encoded and signaled at the level of a block such as a coding tree unit (CTU) or a coding unit (CU). Optionally, multipred_enabled_flag indicating whether multi-prediction is allowed may be encoded and signaled at a higher level, and when multipred_enabled_flag is 1, multipred_flag may be encoded and signaled at the block level. Optionally, multipred_enabled_flag indicating whether multi-prediction is applied to each subblock may be encoded and signaled.
[0886] Multi-prediction can be performed only when the current block meets the predefined conditions. As an example, multipred_flag can be encoded / decoded only when the current block meets the predefined conditions. Optionally, multi-prediction can be performed by default when the current block meets the predefined conditions.
[0887] The predefined condition may be related to the encoding parameters and / or the block information. For example, the predefined condition may be about the prediction mode of the current block, the size of the current block, the size of the block vector, the size of the motion vector, the direction of the inter prediction, the information of the surrounding blocks, the index of the bidirectional prediction (BCW) with CU weight, the index of the adaptive motion vector resolution (AMVR), the indicator of the local illumination compensation (LIC), the indicator of the combined inter-intra prediction (CIIP), the indicator of whether to perform the position-dependent prediction combination (PDPC), and / or at least one of the indicators of the affine mode or the directional intra prediction mode, but is not limited thereto.
[0888] When CIIP is applied, a third prediction block of the current block is obtained by performing weighted summation on a first prediction block obtained through inter prediction and a second prediction block obtained through intra prediction.
[0889] PDPC refers to a technique for correcting a prediction value by at least one of a pixel position in a current block, an intra prediction mode of a current block, or a current block size. When PDPC is applied, the prediction value may be corrected by using a weighted sum of a filtered reference sample and an unfiltered reference sample or by using a difference between reference samples.
[0890] When the prediction value is corrected by using a weighted sum of the filtered reference sample and the unfiltered reference sample, the weight of the weighted sum may be determined based on at least one of the intra prediction mode of the current block or the size of the current block. In this case, the weight applied to the filtered reference sample and the weight applied to the unfiltered reference sample may have the same value or different values.
[0891] PDPC can be used in combination with intra prediction, or can be used in combination with CIIP. Alternatively, in CIIP, PDPC can be configured not to be used.
[0892] When bidirectional prediction is applied to the current block, BCW is a technique for determining a weight combination of reference blocks when generating a prediction block of the current block from a reference block in an L0 direction and a reference block in an L1 direction. For example, an index of a predefined table may be encoded / decoded, and the weight of each reference block may be determined by a weight indicated by an index in the predefined table.
[0893] LIC refers to a technique for correcting at least a portion of a prediction block by using at least one of a weight and an offset after deriving at least one of the weight and the offset by performing a comparison operation between a template of a current block and a template of a reference block. Specifically, the prediction block may be corrected by performing at least one of multiplying a weight by all or a portion of the prediction block or adding a weight by all or a portion of the prediction block.
[0894] The block information may include at least one of information about surrounding blocks, information about reference blocks, or information about reference blocks. Here, the reference block may be a block indicated by a motion vector or an initial motion vector of a current block in a reference picture.
[0895] The block information may include at least one of the encoding parameters. As an example, the information about the block may include at least one of the information used in inter-frame prediction, intra-frame prediction, transform, inverse transform, quantization, inverse quantization, entropy encoding / decoding, or loop filter. In other words, the block information may include at least one value or at least one of the combination forms of the following items: block size, block depth, block partition information, block shape (square or non-square), whether to perform quadtree-shaped partitioning, whether to perform binary tree-shaped partitioning, binary tree-shaped partition direction (horizontal direction or vertical direction), binary tree-shaped partition shape (symmetric partitioning or asymmetric partitioning), prediction mode (intra-frame prediction or inter-frame prediction), intra-frame luminance prediction mode / direction, intra-frame chrominance prediction mode / direction, intra-frame partition information, inter-frame partition information, coding block partition flag, block partition flag, transform block partition flag, reference sample filter label, reference sample filter coefficient, block filter label, block filter coefficient, block boundary filter label, block boundary filter coefficient, motion vector (motion vector of at least one of L0, L1, L2, L3, etc.), motion vector difference (motion vector difference of at least one of L0, L1, L2, L3, etc.), inter-frame prediction direction (for unidirectional prediction, bidirectional prediction, etc.), inter prediction of at least one of L0, L1, L2, L3, etc.), reference picture index (reference picture index for at least one of L0, L1, L2, L3, etc.), inter prediction indicator, prediction list utilization flag, reference picture list, motion vector prediction index, motion vector prediction candidate, motion vector candidate list, whether to use merge mode, merge index, merge candidate, merge candidate list, whether to use skip mode, interpolation filter type, interpolation filter tap, interpolation filter coefficient, motion vector size, motion vector representation accuracy (motion vector representation unit, such as integer sample, 1 / 2 sample, 1 / 4 sample, 1 / 8 sample, 1 / 16 sample, 1 / 32 sample, etc.), transform type, transform size, information on whether to use primary transform, information on whether to use secondary transform, primary transform index, secondary transform index, information on whether there is a residual signal, coding block mode, coding block flag, quantization parameter, residual quantization parameter, quantization matrix, whether to apply an in-loop filter,In-loop filter coefficients, in-loop filter label, in-loop filter shape / form, whether to apply deblocking filter, deblocking filter coefficients, deblocking filter label, deblocking filter strength, deblocking filter shape / form, whether to apply adaptive sample offset, adaptive sample offset value, adaptive sample offset category, adaptive sample offset type, whether to apply adaptive loop filter, adaptive loop filter coefficients, adaptive loop filter taps, adaptive loop filter shape / form, binarization / debinarization method, context model determination method, context model update method, whether to execute normal mode, whether to execute bypass mode, context binary bit, bypass binary bit, valid coefficient flag, last valid coefficient flag, coefficient group unit encoding flag, last valid coefficient position, flag whether the coefficient value is greater than 1 , a flag indicating whether the coefficient value is greater than 2, a flag indicating whether the coefficient value is greater than 3, information about the remaining coefficient value, positive and negative sign information, reconstructed luminance samples, reconstructed chrominance samples, residual luminance samples, residual chrominance samples, luminance transform coefficients, chrominance transform efficiency, luminance quantization level, chrominance quantization level, transform coefficient level scanning method, size of decoder horizontal motion vector search area, shape of decoder horizontal motion vector search area, number of decoder horizontal motion vector searches, CTU size information, minimum block size information, maximum block size information, maximum block depth information, minimum block depth information, slice identification information, slice partition information, parallel block identification information, parallel block type, parallel block partition information, input sample bit depth, reconstructed sample bit depth, residual sample bit depth, transform coefficient bit depth or quantization level bit depth. ,
[0896] At the block level, information about the prediction mode can be encoded and signaled. It can be used to determine whether multi-prediction is allowed based on the prediction mode of the current block.
[0897] As an example, multi-prediction may be performed only when the prediction mode of the current block is intra-prediction. For example, multipred_flag may be encoded / decoded only when the prediction mode of the current block is intra-prediction. Alternatively, whether to perform multi-prediction may be true when the prediction mode of the current block is intra-prediction.
[0898] As an example, multi-prediction may be performed only when the prediction mode of the current block is inter-prediction. For example, multipred_flag may be encoded / decoded only when the prediction mode of the current block is inter-prediction. Alternatively, whether to perform multi-prediction may be true when the prediction mode of the current block is inter-prediction.
[0899] As an example, multi-prediction may be performed only when the prediction mode of the current block is intra block copy. For example, multipred_flag may be encoded / decoded only when the prediction mode of the current block is intra block copy. Alternatively, whether to perform multi-prediction may be true when the prediction mode of the current block is intra block copy.
[0900] As an example, multi-prediction may be performed only when CIIP is not applied to the current block. For example, multipred_flag may be encoded / decoded only when CIIP is not applied to the current block. Alternatively, CIIP may be applied to the current block only when multi-prediction is not applied to the current block. Therefore, an indicator indicating whether CIIP is applied may be encoded / decoded only when multi-prediction is not applied to the current block.
[0901] As an example, when CIIP is applied to the current block, multi-prediction may always be applied. When CIIP is applied to the current block, encoding / decoding of multipred_flag may be omitted, and its value may be inferred to indicate that multi-prediction is performed. Alternatively, multipred_flag may be encoded / decoded only when CIIP is not applied to the current block.
[0902] As an example, multi-prediction may be performed only when PDPC is not applied to the current block. For example, multipred_flag may be encoded / decoded only when PDPC is not applied to the current block. Alternatively, PDPC may be applied to the current block only when multi-prediction is not applied to the current block. Therefore, an indicator indicating whether PDPC is applied may be encoded / decoded only when multi-prediction is not applied to the current block.
[0903] As an example, when PDPC is applied to the current block, multi-prediction may always be applied. When PDPC is applied to the current block, the encoding / decoding of multipred_flag may be omitted and its value may be inferred to indicate that multi-prediction is performed. Alternatively, multipred_flag may be encoded / decoded only when PDPC is not applied to the current block.
[0904] As an example, multi-prediction may be performed only when the geometry partition mode is not applied to the current block. For example, multipred_flag may be encoded / decoded only when the geometry partition mode is not applied to the current block. Alternatively, the geometry partition mode may be applied to the current block only when multi-prediction is not applied to the current block. Thus, the indicator indicating whether the geometry partition mode is applied may be encoded / decoded only when multi-prediction is not applied to the current block.
[0905] As an example, when the geometry partition mode is applied to the current block, multi-prediction may always be applied. When the geometry partition mode is applied to the current block, the encoding / decoding of multipred_flag may be omitted and its value may be inferred to indicate that multi-prediction is performed. Alternatively, multipred_flag may be encoded / decoded only when the geometry partition mode is not applied to the current block.
[0906] In addition, the geometric partitioning mode may include at least one of a geometric partitioning mode based on intra-frame prediction, a geometric partitioning mode based on inter-frame prediction, and a geometric partitioning mode based on a combination of intra-frame prediction and inter-frame prediction. Here, the geometric partitioning mode based on intra-frame prediction means that intra-frame prediction is performed on each of the first partition and the second partition generated by geometric partitioning. And, the geometric partitioning mode based on inter-frame prediction means that inter-frame prediction is performed on each of the first partition and the second partition generated by geometric partitioning. In addition, the geometric partitioning mode based on a combination of intra-frame prediction and inter-frame prediction means that intra-frame prediction is performed on one of the first partition and the second partition generated by geometric partitioning and inter-frame prediction is performed on the other.
[0907] In this case, multi-prediction may be performed only in a geometric partition mode that does not use intra prediction. In other words, the operation of determining whether to perform multi-prediction in the above-mentioned geometric mode may be applied only to a geometric partition mode based on inter prediction.
[0908] As an example, multi-prediction may be performed only when LIC is not applied to the current block. For example, multipred_flag may be encoded / decoded only when LIC is not applied to the current block. Alternatively, LIC may be applied to the current block only when multi-prediction is not applied to the current block. Therefore, an indicator indicating whether LIC is applied may be encoded / decoded only when multi-prediction is not applied to the current block.
[0909] As an example, when LIC is applied to the current block, multi-prediction may be always applied. When LIC is applied to the current block, encoding / decoding of multipred_flag may be omitted and its value may be inferred to indicate that multi-prediction is performed. Alternatively, when LIC is applied to the current block, whether to perform multi-prediction may be determined by multipred_flag. Alternatively, when LIC is applied to the current block and the size of the current block is equal to or greater than a threshold, it may be determined that multi-prediction is always performed.
[0910] According to an embodiment of the present disclosure, whether to allow multi-prediction may be determined according to the size of the current block or prediction information.
[0911] As an example, multi-prediction can be performed only when at least one of the width or height of the current block is equal to or greater than a threshold. For example, multipred_flag can be encoded / decoded only when both the width and height of the current block are equal to or greater than MULTIPREDMODE_MINSIZE. Optionally, whether to perform multi-prediction may be true when both the width and height of the current block are equal to or greater than MULTIPREDMODE_MINSIZE. At this time, MULTIPREDMODE_MINSIZE may be a positive integer equal to or greater than 0. MULTIPREDMODE_MINSIZE may be predefined in the encoding device and the decoding device. Optionally, information indicating MULTIPREDMODE_MINSIZE may be explicitly encoded and transmitted with a signal.
[0912] As an example, multi-prediction may be performed only when at least one of the width or height of the current block is less than or equal to a threshold. For example, multipred_flag may be encoded / decoded only when both the width and height of the current block are less than or equal to MULTIPREDMODE_MAXSIZE. Optionally, whether to perform multi-prediction may be true when both the width and height of the current block are equal to or greater than MULTIPREDMODE_MAXSIZE. In this case, MULTIPREDMODE_MAXSIZE may be a positive integer equal to or greater than 0. MULTIPREDMODE_MAXSIZE may be predefined in the encoding device and the decoding device. Optionally, information indicating MULTIPREDMODE_MAXSIZE may be explicitly encoded and signaled.
[0913] As an example, multi-prediction may be performed only when the number of samples belonging to the current block is equal to or greater than a threshold value. For example, multipred_flag may be encoded / decoded only when the number of samples belonging to the current block is equal to or greater than MULTIPREDMODE_MINSAMPLE. Optionally, whether to perform multi-prediction may be true when the number of samples belonging to the current block is equal to or greater than MULTIPREDMODE_MINSAMPLE. In this case, MULTIPREDMODE_MINSAMPLE is a positive integer equal to or greater than 0, and may be a value such as 64, 128, or 256. MULTIPREDMODE_MINSAMPLE may be predefined in an encoding device and a decoding device. Optionally, information indicating MULTIPREDMODE_MINSAMPLE may be explicitly encoded and signaled.
[0914] As an example, multi-prediction can be performed only when the number of samples belonging to the current block is less than or equal to a threshold. For example, multipred_flag can be encoded / decoded only when the number of samples belonging to the current block is less than or equal to MULTIPREDMODE_MAXSAMPLE. Optionally, whether to perform multi-prediction may be true when the number of samples belonging to the current block is less than or equal to MULTIPREDMODE_MAXSAMPLE. In this case, MULTIPREDMODE_MAXSAMPLE is a positive integer equal to or greater than 0, and may be a value such as 64, 128, or 256. MULTIPREDMODE_MAXSAMPLE may be predefined in the encoding device and the decoding device. Optionally, information indicating MULTIPREDMODE_MAXSAMPLE may be explicitly encoded and signaled.
[0915] As an example, multi-prediction may be performed only when the motion vector resolution of the current block is less than or equal to a threshold. For example, multipred_flag may be encoded / decoded only when the motion vector resolution of the current block is less than or equal to MULTIPRED_AMVRTHRES. Optionally, whether to perform multi-prediction may be true when the motion vector resolution of the current block is less than or equal to MULTIPRED_AMVRTHRES. In this case, MULTIPRED_AMVRTHRES may be an integer such as 1, 2, 4, 8, etc. or may be a real number less than 1. MULTIPRED_AMVRTHRES may be predefined in the encoding device and the decoding device. Optionally, information indicating MULTIPRED_AMVRTHRES may be explicitly encoded and signaled.
[0916] As an example, multi-prediction can be performed only when the motion vector resolution of the current block is equal to or greater than a threshold. For example, multipred_flag can be encoded / decoded only when the motion vector resolution of the current block is equal to or greater than MULTIPREDMODE_AMVRTHRES. Optionally, whether to perform multi-prediction may be true when the motion vector resolution of the current block is equal to or greater than MULTIPREDMODE_AMVRTHRES. In this case, MULTIPREDMODE_AMVRTHRES may be an integer such as 1, 2, 4, 8, etc. or may be a real number less than 1. MULTIPREDMODE_AMVRTHRES may be predefined in the encoding device and the decoding device. Optionally, information indicating MULTIPREDMODE_AMVRTHRES may be explicitly encoded and transmitted with a signal. As an example, when MULTIPREDMODE_AMVRTHRES is 4, it means that the motion vector resolution is 4 pixels.
[0917] As an example, whether to perform multi-prediction may be determined based on an inter-prediction indicator of the current block. For example, multi-prediction may be performed only when the prediction direction indicated by the intra-prediction direction of the current block is at least one of unidirectional, bidirectional, 3-directional, or 4-directional, or the number of prediction directions is equal to or greater than one of them. As an example, multi-prediction may be performed only when unidirectional prediction is applied to the current block, or multi-prediction may be performed only when bidirectional prediction is applied to the current block.
[0918] According to an embodiment of the present disclosure, whether to allow multi-prediction may be determined based on a comparison result between prediction information for a current block or a sub-block and prediction information for a surrounding block. Here, the surrounding block may be adjacent to the current block or may be adjacent to the sub-block. When determining whether to perform multi-prediction in units of sub-blocks, it may be determined that multi-prediction is performed only for some sub-blocks within the current block.
[0919] As an example, whether to perform multi-prediction may be determined according to whether the prediction modes of the current block (or sub-block) and the surrounding blocks are the same or different. For example, multipred_flag may be encoded / decoded only when the prediction mode of the current block (or sub-block) is different from the prediction mode of the surrounding blocks. Alternatively, whether to perform multi-prediction may be true when the prediction mode of the current block (or sub-block) is different from the prediction mode of the surrounding blocks.
[0920] As an example, whether to perform multi-prediction may be determined based on a result of comparing a difference between a motion vector of a current block (or sub-block) and a surrounding block or a difference between block vectors with a threshold. As an example, multi-prediction may be performed only when a difference between a motion vector of a current block (or sub-block) and a surrounding block or a difference between block vectors is equal to or greater than a threshold. For example, multipred_flag may be encoded / decoded only when a difference between a motion vector of a current block (or sub-block) and a surrounding block or a difference between block vectors is equal to or greater than MULTIPREDMODE_VECTORDIFFTHRES. Alternatively, whether to perform multi-prediction may be true when a difference between a motion vector of a current block (or sub-block) and a surrounding block or a difference between block vectors is equal to or greater than MULTIPREDMODE_VECTORDIFFTHRES. In this case, MULTIPREDMODE_VECTORDIFFTHRES may be equal to or greater than 0. MULTIPREDMODE_VECTORDIFFTHRES may be predefined in an encoding device and a decoding device. Alternatively, information indicating MULTIPREDMODE_VECTORDIFFTHRES may be explicitly encoded and signaled.
[0921] As an example, when a directional intra prediction mode is applied to each of the current block (or sub-block) and the surrounding blocks, it may be determined whether to perform multi-prediction based on a result of comparing the difference between the intra prediction modes of the current block (or sub-block) and the surrounding blocks with a threshold. As an example, multi-prediction may be performed only when the difference between the intra prediction modes of the current block (or sub-block) and the surrounding blocks is equal to or greater than the threshold. For example, multipred_flag may be encoded / decoded only when the difference between the intra prediction modes of the current block (or sub-block) and the surrounding blocks is equal to or greater than MULTIPREDMODE_INTRAMODETHRES. Alternatively, whether to perform multi-prediction may be true when the difference between the intra prediction modes of the current block (or sub-block) and the surrounding blocks is equal to or greater than MULTIPREDMODE_INTRAMODETHRES. In this case, MULTIPREDMODE_INTRAMODETHRES may be an integer equal to or greater than 0. MULTIPREDMODE_INTRAMODETHRES may be predefined in the encoding device and the decoding device. Alternatively, information indicating MULTIPREDMODE_INTRAMODETHRES may be explicitly encoded and signaled.
[0922] As an example, multi-prediction may be performed only when the difference between the intra-prediction modes of the current block (or sub-block) and the surrounding blocks is less than or equal to a threshold. For example, multipred_flag may be encoded / decoded only when the difference between the intra-prediction modes of the current block (or sub-block) and the surrounding blocks is less than or equal to MULTIPRED_INTRAMODETHRES. Alternatively, whether to perform multi-prediction may be true when the difference between the intra-prediction modes of the current block (or sub-block) and the surrounding blocks is less than or equal to MULTIPRED_INTRAMODETHRES. In this case, MULTIPRED_INTRAMODETHRES may be an integer equal to or greater than 0. MULTIPRED_INTRAMODETHRES may be predefined in the encoding device and the decoding device. Alternatively, information indicating MULTIPRED_INTRAMODETHRES may be explicitly encoded and signaled.
[0923] According to an embodiment of the present disclosure, whether to allow multi-prediction may be determined based on prediction information of surrounding blocks.
[0924] As an example, when an intra prediction mode is used for at least one of the surrounding blocks referenced in order to derive the coding parameters of the current block (or subblock), multi-prediction may be applied to the current block (or subblock). Here, the surrounding block may be a surrounding block adjacent to the current block (or subblock) in a specific direction or any one of the surrounding blocks among the plurality of surrounding blocks.
[0925] Fig.25 Surrounding blocks adjacent to the current block are shown.
[0926] The surrounding blocks can be Fig.25 , or one of A0, A1, B0, B1, B2, C3, or H shown in . Alternatively, whether to perform multi-prediction may be determined by referring to surrounding blocks at positions different from the positions shown.
[0927] According to an embodiment of the present disclosure, whether to perform multi-prediction or whether to encode / decode multipred_flag may be determined based on at least one sample in a prediction block of a current block (or subblock), at least one sample in a reference block of the current block (or subblock), at least one sample in a surrounding block adjacent to the current block (or subblock), or at least one of a prediction mode, motion information, encoding parameters, or size of the surrounding blocks.
[0928] As an example, multi-prediction may be performed or multipred_flag may be encoded / decoded only when the difference between the maximum value and the minimum value among the prediction samples generated by predicting the current block (or subblock) is less than or equal to the threshold value. In other words, a prediction block other than the first prediction block may be additionally generated only when the difference between the maximum value and the minimum value of the prediction samples in the first prediction block generated by predicting the current block (or subblock) is less than or equal to the threshold value.
[0929] As an example, multi-prediction may be performed or multipred_flag may be encoded / decoded only when the difference between the maximum value and the minimum value in the prediction sample of the current block (or subblock) is equal to or greater than the threshold value. In other words, only when the difference between the maximum value and the minimum value of the prediction sample in the first prediction block generated by predicting the current block (or subblock) is equal to or greater than the threshold value, a prediction block other than the first prediction block may be additionally generated.
[0930] As an example, multi-prediction may be performed or multipred_flag may be encoded / decoded only when the difference between the maximum value and the minimum value in the sample of the reference block belonging to the current block (or subblock) is less than or equal to the threshold value. In other words, a prediction block other than the first prediction block obtained from the reference block may be additionally generated only when the difference between the maximum value and the minimum value of the prediction sample in the reference block specified by the prediction information for the current block (or subblock) is less than or equal to the threshold value.
[0931] As an example, multi-prediction may be performed or multipred_flag may be encoded / decoded only when the difference between the maximum value and the minimum value in the sample of the reference block belonging to the current block (or subblock) is equal to or greater than the threshold value. In other words, a prediction block other than the first prediction block obtained from the reference block may be additionally generated only when the difference between the maximum value and the minimum value of the prediction sample in the reference block specified by the prediction information for the current block (or subblock) is equal to or greater than the threshold value.
[0932] As an example, when there are multiple reference blocks (for example, L0 reference block and L1 reference block) for the current block (or subblock), for all reference blocks, multi-prediction can be performed for the current block (or subblock) or multipred_flag can be encoded / decoded only when the condition that the difference between the maximum value and the minimum value in the reference block is equal to or greater than the threshold is met.
[0933] As an example, when there are multiple reference blocks (for example, L0 reference block and L1 reference block) for the current block (or subblock), for all reference blocks, multi-prediction can be performed for the current block (or subblock) or multipred_flag can be encoded / decoded only when the condition that the difference between the maximum value and the minimum value in the reference block is less than or equal to the threshold is met.
[0934] In the above example, the threshold value may be a positive integer. The threshold value may be predefined in the encoding device and the decoding device. Optionally, information representing the threshold value may be encoded and transmitted with a signal. Optionally, the threshold value may be adaptively determined based on at least one of a prediction mode of the current block, an encoding parameter of the current block, a height of the current block, a width of the current block, or a number of samples in the current block.
[0935] As an example, when surrounding blocks of a current block (or subblock) are not in a predefined prediction mode, multi-prediction may be performed on the current block (or subblock) or multipred_flag may be encoded / decoded.
[0936] As an example, when surrounding blocks of a current block (or subblock) are in a predefined prediction mode, multi-prediction may be performed on the current block (or subblock) or multipred_flag may be encoded / decoded.
[0937] In the above example, the predefined prediction mode may be the intra block copy mode.
[0938] According to an embodiment of the present disclosure, after obtaining the first prediction block of the current block, it may be determined whether to perform multi-prediction based on the characteristics of the first prediction block. Here, the first prediction block may be generated by inter-frame prediction, intra-frame prediction, or intra-frame block copying. In addition, when the prediction direction indicator of the current block indicates bidirectional prediction, the first prediction block may be obtained by performing a weighted sum on the L0 prediction block and the L1 prediction block. When CIIP is applied to the current block, the first prediction block may be obtained by performing a weighted sum on the prediction block generated by inter-frame prediction and the prediction block generated by intra-frame prediction.
[0939] In addition, a characteristic shows the number of main gradients. The main gradients can be derived from a gradient histogram obtained based on the gradients of the prediction samples in the first prediction block. As an example, when the magnitude value in the gradient histogram is greater than a threshold, it can be selected as the main gradient.
[0940] By comparing the number of main gradients with a threshold, it can be determined whether to perform multi-prediction for the current block. As an example, multi-prediction can be performed for the current block only when the number of main gradients is greater than or less than the threshold.
[0941] Alternatively, whether to perform multi-prediction may be determined based on characteristics of a reference block used to obtain the first prediction block. Specifically, when the reference block is determined by a block vector or motion information of the current block, whether to perform multi-prediction for the current block may be determined based on the number of main gradients of the determined reference block. The number of main gradients of the reference block may be calculated from a gradient histogram obtained based on slopes of samples in the first reference block.
[0942] In addition, when the inter prediction indicator of the current block indicates bi-prediction, whether to perform multi-prediction may be determined based on the characteristics of one of the L0 reference block and the L1 reference block. Alternatively, multi-prediction may be set to be performed when both the L0 reference block and the L1 reference block satisfy the condition for performing multi-prediction (i.e., when the number of main gradients is greater than or less than a threshold value). Alternatively, multi-prediction may be set to be performed when only any one of the L0 reference block and the L1 reference block satisfies the condition for performing multi-prediction.
[0943] Alternatively, when the prediction direction indicator of the current block indicates bi-prediction, whether to perform multi-prediction may be determined based on characteristics of at least one of the L0 prediction block or the L1 prediction block. Alternatively, multi-prediction may be set to be performed when both the L0 prediction block and the L1 prediction block satisfy a condition for performing multi-prediction (i.e., when the number of main gradients is greater than or less than a threshold value). Alternatively, multi-prediction may be set to be performed when only any one of the L0 prediction block and the L1 prediction block satisfies the condition for performing multi-prediction.
[0944] When CIIP is applied to the current block, whether to perform multi-prediction may be determined based on characteristics of at least one of an intra-prediction block obtained by intra-prediction and an inter-prediction block obtained by inter-prediction. Alternatively, multi-prediction may be set to be performed when both the intra-prediction block and the inter-prediction block satisfy conditions for performing multi-prediction. Alternatively, multi-prediction may be set to be performed when only any one of the intra-prediction block and the inter-prediction block satisfies the conditions for performing multi-prediction.
[0945] In addition, when multi-prediction is applied, multiple prediction blocks may be generated for the current block or subblock. As an example, when the number of prediction blocks is MULTIPRED_NUM, MULTIPRED_NUM may be 1, 2, or a positive integer greater than 2. MULTIPRED_NUM may be predefined in the encoding device and the decoding device.
[0946] The prediction mode and encoding parameters for each of the plurality of prediction blocks may be the same or different.Here, the prediction mode may be one of intra prediction, inter prediction or intra block copy mode.
[0947] Hereinafter, for convenience of description, each prediction block generated by multi-prediction is referred to as an i-th prediction block (i is a positive integer having a value from 1 to MULTIPRED_NUM). However, i may not represent an order of at least one of encoding, decoding, prediction, reconstruction, or weighted sum for each prediction block. i is a positive integer having a value from 1 to MULTIPRED_NUM.
[0948] When performing multi-prediction, information about a prediction mode used to obtain at least one prediction block among a plurality of prediction blocks may be encoded and signaled. As an example, an indicator indicating whether a specific prediction mode is used or available may be encoded and signaled.
[0949] For at least one prediction mode, the indicator may be encoded and signaled. As an example, the indicator may be encoded and signaled for at least one of intra prediction, inter prediction, intra block copy, unidirectional intra prediction, or bidirectional inter prediction. As an example, the indicator of the intra prediction mode may indicate whether intra prediction is performed when multi prediction is performed. Through encoding / decoding of the indicator, intra prediction may be determined to be not used in multi prediction for the current block or may be determined to be used in multi prediction for the current block.
[0950] When the indicator indicating whether a specific prediction mode is used or available is a first value, it may indicate that when multi-prediction is applied, all prediction blocks are predicted according to the specific prediction mode or at least one prediction block is predicted according to the specific prediction mode. The first value may be true or 1.
[0951] When the indicator indicating whether a specific prediction mode is used or available is a second value, it may indicate that all prediction blocks are not predicted according to the specific prediction mode when multi-prediction is applied. The second value may be false or zero.
[0952] For each prediction block, an indicator may be encoded and signaled.The indicator may specify a prediction mode used to obtain the prediction block corresponding to the indicator.
[0953] Alternatively, whether at least one prediction mode is used or available may be determined from an indicator.
[0954] As an example, when the value of the indicator is a first value, it means that when multi-prediction is performed for the current block, at least two of intra prediction, inter prediction, and intra block copy are used or available. As an example, when the value of the indicator is a second value, it means that when multi-prediction is performed for the current block, only intra block copy is available.
[0955] Here, one of the first value and the second value may be 0 and the other may be an integer greater than 0.
[0956] When multi-prediction is performed, whether a specific prediction mode will be used or whether a specific prediction mode is available may be determined according to predefined conditions.
[0957] The predefined condition may be related to at least one of the encoding parameters or the block information. As an example, the predefined condition may be related to at least one of the following items: the size of the current block, the size of the block vector, the size of the motion vector, the direction of the inter prediction, the surrounding block information, the index of the bidirectional prediction (BCW) with CU weight, the index of the adaptive motion vector resolution (AMVR), the indicator of the local illumination compensation (LIC), the indicator of the combined inter-intra prediction (CIIP), the indicator of whether to perform the position-dependent prediction combination (PDPC) and / or the indicator of the affine mode or the directional intra prediction mode. But not limited to this.
[0958] According to the prediction mode, a condition for determining whether the corresponding prediction mode is used or available may be set. In other words, the condition for determining whether the first prediction mode is used or available may be different from the condition for determining whether the second prediction mode is used or available.
[0959] Alternatively, whether each of the plurality of prediction modes is used or available may be determined under the same condition.
[0960] When multi-prediction is applied, encoding parameters for at least one of a plurality of prediction blocks may be encoded / decoded.
[0961] Based on the information used to obtain a specific prediction block, it may be determined whether the corresponding prediction mode is used or available for multi-prediction, ie, another prediction block.
[0962] As an example, whether inter prediction is used or available in multi-prediction may be determined by comparing the difference between the motion vector of the current block and the motion vector of the surrounding blocks with a threshold. As an example, when the difference between the motion vector of the current block and the motion vector of the surrounding blocks is equal to or greater than MULTIPREDSPECIFIC_VECTORDIFFTHRES, inter prediction may be applied in multi-prediction. Alternatively, when the difference between the motion vector of the current block and the motion vector of the surrounding blocks is equal to or greater than MULTIPREDSPECIFIC_VECTORDIFFTHRES, an indicator indicating whether inter prediction is used in multi-prediction may be encoded / decoded.
[0963] As an example, whether intra block copying is used or available in multi-prediction may be determined by comparing the difference between the block vector of the current block and the block vectors of the surrounding blocks with a threshold. As an example, when the difference between the block vector of the current block and the block vectors of the surrounding blocks is equal to or greater than MULTIPREDSPECIFIC_VECTORDIFFTHRES, intra block copying may be applied in multi-prediction. Alternatively, when the difference between the block vector of the current block and the block vectors of the surrounding blocks is equal to or greater than MULTIPREDSPECIFIC_VECTORDIFFTHRES, an indicator indicating whether intra block copying is used in multi-prediction may be encoded / decoded.
[0964] In the above example, MULTIPREDSPECIFIC_VECTORDIFFTHRES may be a positive integer equal to or greater than 0. MULTIPREDSPECIFIC_VECTORDIFFTHRES may be predefined in the encoding device and the decoding device. Optionally, information representing MULTIPREDSPECIFIC_VECTORDIFFTHRES may be encoded and transmitted with a signal. MULTIPREDSPECIFIC_VECTORDIFFTHRES may be set to a different value for each prediction mode or may be set equally for multiple prediction modes.
[0965] As an example, whether inter prediction is used or available in multi-prediction may be determined by comparing the motion vector resolution of the current block with a value. For example, when the motion vector resolution of the current block is equal to or greater than MULTIPREDSPECIFIC_AMVRTHRES, inter prediction may be applied in multi-prediction. Alternatively, when the motion vector resolution of the current block is equal to or greater than MULTIPREDSPECIFIC_AMVRTHRES, an indicator indicating whether inter prediction is used in multi-prediction may be encoded / decoded.
[0966] For example, when the motion vector resolution of the current block is less than or equal to MULTIPREDSPECIFIC_AMVRTHRES, inter-frame prediction may be applied in multi-prediction. Optionally, when the motion vector resolution of the current block is less than or equal to MULTIPREDSPECIFIC_AMVRTHRES, an indicator indicating whether inter-frame prediction is used in multi-prediction may be encoded / decoded.
[0967] As an example, whether intra prediction is used or available in multi-prediction may be determined by comparing the difference between the directional intra prediction mode of the current block and the directional intra prediction mode of the surrounding blocks with a threshold. As an example, when the difference between the directional intra prediction mode of the current block and the directional intra prediction mode of the surrounding blocks is equal to or greater than MULTIPREDSPECIFIC_INTRAMODETHRES, intra prediction may be applied in multi-prediction. Alternatively, when the difference between the directional intra prediction mode of the current block and the directional intra prediction mode of the surrounding blocks is equal to or greater than MULTIPREDSPECIFIC_INTRAMODETHRES, an indicator indicating whether intra prediction is used in multi-prediction may be encoded / decoded.
[0968] As an example, when the difference between the directional intra prediction mode of the current block and the directional intra prediction mode of the surrounding blocks is less than or equal to MULTIPREDSPECIFIC_INTRAMODETHRES, intra prediction may be applied in multi-prediction. Optionally, when the difference between the directional intra prediction mode of the current block and the directional intra prediction mode of the surrounding blocks is less than or equal to MULTIPREDSPECIFIC_INTRAMODETHRES, an indicator indicating whether intra prediction is used in multi-prediction may be encoded / decoded.
[0969] In the above example, MULTIPREDSPECIFIC_INTRAMODETHRES may be an integer equal to or greater than 0. MULTIPREDSPECIFIC_INTRAMODETHRES may be predefined in the encoding device and the decoding device. Optionally, information indicating MULTIPREDSPECIFIC_INTRAMODETHRES may be encoded and signaled. Optionally, MULTIPREDSPECIFIC_INTRAMODETHRES may be adaptively determined according to the directional intra prediction mode of the current block (or subblock).
[0970] An indicator indicating whether a prediction mode is used or available in multi-prediction may be encoded and signaled.
[0971] As an example, when the prediction mode of the current block is not the specific prediction mode, the specific prediction mode may be available in multi-prediction. Here, the specific prediction mode may be inter-frame prediction, intra-frame prediction, or intra-frame block copy. As an example, when a first prediction block of the current block is generated in a prediction mode different from inter-frame prediction, a second prediction block may be generated based on inter-frame prediction. To this end, when the prediction mode of the current block is not inter-frame prediction, an indicator indicating whether inter-frame prediction is applied in multi-prediction or whether inter-frame prediction is available in multi-prediction may be encoded and signaled. Optionally, when the prediction mode of the current block is not inter-frame prediction, inter-frame prediction may be set by default to be available in multi-prediction.
[0972] As an example, when the CIIP is not applied to the current block, the CIIP may be available in multi-prediction. As an example, when a first prediction block ...
Claims
1. A method for decoding an image, the method comprising: Generate a first prediction block of the current block; generating a second prediction block of the current block or a subblock within the current block; as well as generating a final prediction block of the sub-block or the current block by a weighted sum of the first prediction block and the second prediction block, The prediction information used to generate the first prediction block is different from the prediction information used to generate the second prediction block.
2. The method according to claim 1, wherein: The second prediction block is obtained based on prediction information of surrounding blocks located outside the current block and adjacent to the subblock.
3. The method according to claim 1, wherein: When the difference between the maximum value and the minimum value of the prediction samples within the first prediction block is less than a threshold value, the generation of the second prediction block is allowed.
4. The method according to claim 1, wherein: determining a number of dominant gradients from a gradient histogram generated based on the gradients of each prediction sample in the first prediction block, When the number of main gradients is less than a threshold, generation of a second prediction block is allowed.
5. The method according to claim 1, wherein: The first prediction block is generated based on the first intra prediction mode, The second prediction block is generated based on a second intra prediction mode different from the first intra prediction mode.
6. The method according to claim 5, wherein: The second intra prediction mode is used to decode surrounding blocks adjacent to the subblock.
7. The method according to claim 1, wherein: When a plurality of prediction blocks other than the first prediction block are additionally generated, the weighted sum operation is sequentially performed with respect to the plurality of prediction blocks.
8. The method according to claim 7, wherein: An order in which the weighted sum operation is performed for the plurality of prediction blocks is determined based on priorities of surrounding blocks from which prediction information for each of the plurality of prediction blocks is derived.
9. The method according to claim 7, wherein: An order in which the weighted sum operation is performed for the plurality of prediction blocks is determined according to a prediction mode for each of the plurality of prediction blocks.
10. The method of claim 1, wherein: Weights applied to a first prediction block and a second prediction block are determined by comparing a first matching cost obtained by applying first prediction information for the first prediction block to a surrounding template of the sub-block with a second matching cost obtained by applying second prediction information for the second prediction block to the template.
11. The method according to claim 10, wherein: The weight is determined by additionally considering a third matching cost obtained by performing a weighted sum on the first matching cost and the second matching cost.
12. The method according to claim 10, wherein: A first matching cost and a second matching cost are obtained by using the sub-sampling positions within the template.
13. The method of claim 1, wherein: Additional prediction blocks are obtained only for sub-blocks adjacent to the boundary within the current block.
14. A method for encoding an image, the method comprising: Generate a first prediction block of the current block; generating a second prediction block of the current block or a subblock within the current block; as well as generating a final prediction block of the sub-block or the current block by a weighted sum of the first prediction block and the second prediction block, The prediction information used to generate the first prediction block is different from the prediction information used to generate the second prediction block.
15. A computer-readable recording medium, the recording medium comprising: Generate a first prediction block of the current block; generating a second prediction block of the current block or a subblock within the current block; as well as generating a final prediction block of the sub-block or the current block by a weighted sum of the first prediction block and the second prediction block, The prediction information used to generate the first prediction block is different from the prediction information used to generate the second prediction block.
