KR20210070942A - Method, apparatus and recording medium for region differential image encoding/decoding - Google Patents

Abstract

The present invention relates to a decoding method of a video, a decoding device, an encoding method, and an encoding device. Disclosed are a method and a device for performing region differential image encoding/decoding using a reconstructed image. According to the encoding method of the embodiment, a reconstructed low-quality image is generated by encoding the original image, and a reconstructed high-quality image is generated using the reconstructed low-quality image. The image is divided into a plurality of regions, and encoded reconstruction information for generating the reconstructed high-quality image is generated by performing encoding on the region.

Description

Method, apparatus and recording medium for region differential image encoding/decoding {METHOD, APPARATUS AND RECORDING MEDIUM FOR REGION DIFFERENTIAL IMAGE ENCODING/DECODING}

The present invention relates to a method, apparatus and recording medium for video encoding/decoding. Specifically, the present invention relates to a method, an apparatus, and a recording medium for providing region differential image encoding/decoding using a reconstructed image.

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

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

As an image compression technique, various techniques, such as an inter prediction technique, an intra prediction technique, a transform and quantization technique, and an entropy encoding technique, exist.

The inter prediction technique is a technique of predicting a value of a pixel included in a current picture by using a picture before and/or after a picture of the current picture. The intra prediction technique is a technique of predicting the value of a pixel included in the current picture by using information about the pixel in the current picture. Transformation and quantization techniques are techniques for compressing the energy of the residual image. The entropy encoding technique is a technique in which a short code is assigned to a value having a high frequency of occurrence and a long code is assigned to a value having a low frequency of occurrence.

Using this image compression technique, data for an image can be effectively compressed, transmitted, and stored.

A technology for reconstructing a super-resolution image is a technology for reconstructing a high-quality image using a low-resolution image. A machine learning method may be applied to the reconstruction of an ultra-high-resolution image.

An embodiment may provide a method and an apparatus for performing encoding and decoding using content of an image.

An embodiment may provide a method and apparatus for dividing an image into a plurality of regions according to features of the image and semantic analysis of the image, and then performing region differential encoding and decoding on each of the divided regions. have.

An embodiment may provide a method and apparatus for dividing an image into regions and then performing encoding and decoding using a method adapted to the characteristics of each region of the divided regions.

An embodiment may provide a method and an apparatus for selecting a super-resolution reconstructed image as a prediction image through various methods in order to improve encoding efficiencies for regions of the image.

An embodiment provides a method and apparatus for performing distortion-reducing processing on a low-bit-rate image having high encoding distortion in order to improve encoding efficiencies for regions of the image, and using the image with reduced distortion as a prediction image. can provide

An embodiment may provide a method and an apparatus for encoding and decoding a prediction image in block units.

In one aspect, the method comprising: generating a reconstructed low-quality image by performing encoding on the original image; generating a reconstructed high-quality image by using the reconstructed low-quality image; dividing the image into a plurality of regions; generating encoded reconstruction information by performing encoding on the region; and generating a bitstream including encoded reconstruction information, wherein the region is at least one of the plurality of regions, and the image is the original image, the restored low-quality image, or the restored high-quality image. An encoding method is provided.

The bitstream may include an encoded low-quality image.

The encoded low-quality image may be an encoded low-resolution image or an encoded low-bit rate image.

The bitstream may include type information indicating which image of the encoded low-resolution image and the encoded low-bit-rate image is included in the bitstream.

If the encoded low-quality image is the encoded low-resolution image, the reconstructed high-quality image may be a reconstructed high-resolution image.

If the encoded low-quality image is the encoded low-bit rate image, the reconstructed high-quality image may be a coding distortion correction image.

By performing classification on the image, a map for the image may be generated.

The map may indicate a plurality of regions.

Different encoding methods may be used for the plurality of regions, respectively.

The edge power of the image may be calculated, and the map may be modified based on the edge power.

The map may be modified based on the semantic classification in the image.

If the encoded low-quality image is an encoded low-resolution image, differential restoration from a low resolution to a high resolution may be performed on the image based on the map.

If the encoded low-quality image is an encoded low-bit rate image, differential correction may be performed on coding distortion generated in the image based on the map.

Encoding may be performed on each of the plurality of regions using different neural networks.

In another aspect, there is provided a recording medium for recording the bitstream generated by the encoding method.

In another aspect, the method comprising: receiving a bitstream including an encoded low-quality image; generating a reconstructed low-quality image by performing decoding on the encoded low-quality image; generating a reconstructed high-quality image by using the reconstructed low-quality image; dividing the image into a plurality of regions; generating reconstruction information by performing decoding on the region; and generating a reconstructed image using the reconstruction information, wherein the area is at least one of the plurality of areas, and the image is the reconstructed low-quality image or the reconstructed high-quality image. A method is provided.

The encoded low-quality image may be an encoded low-resolution image or an encoded low-bit rate image.

The bitstream may include type information indicating which image of the encoded low-resolution image and the encoded low-bit-rate image is included in the bitstream.

If the encoded low-quality image is the encoded low-resolution image, the reconstructed high-quality image may be a reconstructed high-resolution image.

If the encoded low-quality image is the encoded low-bit rate image, the reconstructed high-quality image may be a coding distortion correction image.

By performing classification on the image, a map for the image may be generated.

The map may indicate a plurality of regions.

Different decoding methods may be used for a plurality of regions, respectively.

The edge power of the image may be calculated, and the map may be modified based on the edge power.

The map may be modified based on the semantic classification in the image.

If the encoded low-quality image is an encoded low-resolution image, differential restoration from a low resolution to a high resolution may be performed on the image based on the map.

If the encoded low-quality image is an encoded low-bit rate image, differential correction may be performed on coding distortion generated in the image based on the map.

Decoding may be performed on each of the plurality of regions using different neural networks.

In another aspect, in a computer-readable recording medium for storing a bitstream, the bitstream includes an encoded low-quality image, and a reconstructed low-quality image is generated by performing decoding on the encoded low-quality image A high-quality image reconstructed using the reconstructed low-quality image is generated, the image is divided into a plurality of regions, and reconstruction information is generated by performing decoding on the region, and reconstruction using the reconstruction information A constructed image is generated, the area is at least one of the plurality of areas, and the image is the reconstructed low-quality image or the reconstructed high-quality image.

A method and apparatus for performing encoding and decoding using content of an image are provided.

A method and apparatus are provided for dividing an image into a plurality of regions according to features of the image and semantic analysis of the image, and then performing region differential encoding and decoding on each of the divided regions.

A method and apparatus are provided for segmenting an image into regions and performing encoding and decoding using a method adapted to the characteristics of each region of the divided regions.

A method and apparatus are provided for selecting an ultra-high-resolution reconstructed image as a prediction image through various methods in order to improve encoding efficiencies for regions of the image.

A method and apparatus for performing distortion-reducing processing on a low-bit-rate image having high encoding distortion in order to improve encoding efficiencies for regions of an image and using the reduced-distortion image as a prediction image are provided.

A method and apparatus for encoding and decoding a prediction image in block units are provided.

1 is a block diagram illustrating a configuration of an encoding apparatus to which the present invention is applied according to an embodiment.
2 is a block diagram illustrating a configuration of a decoding apparatus to which the present invention is applied according to an embodiment.
FIG. 3 is a diagram schematically illustrating a split structure of an image when encoding and decoding an image.
4 is a diagram illustrating a form of a prediction unit that a coding unit may include.
5 is a diagram illustrating a form of a transform unit that may be included in a coding unit.
6 illustrates the division of a block according to an example.
7 is a diagram for explaining an embodiment of an intra prediction process.
8 is a diagram for describing a reference sample used in an intra prediction process.
9 is a diagram for explaining an embodiment of an inter prediction process.
10 illustrates spatial candidates according to an example.
11 illustrates an order of adding motion information of spatial candidates to a merge list according to an example.
12 illustrates a process of transformation and quantization according to an example.
13 illustrates diagonal scanning according to an example.
14 illustrates horizontal scanning according to an example.
15 illustrates vertical scanning according to an example.
16 is a structural diagram of an encoding apparatus according to an embodiment.
17 is a structural diagram of a decoding apparatus according to an embodiment.
18 is a flowchart of an encoding method according to an embodiment.
19 is a flowchart of a method for generating coded information using a map according to an embodiment.
20 is a flowchart of a decoding method according to an embodiment.
21 is a flowchart of a method of generating decrypted information using a map according to an embodiment.
22 illustrates a processing process of an image encoding method using a reconstructed super-resolution image according to an example.
23 illustrates a processing process of an image decoding method using an ultra-high-resolution reconstructed image according to an example.
24 shows an original image used for semantic classification according to an example.
25 shows a result of semantic classification according to an example.
26 shows an original image in classification of a texture region according to an example.
27 shows a result of classification of a texture area according to an example.
28 illustrates generation of a map using a neural network according to an example.
29 illustrates generation of a region differential super-resolution reconstructed image according to an exemplary embodiment.
30 and 31 illustrate degradation of image quality in a low bit rate environment according to an example.
30 shows a reconstructed image of HEVC qp37 according to an example.
31 shows a reconstructed high-quality image of a low-resolution HEVC qp37 according to an example.
32 illustrates a region in which an edge power exceeding a threshold value occurs when a reconstructed high-quality image is generated using a low-resolution image with a low bit rate according to an example.
33, 34, 35, and 36 show images for classifying a texture region according to an example.
33 shows an original image according to an example.
34 shows a high-resolution image reconstructed by SR reconstruction according to an example.
35 illustrates edge power of a reconstructed high-resolution image according to an example.
36 illustrates a result of division of regions for a reconstructed high-resolution image according to an example.
37 is a flowchart of generation and use of a map for classification of regions according to an embodiment.
38 shows a syntax for signaling of a type of a base bitstream according to an example.
39 shows signaling information for coding of SR copy and SR prediction according to an example.
40 shows the structure of a decoding apparatus according to an embodiment.
41 is a flowchart of a decoding method using a base bitstream according to an example.
42 is a flowchart of a decoding method for a low-resolution bitstream according to an example.
43 is a flowchart of a decoding method for a low bit rate bitstream according to an example.

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

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] Reference is made to the accompanying drawings, which illustrate specific embodiments by way of example. These embodiments are described in sufficient detail to enable those skilled in the art to practice the embodiments. It should be understood that various embodiments are different but need not be mutually exclusive. For example, certain shapes, structures, and characteristics described herein with respect to one embodiment may be embodied in other embodiments without departing from the spirit and scope of the invention. In addition, it should be understood that the location or arrangement of individual components within each disclosed embodiment may be changed without departing from the spirit and scope of the embodiment. Accordingly, the detailed description set forth below is not intended to be taken in a limiting sense, and the scope of exemplary embodiments, if properly described, is limited only by the appended claims, along with all scope equivalents to those claimed.

Like reference numerals in the drawings refer to the same or similar functions throughout the various aspects. Shapes and sizes of elements in the drawings may be exaggerated for clearer description.

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

When a component is referred to as being “connected” or “connected” to another component, the two components may be directly connected or connected to each other, but in the above 2 It should be understood that other components may exist in the middle of the components. On the other hand, when it is mentioned that a component is "directly connected" or "directly connected" to another component, it should be understood that no other component is present in the middle of the two components. something to do.

The components shown in the embodiment of the present invention are shown independently to represent different characteristic functions, and it does not mean that each component is composed of separate hardware or a single software component. That is, each component is included by listing as each component for convenience of description, and at least two components of each component are combined to form one component, or one component is divided into a plurality of components to provide a function. As long as it can be carried out and integrated embodiments and separate embodiments of each of these components do not depart from the essence of the present invention, they are included in the scope of the present invention.

In addition, the description of "including" a specific configuration in the exemplary embodiments does not exclude the configuration other than the above specific configuration, and the additional configuration is the implementation of the exemplary embodiments or the technical spirit of the exemplary embodiments. It means that it can be included in the scope.

The terms used in the present invention are only used to describe specific embodiments, and are not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise. In the present invention, terms such as "comprises" or "have" are intended to designate that the features, numbers, steps, operations, components, parts, or combinations thereof described in the specification exist, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof. That is, the description of "including" a specific configuration in the present invention does not exclude configurations other than the corresponding configuration, and additional configurations may also be included in the practice of the present invention or the scope of the technical spirit of the present invention.

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

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

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

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

Hereinafter, the target image may be an encoding target image to be encoded and/or a decoding target image to be decoded. In addition, the target image may be an input image input to the encoding apparatus or may be an input image input to the decoding apparatus. Also, the target image may be a current image that is a target of current encoding and/or decoding. For example, the terms "target image" and "current image" may be used interchangeably and may be used interchangeably.

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

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

Hereinafter, the terms “block” and “unit” may be used with the same meaning, and may be used interchangeably. Or “block” may indicate a specific unit.

Hereinafter, the terms “region” and “segment” may be used interchangeably.

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

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

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

In the following, terms used in the embodiments are described.

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

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

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

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

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

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

- Depending on the function, the type of unit is a macro unit, a coding unit (CU), a prediction unit (PU), a residual unit, and a transform unit (TU), etc. can be classified as Or, depending on the function, the unit is a block, a macroblock (Macroblock), a coding tree unit (Coding Tree Unit), a coding tree block (Coding Tree Block), a coding unit (Coding Unit), a coding block (Coding Block), a prediction unit (Prediction Unit), a prediction block (Prediction Block), a residual unit (Residual Unit), a residual block (Residual Block), a transform unit (Transform Unit), a transform block (Transform Block), etc. may mean. For example, the target unit may be at least one of a CU, a PU, a residual unit, and a TU that are encoding and/or decoding targets.

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

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

- Also, the unit information may include at least one of a unit type, a unit size, a unit depth, an encoding order of a unit, a decoding order of a unit, and the like. For example, the type of unit may indicate one of CU, PU, residual unit and TU, and the like.

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

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

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

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

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

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

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

- A node having a depth of level n may represent a unit generated as the original unit is divided n times.

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

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

Sample (sample): A sample may be a base unit constituting a block. A sample may be expressed as values ^{from 0 to 2 Bd} −1 depending on the bit depth (Bd).

- Samples can be pixels or pixel values.

- Hereinafter, the terms "pixel", "pixel" and "sample" may be used interchangeably and may be used interchangeably.

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

- Each coding tree unit includes a quad tree (QT), a binary tree (BT), and a ternary tree (TT) to configure subunits such as a coding unit, a prediction unit, and a transform unit. It may be partitioned using one or more partitioning schemes. The quad tree may mean a quarternary tree. In addition, each coding tree unit may be partitioned using a MultiType Tree (MTT) using one or more partitioning schemes.

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

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

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

- Hereinafter, the terms "neighboring block" and "adjacent block" may be used with the same meaning, and may be used interchangeably.

- The neighbor block may mean a reconstructed neighbor block.

Spatial neighbor block: A spatial neighbor block may be a block spatially adjacent to a target block. The neighboring blocks may include spatial neighboring blocks.

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

- The spatial neighboring block may mean a block having a boundary abutting on the target block or a block located within a predetermined distance from the target block.

- The spatial neighboring block may mean a block adjacent to the vertex of the target block. Here, the block adjacent to the vertex of the target block may be a block vertically adjacent to a neighbor block horizontally adjacent to the target block or a block horizontally adjacent to a neighbor block vertically adjacent to the target block.

Temporal neighbor block: A temporal neighbor block may be a block temporally adjacent to a target block. A neighboring block may include a temporal neighboring block.

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

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

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

Prediction mode: The prediction mode may be information indicating a mode encoded and/or decoded for intra prediction or a mode encoded and/or decoded for inter prediction.

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

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

Prediction unit partition: A prediction unit partition may refer to a form in which a prediction unit is divided.

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

- The reconstructed neighboring unit may be a spatial neighboring unit or a temporal neighboring unit for the target unit.

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

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

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

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

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

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

- A tile may contain one or more CTUs.

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

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

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

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

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

Parameter set: The parameter set may correspond to header information among structures in the bitstream.

- The parameter set includes a video parameter set (VPS), a sequence parameter set (SPS), a picture parameter set (PPS), an adaptation parameter set (APS), and a decoding parameter. It may include at least one of a set (Decoding Parameter Set; DPS).

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

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

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

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

- The rate-distortion optimization method may calculate a rate-distortion cost of each combination in order to select an optimal combination from among the above combinations. The rate-distortion cost can be calculated using the formula "D+λ*R". In general, the combination in which the rate-distortion cost is minimized by the expression "D+λ*R" may be selected as the optimal combination in the rate-distortion optimization method.

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

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

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

- The encoding apparatus may perform processes such as inter prediction, intra prediction, transformation, quantization, entropy encoding, inverse quantization, and/or inverse transformation in order to accurately calculate D and R. These processes may greatly increase the complexity of the encoding apparatus.

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

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

Symbol: It may mean at least one of a syntax element, a coding parameter, and a transform coefficient of an encoding object unit and/or a decoding object unit. Also, the symbol may mean an object of entropy encoding or a result of entropy decoding.

Reference picture: A reference picture may mean an image referenced by a unit for inter prediction or motion compensation. Alternatively, the reference picture may be an image including the reference unit referenced by the target unit for inter prediction or motion compensation.

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

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

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

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

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

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

- An inter prediction indicator may be derived using the prediction list utilization flag. Conversely, a prediction list utilization flag may be derived using the inter prediction indicator. For example, indicating that the prediction list utilization flag has a first value of 0 may indicate that a prediction block is not generated using a reference picture in the reference picture list with respect to the target unit. Indicating 1, which is the second value, of the prediction list utilization flag may indicate that the prediction unit is generated using the reference picture list with respect to the target unit.

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

Picture order count (POC): The POC of a picture may indicate the display order of the picture.

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

- For example, MV may be expressed in the form _{(mv x} , mv _{y ).} mv _x may represent a horizontal component, and mv _y may represent a vertical component.

Search range: The search range may be a two-dimensional area in which an MV is searched during inter prediction. For example, the size of the search area may be MxN. M and N may each be a positive integer.

Motion vector candidate: A motion vector candidate may mean a block that is a prediction candidate or a motion vector of a block that is a prediction candidate when a motion vector is predicted.

- The motion vector candidate may be included in the motion vector candidate list.

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

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

Motion information: Motion information includes motion vectors, reference picture indexes, and inter prediction indicators, as well as reference picture list information, reference pictures, motion vector candidates, motion vector candidate indexes, merge candidates and merge indexes, etc. It may mean information including at least one of

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

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

Merge index: The merge index may be an indicator indicating a merge candidate in the merge candidate list.

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

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

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

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

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

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

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

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

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

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

Quantized level: A quantized level may mean a value generated by performing quantization on a transform coefficient or a residual signal in an encoding apparatus. Alternatively, the quantized level may mean a value that is a target of inverse quantization when the decoding apparatus performs inverse quantization.

- A quantized transform coefficient level resulting from transform and quantization may also be included in the meaning of the quantized level.

Non-zero transform coefficient: The non-zero transform coefficient may mean a transform coefficient having a non-zero value or a transform coefficient level having a non-zero value. Alternatively, the non-zero transform coefficient may mean a transform coefficient having a non-zero value or a transform coefficient level having a non-zero value.

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

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

Default matrix: The default matrix may be a quantization matrix predefined in the encoding apparatus and the decoding apparatus.

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

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

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

- The encoding apparatus and the decoding apparatus may determine one or more MPMs based on the intra prediction mode of the reference block. The reference block may be plural. The plurality of reference blocks may include a spatial neighboring block adjacent to a left side of the target block and a spatial neighboring block adjacent to an upper end of the target block. That is, one or more different MPMs may be determined according to which intra prediction modes are used for the reference blocks.

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

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

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

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

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

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

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

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

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

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

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

The encoding apparatus 100 may be an encoder, a video encoding apparatus, or an image encoding apparatus. A video may include one or more images. The encoding apparatus 100 may sequentially encode one or more images of a video.

Referring to FIG. 1 , the encoding apparatus 100 includes an inter prediction unit 110 , an intra prediction unit 120 , a switch 115 , a subtractor 125 , a transform unit 130 , a quantization unit 140 , and entropy encoding. It may include a unit 150 , an inverse quantizer 160 , an inverse transform unit 170 , an adder 175 , a filter unit 180 , and a reference picture buffer 190 .

The encoding apparatus 100 may encode the target image using the intra mode and/or the inter mode. In other words, the prediction mode for the target block may be one of an intra mode and an inter mode.

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

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

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

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

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

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

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

The inter prediction unit 110 may include a motion prediction unit and a motion compensator.

When the prediction mode is the inter mode, the motion predictor may search for a region that best matches the target block from the reference image in the motion prediction process, and derive a motion vector for the target block and the searched region using the searched region. can do. In this case, the motion prediction unit may use the search area as the area to be searched.

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

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

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

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

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

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

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

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

The plurality of predefined transform methods may include a Discrete Cosine Transform (DCT), a Discrete Sine Transform (DST), and a Karhunen-Loeve Transform (KLT)-based transform. have.

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

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

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

The quantization unit 140 may generate a quantized transform coefficient level (that is, a quantized level or a quantized coefficient) by quantizing the transform coefficient according to the quantization parameter. The quantizer 140 may output the generated quantized transform coefficient level. In this case, the quantization unit 140 may quantize the transform coefficients using a quantization matrix.

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

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

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

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

In order to encode the quantized transform coefficient level, the entropy encoder 150 may change a coefficient of a two-dimensional block form into a form of a one-dimensional vector through a transform coefficient scanning method.

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

A coding parameter may include information (or flags and indexes, etc.) that is encoded by the encoding device and signaled from the encoding device to the decoding device like a syntax element, as well as information derived from the encoding process or decoding process. have. Also, the coding parameter may include information required for encoding or decoding an image. For example, the size of the unit/block, the form of the unit/block, the depth of the unit/block, the division information of the unit/block, the division structure of the unit/block, information indicating whether the unit/block is divided in a quad tree form, Information indicating whether a unit/block is partitioned in the form of a binary tree, the partitioning direction of the binary tree (horizontal or vertical), the partitioning of the binary tree (symmetrical or asymmetrical partitioning), the unit/block is in the form of a ternary tree Information indicating whether or not to be partitioned into a ternary tree type of division direction (horizontal direction or vertical direction), a ternary tree type division type (symmetric division or asymmetric division, etc.), if a unit/block is a multi-type tree Information indicating whether or not the partition is divided in the form of a multi-type tree, the combination and direction of the division (horizontal direction or vertical direction, etc.) Tree-shaped split tree (binary tree or ternary tree), type of prediction mode (intra prediction or inter prediction), intra prediction mode/direction, intra luma prediction mode/direction, intra chroma prediction mode/direction, intra split information, inter split information, coding block split flag, predictive block split flag, transform block split flag, reference sample filtering method, reference sample filter tap, reference sample filter coefficient, predictive block filtering method, predictive block filter tap, predictive block filter coefficient , prediction block boundary filtering method, prediction block boundary filter tap, prediction block boundary filter coefficients, inter prediction mode, motion information, motion vector, motion vector difference, reference picture index, inter prediction direction, inter prediction indicator, prediction list utilization ) flag, reference picture list, reference picture, POC, motion vector predictor, motion vector prediction index, motion vector prediction candidate, motion vector candidate list, information indicating whether to use merge mode, merge index, merge candidate, merge candidate list , whether to use skip mode Information indicating whether or not interpolation filter type, filter tab of interpolation filter, filter coefficients of interpolation filter, motion vector size, motion vector expression accuracy, transform type, transform size, information indicating whether to use first-order transform, add ( Information indicating whether to use a secondary) transform, primary transformation selection information (or primary transformation index), secondary transformation selection information (or secondary transformation index), information indicating the presence or absence of a residual signal, coded A coded block pattern, a coded block flag, a quantization parameter, a residual quantization parameter, a quantization matrix, information about an intra-loop filter, information indicating whether to apply an intra-loop filter, intra- The coefficient of the loop filter, the filter tap of the intra-loop, the shape/form of the intra-loop filter, information indicating whether to apply the deblocking filter, the coefficient of the deblocking filter, the filter tap of the deblocking filter, Deblocking filter strength, shape/form of deblocking filter, information indicating whether adaptive sample offset is applied, adaptive sample offset value, adaptive sample offset category, adaptive sample offset type, adaptive intra-loop ( information indicating whether in-loop) filters are applied, coefficients of adaptive in-loop filters, filter taps of adaptive in-loop filters, shape/shape of adaptive in-loop filters, binarization/inverse binarization methods, context Model, context model determination method, context model update method, information indicating whether to perform regular mode, information indicating whether to perform bypass mode, significant coefficient flag, last significant coefficient flag, coefficient group unit coding flag , last significant coefficient position, flag indicating whether coefficient value is greater than 1, flag indicating whether coefficient value is greater than 2, flag indicating whether coefficient value is greater than 3, remaining coefficient value information, sign ) information, reconstructed luma samples, reconstructed chroma samples, context beans, Bypass bin, residual luma samples, residual chroma samples, transform coefficients, luma transform coefficients, chroma transform coefficients, quantized level, luma quantized level, chroma quantized level, transform coefficient level, luma transform coefficient level, chroma transform coefficient level , the transform coefficient level scanning method, the size of the motion vector search region on the side of the decoding device, the shape of the motion vector search region on the side of the decoding device, the number of motion vector searches on the side of the decoding device, CTU size, minimum block size , maximum block size, maximum block depth, minimum block depth, image display/output order, slice identification information, slice type, slice division information, tile group identification information, tile group type, tile group division information, tile identification information, tile Type, tile partitioning information, picture type, bit depth, input sample bit depth, reconstructed sample bit depth, residual sample bit depth, transform coefficient bit depth, quantized level bit depth, information about luma signal and chroma signal Information, a value of at least one of a color space of a target block and a color space of a residual block, a combined form, or statistics may be included in the coding parameter. In addition, information related to the above-described coding parameters may also be included in the coding parameters. Information used to calculate and/or derive the above-described coding parameters may also be included in the coding parameters. Information calculated or derived using the above-described coding parameters may also be included in the coding parameters.

The prediction method may indicate one of an intra prediction mode and an inter prediction mode.

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

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

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

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

A signal may mean signaled information. Hereinafter, information about an image and a block may be referred to as a signal. Also, hereinafter, the terms “information” and “signal” may be used with the same meaning, and may be used interchangeably. For example, the specific signal may be a signal representing a specific block. The original signal may be a signal representing the target block. A prediction signal may be a signal indicating a prediction block. The residual signal may be a signal representing a residual block.

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

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

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

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

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

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

When the deblocking filter is applied to the target block, the applied filter may vary depending on the required strength of the deblocking filtering. That is, a filter determined according to the strength of deblocking filtering among different filters may be applied to the target block. When a deblocking filter is applied to the target block, a long-tap filter, a strong filter, a weak filter, and a Gaussian filter according to the required strength of deblocking filtering ), one or more filters may be applied to the target block.

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

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

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

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

The reconstructed block or reconstructed image passing through the filter unit 180 may be stored in the reference picture buffer 190 as a reference picture. The reconstructed block passing through 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 reconstructed blocks that have passed through the filter unit 180 . The stored reference picture may then be used for inter prediction or motion compensation.

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

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

Referring to FIG. 2 , the decoding apparatus 200 includes an entropy decoding unit 210 , an inverse quantization unit 220 , an inverse transform unit 230 , an intra prediction unit 240 , an inter prediction unit 250 , and a switch 245 . , an adder 255 , a filter unit 260 , and a reference picture buffer 270 .

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

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

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

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

The entropy decoding unit 210 may generate symbols by performing entropy decoding on the bitstream based on the probability distribution of the bitstream. The generated symbols may include symbols in the form of quantized transform coefficient levels (ie, quantized levels or quantized coefficients). Here, the entropy decoding method may be similar to the entropy encoding method described above. For example, the entropy decoding method may be a reverse process of the entropy encoding method described above.

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

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

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

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

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

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

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

The reconstructed residual block and prediction block may be added via an adder 255 . The adder 255 may generate a reconstructed block by adding the reconstructed residual block and the prediction block.

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

The filter unit 260 may output the reconstructed image.

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

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

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

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

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

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

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

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

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

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

The partition structure may mean a distribution of CUs for efficiently encoding an image in the LCU 310 . Such a distribution may be determined according to whether one CU is divided into a plurality of CUs. The number of partitioned CUs may be a positive integer of 2 or more, including 2, 4, 8, and 16, and the like.

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

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

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

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

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

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

For example, for each depth, an undivided CU may have a size of 2Nx2N. In addition, in the case of a divided CU, a CU having a size of 2Nx2N may be divided into four CUs having a size of NxN. The size of N can be halved for each increase in depth by 1.

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

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

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

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

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

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

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

Through recursive segmentation for CUs, an optimal segmentation method that produces the minimum rate-distortion ratio can be selected.

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

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

For example, quad-tree splitting may be preferentially applied to the CTU. A CU that can no longer be divided into a quad tree may correspond to a leaf node of the quad tree. A CU corresponding to a leaf node of a quad tree may be a root node of a binary tree and/or a ternary tree. That is, a CU corresponding to a leaf node of a quad tree may be split into a binary tree form or a ternary tree form, or may not be split any more. At this time, by applying binary tree splitting or ternary tree splitting to the CU corresponding to the leaf node of the quad tree so that quad tree splitting is not applied again to the CU, the signaling of block splitting and/or block splitting information is can be performed effectively.

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

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

A partition in the case where there is no priority between a binary tree partition and a ternary tree partition may be referred to as a multi-type tree partition. That is, a CU corresponding to a leaf node of a quad tree may be a root node of a multi-type tree. The splitting of a CU corresponding to each node of the multi-type tree may be signaled using at least one of information indicating whether the multi-type tree is split, split direction information, and split tree information. In order to split a CU corresponding to each node of the multi-type tree, information indicating whether to sequentially split, split direction information, and split tree information may be signaled.

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

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

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

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

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

Here, each of the information indicating whether to split, the split tree information, and the split direction information described above may be a flag having a specified length (eg, 1 bit).

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

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

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

A CU that is no longer split by quad tree splitting, binary tree splitting, and/or ternary tree splitting may be a unit of encoding, prediction, and/or transformation. That is, for prediction and/or transformation, the CU may no longer be split. Accordingly, a partition structure and partition information for partitioning a CU into a prediction unit and/or a transform unit may not exist in the bitstream.

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

In this case, information on whether a CU is split for transformation may not be separately signaled. Without signaling, whether the CU is divided may be determined by comparison between the horizontal size (and/or vertical size) of the CU and the horizontal size (and/or vertical size) of the maximum transform block. For example, if the horizontal size of the CU is larger than the horizontal size of the maximum transform block, the CU may be vertically bisected. Also, when the vertical size of the CU is larger than the vertical size of the maximum transform block, the CU may be horizontally divided into two.

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

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

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

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

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

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

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

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

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

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

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

- Horizontal binary tree split for 128xN (N <= 64) CUs

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

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

Or, for a CU corresponding to a node of the multi-type tree, 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, the multi-type tree Information indicating whether to split may be signaled. Otherwise, the CU may not be split into a binary tree form and/or a ternary tree form. According to this determination method, information indicating whether the multi-type tree is split may not be signaled and may be inferred as the second value.

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

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

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

A CU that is no longer split among CUs split from an LCU may be split into one or more prediction units (PUs).

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

A CU may be partitioned in a manner other than that shown in FIG. 5 .

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

As an example, when a CU having a size of 32x32 is vertically divided into three CUs, the sizes of the three divided CUs may be 8x32, 16x32, and 8x32, respectively. As such, when one CU is divided into three CUs, it can be considered that the CU is divided in the form of a ternary tree.

One of the illustrated partitioning in the form of a quad tree, partitioning in the form of a binary tree, and partitioning in the form of a ternary tree may be applied for partitioning of a CU, and a plurality of partitioning methods may be combined together and used for partitioning of the CU . In this case, a case in which a plurality of partitioning methods are combined and used may be referred to as partitioning in the form of a complex tree.

6 illustrates the division of a block according to an example.

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

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

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

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

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

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

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

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

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

[Table 1]

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

[Table 2]

The partitioning method may be limited only to a quad tree, or only to a binary tree, depending on the size and/or shape of the block. When this restriction is applied, split_flag may be a flag indicating whether to split in a quad tree form or a flag indicating whether to split in a binary tree form. The size and shape of the block may be derived according to the depth information of the block, and the depth information may be signaled from the encoding apparatus 100 to the decoding apparatus 200 .

When the size of the block falls within a specified range, only the quad tree type division may be possible. For example, the specified range may be defined by at least one of a maximum block size and a minimum block size that can only be divided in a quad tree form.

Information indicating the maximum block size and/or the minimum block size that can be split only in the form of a quart tree may be signaled from the encoding apparatus 100 to the decoding apparatus 200 through a bitstream. In addition, such information may be signaled for at least one unit of a video, a sequence, a picture, a parameter, a tile group, and a slice (or segment).

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

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

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

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

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

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

The description of the quad-tree type division described above may be equally applied to the binary tree type and/or ternary tree type division.

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

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

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

Arrows outward from the center of the graph of FIG. 7 may indicate prediction directions of directional intra prediction modes. Also, a number indicated adjacent to the arrow may indicate an example of an intra prediction mode or a mode value assigned to a prediction direction of the intra prediction mode.

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

Intra encoding and/or decoding may be performed using reference samples of neighboring units of the target block. The neighboring block may be a reconstructed neighboring block. The reference sample may refer to a neighboring sample.

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

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

The prediction block may mean 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.

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

Alternatively, the prediction block may be a square-shaped block having a size such as 2x2, 4x4, 8x8, 16x16, 32x32 or 64x64, or a rectangular block having a size such as 2x8, 4x8, 2x16, 4x16 and 8x16. have.

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

The intra prediction mode located to the right of the vertical mode may be referred to as a vertical-right mode. The intra prediction mode located at the bottom of the horizontal mode may be called a horizontal-below mode. For example, in FIG. 7 , intra prediction modes whose mode values are one of 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65 and 66 are vertical may be right-hand modes. Intra prediction modes in which the mode value is one of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, and 17 may be horizontal bottom modes.

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

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

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

The number of intra prediction modes and the mode value of each intra prediction mode described above may be merely exemplary. The number of the above-described intra prediction modes and a mode value of each intra prediction mode may be defined differently according to an embodiment, implementation, and/or need.

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

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

The type of filter applied to at least one of the reference sample and the prediction sample may vary according to at least one of an 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 a length of a filter tap, a value of a filter coefficient, and a filter strength. The length of the filter taps may mean the number of filter taps. Also, the number of filter taps may mean the length of the filter.

When the intra prediction mode is the planar mode, in generating the prediction block of the target block, the upper reference sample of the target sample, the left reference sample of the target sample, and the upper right reference sample of the target block according to the position of the prediction target sample in the prediction block and a weight-sum of the lower left reference sample of the target block may be used to generate a sample value of the prediction target sample.

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

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

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

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

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

For example, an indicator indicating 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.

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

8 is a diagram for describing a reference sample used in an intra prediction process.

The reconstructed reference samples used for intra prediction of the target block include lower-left reference samples, left reference samples, upper-left corner reference samples, and above reference samples. and above-right reference samples, and the like.

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

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

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

A reference sample used for intra prediction of the target block may vary depending on the intra prediction mode of the target block. The direction of the intra prediction mode may indicate a dependency relationship between reference samples and pixels of a prediction block. For example, the value of the specified reference sample may be used as the value of the specified one or more pixels of the predictive block. In this case, the specified one or more pixels of the specified reference sample and prediction block may be samples and pixels specified by 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 a value of a pixel located in a direction opposite to the direction of the intra prediction mode. Alternatively, the value of the pixel of the prediction block may be the value of the reference sample located in the direction of the intra prediction mode based on the position of the pixel.

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

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

For example, when the mode value of the intra prediction mode of the target block is 34, at least some of the left reference samples, the upper left corner reference sample, and 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 the pixel of the prediction block may be the value of the reference sample located at the upper left side diagonally with respect to the pixel.

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

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

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

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

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

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

The prediction block generated by prediction may not be the same as the original target block. That is, a prediction error that is a difference between the target block and the prediction block may exist, and a prediction error may also exist between a pixel of the target block and a pixel of the prediction block.

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

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

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

As illustrated in FIG. 8 , at least one of reference lines 0 to 3 may be used for intra prediction of a target block. Each reference line may represent a reference sample line. As the number of the reference line is smaller, it may be a line of reference samples closer to the target block.

The samples of segment A and segment F may be obtained through padding using the closest samples of segment B and segment E, respectively, instead of being obtained from the reconstructed neighboring block.

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

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

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

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

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

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

For example, the parameter of the linear model may be 1) the value of the sample of the first color component having the maximum value among the samples in the template, 2) the value of the sample of the second color component corresponding to the sample of this first color component, 3) the value of the sample of the first color component having the smallest value among the samples in the template, and 4) the value of the sample of the second color component corresponding to this sample of the first color component.

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

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

Whether to perform intra prediction between color components and/or the range of a template may be signaled as an intra prediction mode.

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

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

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

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

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

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

The filtering may be performed by applying a specific weight to the filtering target sample, the left reference sample, the top reference sample, and/or the top left reference sample that are the filtering targets.

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

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

The adjacent diagonal mode may be a mode having a number in which k is added to the number of the diagonal mode, and may be a mode having a number in which k is subtracted from the number of the diagonal mode. In other words, the number of adjacent diagonal modes may be the sum of the number and k of the diagonal modes, and may be the difference between the number and k of the diagonal modes. For example, k may be a positive integer of 8 or less.

The intra prediction mode of the target block may be derived using the intra prediction mode of a neighboring block existing in the vicinity of the target block, and the derived intra prediction mode may be entropy-encoded and/or entropy-decoded.

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

Also, for example, indicator information on 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.

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

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

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

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

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

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

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

P-pictures and B-pictures that are encoded and/or decoded using a reference picture may be regarded as images for which inter prediction is used.

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

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

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

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

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

The spatial candidate may be a reconstructed block spatially adjacent to the target block.

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

In inter prediction, the encoding apparatus 100 and the decoding apparatus 200 may improve encoding efficiency and decoding efficiency by using motion information of a spatial candidate and/or a temporal candidate. 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.

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

Inter prediction may be performed using a reference picture.

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

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

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

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

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

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

Each of the encoding apparatus 100 and the decoding apparatus 200 may identify a block existing at a location 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 collocated picture may correspond to each other.

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

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

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

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

A method of deriving motion information may vary according to the inter prediction mode of the target block. For example, as an inter prediction mode applied for inter prediction, an advanced motion vector predictor (AMVP) mode, a merge mode and a skip mode, a merge mode with a motion vector difference, There may be a sub-block merge mode, a triangulation mode, an inter-intra joint prediction mode, an affine inter mode, and a current picture reference mode. The merge mode may be referred to as a motion merge mode. In the following, each of the modes is described in detail.

1) AMVP mode

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

1-1) Creation of a prediction motion vector candidate list

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

Hereinafter, the terms “prediction motion vector (candidate)” and “motion vector (candidate)” may be used with the same meaning, and may be used interchangeably.

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

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

A spatial candidate may include a reconstructed spatial neighboring block. That is, the motion vector of the reconstructed neighboring block may be referred to as a spatial prediction motion vector candidate.

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

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

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

1-2) Searching for a motion vector using the predicted motion vector candidate list

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

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

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

1-3) Transmission of inter prediction information

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

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

Hereinafter, the terms “prediction motion vector index” and “AMVP index” may be used with the same meaning, and may be used interchangeably.

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

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

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

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

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

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

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

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

The MVD may be a difference between a motion vector of a target block and a prediction motion vector. The encoding apparatus 100 may calculate an 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 MDV.

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

The decoding apparatus 200 may derive the motion vector of the target block by summing the MVD and the prediction motion vector. That is, the motion vector of the target block derived from the decoding apparatus 200 may be the sum of the MVD and the motion vector candidate.

Also, the encoding apparatus 100 may generate entropy-encoded MVD resolution information by applying entropy encoding to the calculated MVD resolution information, and may generate a bitstream including the entropy-encoded MVD resolution information. The decoding apparatus 200 may extract entropy-encoded MVD resolution information from the bitstream, and may obtain MVD resolution information by applying entropy decoding to the entropy-encoded MVD resolution information. The decoding apparatus 200 may adjust the resolution of the MVD by using the MVD resolution information.

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

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

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

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

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

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

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

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

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

Motion information of the reconstructed neighboring block for the target block may be used. In a specific inter prediction mode, the encoding apparatus 100 may not separately encode the motion information for the target block. The motion information of the target block is not coded, but other information capable of deriving the motion information of the target block through the reconstructed motion information of the neighboring block may be coded instead. As other information is encoded instead, the amount of bits transmitted to the decoding apparatus 200 may be reduced, and encoding efficiency may be improved.

For example, as an inter 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. In this case, the encoding apparatus 100 and the decoding apparatus 200 may use an identifier and/or an index indicating which unit of the reconstructed neighboring units motion information is used as the target unit's motion information.

2) Merge mode

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

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

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

2-1) Preparation of merge candidate list

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

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

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

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

Also, 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.

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

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

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

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

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

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

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

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

The number of merge candidates in the merge candidate list may be predefined. The encoding apparatus 100 and the decoding apparatus 200 may add merge candidates to the merge candidate list according to a predefined method and a predefined order so that the merge candidate list has a predefined number of merge candidates. The merge candidate list of the encoding apparatus 100 and the merge candidate list of the decoding apparatus 200 may become the same through a predefined scheme and a predefined rank.

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

2-2) Search for motion vector using merge candidate list

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

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

2-3) Transmission of inter prediction information

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

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

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

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

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

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

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

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

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

The correction information may include at least one of information indicating whether correction is made, correction direction information, and correction size information. A prediction mode that corrects a motion vector based on the signaled correction information may be referred to as a merge mode having a motion vector difference.

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

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

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

3) Skip mode

The skip mode may be a mode in which motion information of a spatial candidate or motion information of a temporal candidate is directly applied to a target block. Also, the skip mode may be a mode that does not use a residual signal. In other words, when the skip mode is used, the reconstructed block may be the same as the prediction block.

The difference between the merge mode and the skip mode may be whether a residual signal is transmitted or used. In other words, skip mode may be similar to merge mode except that no residual signal is transmitted or used.

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

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

3-1) Preparation of merge candidate list

The skip mode may also use the merge candidate list. In other words, the merge candidate list can be used in both the merge mode and the skip mode. In this aspect, the merge candidate list may be referred to as a “skip candidate list” or a “merge/skip candidate list”.

Alternatively, the skip mode may use a different candidate list from 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.

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

3-2) Search for motion vector using merge candidate list

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

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

3-3) Transmission of inter prediction information

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

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

The skip index may be the same as the above-described merge index.

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

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

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

3-4) Inter prediction in skip mode using inter prediction information

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

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

4) Current picture reference mode

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

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

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

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 for the target block.

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

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

5) subblock merge mode

The sub-block merge mode may refer to a mode for deriving motion information for a sub-block of a CU.

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

6) Triangle partition mode

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

7) Inter-intra joint prediction mode

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

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

In the above-described modes, the decoding apparatus 200 may compensate for a prediction sample derived through inter prediction using an optical flow.

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

In order to improve encoding efficiency, the encoding apparatus 100 may signal only the index of the element causing the minimum cost in inter prediction of the target block among elements of the list. The encoding apparatus 100 may encode the index and may signal the encoded index.

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

10 illustrates spatial candidates according to an example.

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

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

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

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

The spatial candidate A ₁ may be a block adjacent to the left of the target block. A ₁ may be the lowest block among blocks adjacent to the left of the target block. Alternatively, A ₁ may be a block adjacent to the upper end of A _{0 .} A ₁ may be a block occupying a pixel of coordinates (xP - 1, yP + nPSH).

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

The spatial candidate B ₁ may be a block adjacent to the top of the target block. B ₁ may be a rightmost block among blocks adjacent to the top of the target block. Alternatively, B ₁ may be a block adjacent to the left of B _{0 .} B ₁ may be a block occupying a pixel of coordinates (xP + nPSW, yP - 1).

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

Determination of availability of spatial and temporal candidates

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

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

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

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

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

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

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

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

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

Method of derivation of merge list in merge mode and skip mode

As described above, the maximum number of merge candidates in the merge list may be set. The set maximum number is displayed as N. The set number may be transmitted from the encoding apparatus 100 to the decoding apparatus 200 . A slice header of a slice may include N. That is, 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 5 by default.

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

Step 1) Among the spatial candidates, available spatial candidates may be added to the merge list. Motion information of available spatial candidates may be added to the merge list in the order shown in FIG. 10 . In this case, if motion information of an available spatial candidate overlaps with other motion information already existing in the merge list, the motion information may not be added to the merge list. Checking whether or not it overlaps with other motion information existing in the list may be abbreviated as "redundancy check".

There may be a maximum of N pieces of added motion information.

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

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

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

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

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

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

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

In this case, if 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.

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

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

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

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

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

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

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

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

Method of deriving a predictive motion vector candidate list in AMVP mode

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

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

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

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

Motion information of the 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, if motion information of an available spatial candidate overlaps with other motion information already existing in the prediction motion vector candidate list, the motion information may not be added to the prediction motion vector candidate list. That is, 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.

There may be a maximum of N pieces of added motion information.

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

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

There may be one or more zero vector motion information. Reference picture indices of one or more zero vector motion information may be different from each other.

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

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

The description of the zero vector motion information described above for the merge list may also be applied to the zero vector motion information. A duplicate description is omitted.

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

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

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

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

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

The transform kernel used for the transform may include a variety of DCT kernels such as Discrete Cosine Transform (DCT) type 2 (DCT-II) and a Discrete Sine Transform (DST) kernel. .

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 in which a one-dimensional (1D) transform is performed in each of a horizontal direction and a vertical direction on the residual signal.

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

[Table 3]

[Table 4]

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

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

[Table 5]

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

As illustrated in Table 5, transform sets applied in the horizontal direction and the vertical direction may be predefined according to the intra prediction mode of the target block. The encoding apparatus 100 may perform transform and inverse transform on the residual signal by using a transform included in a transform set corresponding to the intra prediction mode of the target block. Also, the decoding apparatus 200 may perform inverse transform on the residual signal by using a transform included in a transform set corresponding to the intra prediction mode of the target block.

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

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

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

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

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

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

The secondary transformation may be a transformation for improving the energy concentration of a transform coefficient generated by the primary transformation. The quadratic transform can be either a separable transform or a non-separable transform, like the first transform. The non-separable transform may be a Non-Separable Secondary Transform (NSST).

The first-order transformation may be performed using at least one of a plurality of predefined transformation methods. For example, a plurality of predefined transform methods include a discrete cosine transform (DCT), a discrete sine transform (DST), and a Karhunen-Loeve transform (KLT)-based transform. may include

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

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

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

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

[Table 6]

In Table 6, i and j may be integer values of 0 or more and N-1 or less.

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

As in the first-order transform, a transform set can be defined in the second-order transform. Methods for deriving and/or determining a transform set as described above may be applied not only to the first-order transform but also to the second-order transform.

A first-order transform and a second-order transform can be determined for a specified object.

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

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

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

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

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

For example, for one CU, whether the primary transform is used, an index indicating the primary transformation, whether the secondary transformation is used, an index indicating the secondary transformation, etc. may be derived as transformation information in the decoding apparatus 200 . have. Alternatively, transformation information indicating whether primary transformation is used, an index indicating primary transformation, whether secondary transformation is used, and an index indicating secondary transformation, etc. may be signaled for one CU.

Quantized transform coefficients (ie, quantized levels) may be generated by performing quantization on a residual signal or a result generated by performing a first-order transform and/or a second-order transform.

13 illustrates diagonal scanning according to an example.

14 illustrates horizontal scanning according to an example.

15 illustrates vertical scanning according to an example.

The quantized transform coefficients may be scanned according to at least one of (up-right) diagonal scanning, vertical scanning, and horizontal scanning according to at least one of an intra prediction mode, a block size, and a block shape. A block may be a transform unit.

Each scan may start at a specified start point and end at a specified end point.

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

The vertical scanning may be to scan the two-dimensional block shape coefficient in the column direction. Horizontal scanning may be a row-wise scanning of two-dimensional block shape coefficients.

That is, according to the size of the block and/or the inter prediction mode, it may be determined which scanning among diagonal scanning, vertical scanning, and horizontal scanning will be used.

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

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

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

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

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

The decoding apparatus 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 arranged in a two-dimensional block form through inverse scanning. In this case, as a method of inverse scanning, at least one of a (upper right) diagonal scan, a vertical scan, and a horizontal scan may be performed.

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

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

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

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

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

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

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

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

For example, scaling may be applied only when mapping for a luma component is available, and the division of the luma component and the division of the chroma component follow the same tree structure.

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

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

By performing scaling on the residual block using the finally derived value, the residual block can be converted into a demapped region. Thereafter, with respect to the chroma component block, reconstruction, intra prediction, inter prediction, in-loop filtering, and storage of a reference picture may be performed in the demapped region.

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

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

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

The IBC mode may be applied to a CU of a specified size. For example, the IBC mode may be applied to an MxN CU. Here, M and N may be 64 or less.

The IBC mode may include a skip mode, a merge mode, and an AMVP mode. In the skip mode or merge mode, a merge candidate list may be configured, and one merge candidate may be specified among merge candidates of the merge candidate list by signaling a merge index. The block vector of the specified merge candidate may be used as the block vector of the target block.

In the AMVP mode, a residual block vector may be signaled. Also, the prediction block vector may be derived from a left neighboring block and an upper neighboring block of the target block. In addition, an index regarding which neighboring block is to be used may be signaled.

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

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

The encoding apparatus 1600 may correspond to the above-described encoding apparatus 100 .

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

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

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

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

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

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

Program modules include routines, subroutines, programs, objects, components, and data that perform functions or operations according to an embodiment or implement abstract data types according to an embodiment. It may include, but is not limited to, a data structure and the like.

The program modules may include instructions or codes that are executed by at least one processor of the encoding apparatus 1600 .

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

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

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

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

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

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

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

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

17 is a structural diagram of a decoding apparatus according to an embodiment.

The decoding apparatus 1700 may correspond to the above-described decoding apparatus 200 .

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

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

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

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

Entropy decoding unit 210, inverse quantization unit 220, inverse transform unit 230, intra prediction unit 240, inter prediction unit 250, switch 245, adder 255, filter unit 260 and At least some of the reference picture buffer 270 may be program modules, and may communicate with an external device or system. The program modules may be included in the decryption apparatus 1700 in the form of an operating system, an application program module, and other program modules.

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

Program modules include routines, subroutines, programs, objects, components, and data that perform functions or operations according to an embodiment or implement abstract data types according to an embodiment. It may include, but is not limited to, a data structure and the like.

The program modules may include instructions or codes that are executed by at least one processor of the decryption apparatus 1700 .

The processing unit 1710 includes an entropy decoding unit 210 , an inverse quantization unit 220 , an inverse transform unit 230 , an intra prediction unit 240 , an inter prediction unit 250 , a switch 245 , an adder 255 , and a filter. It can execute instructions or codes of the sub 260 and the reference picture buffer 270 .

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

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

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

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

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

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

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

18 is a flowchart of an encoding method according to an embodiment.

In operation 1810 , the processor 1610 may generate a recovered low-quality image for the original image.

The restored low-quality image may be a recovered low-resolution image and/or a recovered low-bitrate image.

Step 1810 may include steps 1811 , 1812 , and 1813 .

In operation 1811, the processing unit 1610 may generate an encoded low-quality image by performing encoding on the original image.

The encoded low-quality image may be an encoded low-resolution image and/or an encoded low-bitrate image.

For example, the processor 1610 may generate an encoded low-resolution image by performing encoding on the reconstructed low-resolution image of the original image.

In operation 1812 , the processor 1610 may generate a bitstream including an encoded low-quality image.

The storage unit 1730 may store a bitstream.

In operation 1813 , the processor 1610 may generate a reconstructed low-quality image by using the encoded low-quality image.

The processor 1610 may generate a reconstructed low-quality image by performing decoding on the encoded low-quality image in the bitstream.

Unlike steps 1811, 1812, and 1813, the processing unit 1610 performs encoding on the original image in steps 1181 and 1182 to generate an encoded low-quality image, while also generating a reconstructed low-quality image. may be

In operation 1820 , the processor 1610 may generate a recovered high-quality image by using the restored low-quality image.

The reconstructed high-quality image may include a recovered high-resolution image and/or a coding distortion corrected image.

If the encoded low-quality image is an encoded low-resolution image, the reconstructed high-quality image may be a reconstructed high-resolution image. The reconstructed high-resolution image may have a higher resolution than that of the reconstructed low-resolution image.

If the encoded low-quality image is an encoded low-bit-rate image, the reconstructed high-quality image may be a coding distortion correction image. The coding distortion-corrected image may have a higher bit rate than the bit-rate of the reconstructed low bit rate image. The coding distortion correction image may have a higher image quality than that of the reconstructed low bit rate image.

The coding distortion correction image may be an image generated by applying restoration for reducing coding distortion to the restored low bit rate image.

If the reconstructed low-quality image is the reconstructed low-resolution image, the processing unit 1610 may generate a reconstructed high-resolution image by using the reconstructed low-quality image.

If the reconstructed low-quality image is the reconstructed low-bit rate image, the processing unit 1610 may generate a coding distortion correction image by using the reconstructed low-quality image.

For example, the processor 1710 may generate a reconstructed high-resolution image by performing scaling on the reconstructed low-resolution image.

Generating a reconstructed high-resolution image and/or a coding distortion correction image using the reconstructed low-quality image may be considered to generate an ultra-high-resolution image.

Generating a reconstructed high-resolution image using the reconstructed low-resolution image may be referred to as “super-resolution (SR) restoration”. The reconstructed high-resolution image may be referred to as an “SR image”.

Generating a coding distortion correction image using the reconstructed low bit rate image may be referred to as “artifact reduction (AR) restoration”. The coding distortion correction image may be referred to as an “AR image”.

Also, hereinafter, the “reconstructed super-resolution image” may mean a “reconstructed high-resolution image” and/or a “coding distortion-corrected image”.

Also, in the following, a characteristic operation according to an embodiment that is distinguished from a general operation used for encoding and/or decoding of a video may be identified using “SR” and “AR”.

In operation 1830 , the processing unit 1610 may segment the image into a plurality of regions by performing segmentation on the image. Here, the image may be at least one of an original image, a reconstructed low-quality image, a reconstructed high-quality image, a reconstructed high-resolution image, and a coding distortion correction image. Alternatively, the image may be an original image, a reconstructed low-quality image, a reconstructed high-quality image, a reconstructed high-resolution image, or a coding distortion correction image.

In operation 1840, the processing unit 1610 may generate encoded reconstruction information by performing encoding on the region. Here, the region may be at least one of the plurality of regions described above.

The processing unit 1610 may generate the encoded reconstruction information for the region using a specific method selected in consideration of the characteristics of the region. In other words, the processing unit 1610 may perform differential encoding for each region on a plurality of regions. Through the differential encoding for each region, the processing unit 1610 may generate encoded reconstruction information based on the characteristics of each region of the plurality of regions.

The encoded reconstruction information may be information used to generate a reconstructed high-quality image with respect to the original image.

The reconstructed high-quality image may be a reconstructed high-resolution image and/or a reconstructed high-bit rate image.

The reconstructed high-resolution image and/or the coding distortion correction image may be used as a kind of prediction image for generating encoded reconstruction information.

In operation 1850, the processing unit 1610 may generate a bitstream including the encoded reconstruction information.

The bitstream may include the above-described encoded low-quality image and encoded reconstruction information.

The encoded low-quality image may be transmitted through the first bitstream. The encoded reconstruction information may be transmitted through a separate second bitstream, and the second bitstream may be selectively transmitted.

The first bitstream may be a base bitstream.

The first bitstream may include type information indicating the type of the first bitstream. The processing unit 1610 may generate type information indicating the type of the first bitstream, and may include the type information in the first bitstream.

The type information may indicate the type of the first bitstream. The type of the first bitstream may be one of a low-resolution bitstream and a low-bitrate bitstream. The type information may indicate which image of the first bitstream includes an encoded low-resolution image and an encoded low-bit-rate image.

When the low-quality image is a low-resolution image, the type of the first bitstream may be a low-resolution bitstream. That is, if the type of the first bitstream is a low-resolution bitstream, the first bitstream may include an encoded low-resolution image.

When the low-quality image is a low-bit-rate image, the type of the first bitstream may be a low-bit-rate bitstream. That is, if the type of the first bitstream is a low bitrate bitstream, the first bitstream may include an encoded low bitrate image.

The second bitstream may be an additional bitstream.

In operation 1860 , the communication unit 1620 may transmit the bitstream to the decoding apparatus 1700 .

19 is a flowchart of a method for generating coded information using a map according to an embodiment.

The embodiment to be described with reference to FIG. 19 may be an example or a part of the embodiment described above with reference to FIG. 18 .

Step 1830 described above with reference to FIG. 18 may include step 1910 .

In operation 1910, the processing unit 1610 may generate a map for the image by performing segmentation on the image.

The map may indicate a plurality of regions. That is, the map may divide an image into a plurality of regions and may indicate a plurality of divided regions. In this sense, a “map” may be called a region-delimited map.

For example, the map may indicate a plurality of regions with different values, respectively.

For example, the map may indicate a first area among the plurality of areas as “0” and may indicate a second area among the plurality of areas as “1”. In other words, the first area may be an area indicated by "0" in the map. The second area may be an area indicated by "1" on the map.

Alternatively, the map may indicate a first area among the plurality of areas as “1” and a second area among the plurality of areas as “0”. In other words, the first area may be an area indicated by "1" in the map. The second area may be an area indicated by "0" on the map.

A prediction image may be generated for each region of the plurality of regions indicated by the map.

Step 1840 described above with reference to FIG. 18 may include steps 1920 , 1930 , 1940 and 1950 .

In operation 1920, the processing unit 1610 may divide the image into a plurality of regions using the map, and may select an encoding method for each region of the plurality of regions.

The plurality of regions may include a first region and a second region.

In an embodiment, the processor 1610 may generate a reconstructed high-quality image by using “copy” for the first region of the map. In this respect, the first area in which radiation is used may be referred to as a radiation area.

Here, the copying refers to copying the first region of the reconstructed high-quality image as it is to the first region of the reconstructed high-quality image or using the first region of the reconstructed high-quality image as the first region of the reconstructed high-quality image as it is. can mean doing Accordingly, information other than the reconstructed high-quality image and map may not be required for the first region.

For example, the processor 1610 may generate a reconstructed high-quality image using “prediction” for the second region of the map. From this point of view, the second region in which prediction is used may be referred to as a prediction region.

Here, the prediction may mean using the second region of the reconstructed high-quality image as a prediction image for the second region of the reconstructed high-quality image. In other words, the prediction may mean using the second region of the reconstructed high-quality image and the sum of the reconstructed residual image for the second region as the second region of the reconstructed high-quality image. Therefore, for the second region, a residual signal representing a result of subtracting the second region of the reconstructed high-quality image from the second region of the original image may be required, and the encoded residual signal (through a conventional encoding method, etc.) It may have to be transmitted from the encoding apparatus 1600 to the decoding apparatus 1700 . The decoding apparatus 1700 may obtain a reconstructed residual signal by decoding the encoded residual signal.

In an embodiment, the processing unit 1610 may divide the image into 1) a first region to be encoded/decoded according to the aspect of perceived quality, and 2) a second area to be encoded/decoded according to the aspect of a pixel difference.

Here, the first region encoded/decoded according to the aspect of perceptual image quality may be referred to as a “perceptual-oriented region”, and 2) the second region encoded/decoded according to the aspect of the pixel difference may be referred to as a “distortion-oriented region”. can be named

In an embodiment, the processor 1610 may generate a prediction image using a “perception-oriented SR network” for the first region of the map.

That is, the sum of the predicted image for the first region generated using the cognitive-oriented SR network and the reconstructed residual image for the first region may be used as the first region of the reconstructed high-quality image. Therefore, for the first region, a residual signal representing the result of subtracting the predicted image from the first region of the original image is required, and the encoded residual signal (through a conventional encoding method, etc.) is transmitted from the encoding device 1600 to the decoding device. may have to be signaled as (1700). The decoding apparatus 1700 may obtain a reconstructed residual signal by decoding the encoded residual signal.

The processor 1610 may generate a prediction image for the second region of the map using a distortion-oriented SR network.

That is, the sum of the prediction image for the second region generated using the distortion-oriented SR network and the reconstructed residual image for the second region may be used as the second region of the reconstructed high-quality image. Therefore, for the second region, a residual signal representing the result of subtracting the predicted image from the second region of the original image is required, and the encoded residual signal (through a conventional encoding method, etc.) is transmitted from the encoding device 1600 to the decoding device. may have to be signaled as (1700). The decoding apparatus 1700 may obtain a reconstructed residual signal by decoding the encoded residual signal.

In the embodiment, the image is divided into two regions, but the image may be divided into three or more regions, and different encoding methods may be used for a plurality of regions indicated by the map, respectively.

Different encoding methods may use different neural networks. That is, encoding may be performed on a plurality of regions indicated by the map using different neural networks, respectively.

In operation 1930 , the processor 1610 may generate a residual signal with respect to at least one of a plurality of regions of the original image.

The residual signal may be used for a region in which a prediction image is used.

For example, a residual signal may not be generated for a first region where radiation is used.

For example, among the plurality of regions, at least one region in which a residual signal is generated includes 1) a region in which prediction is used, 2) a region in which a prediction image is generated using a cognitive-oriented SR network, and 3) a distortion-oriented SR network. It may include at least a part of an area in which a prediction image is generated using the MSE AR network, 4) an area in which a prediction image is generated using the MSE AR network, and 5) an area in which a prediction image is generated using the GAN AR network.

For example, a residual signal may be generated for the second region where prediction is used.

The residual signal may be a result of subtracting the predicted image from the region of the original image.

In operation 1940 , the processor 1610 may generate an encoded residual signal by performing encoding on the residual signal.

Here, the encoding may include reconstruction and quantization.

Also, the encoding may include operations of encoding an existing encoded image or an encoded residual signal.

The coded reconstruction information described above with reference to FIG. 18 may include an coded residual signal.

The second bitstream may include an encoded residual signal.

In operation 1950, the processor 1610 may generate a bitstream including the encoded residual signal.

20 is a flowchart of a decoding method according to an embodiment.

In operation 2010 , the communication unit 1730 may receive a bitstream from the encoding apparatus 1600 .

The bitstream may include the above-described encoded low-quality image and encoded reconstruction information.

The bitstream may include a first bitstream and a second bitstream.

The first bitstream may be a base bitstream.

The first bitstream may include type information indicating the type of the first bitstream. The processing unit 1710 may obtain type information from the first bitstream and identify the type of the first bitstream using the type information.

The type information may indicate the type of the first bitstream. The type of the first bitstream may be one of a low-resolution bitstream and a low-bitrate bitstream. The type information may indicate which image of the first bitstream includes an encoded low-resolution image and an encoded low-bit-rate image.

When the low-quality image is a low-resolution image, the type of the first bitstream may be a low-resolution bitstream. That is, if the type of the first bitstream is a low-resolution bitstream, the first bitstream may include an encoded low-resolution image.

When the low-quality image is a low-bit-rate image, the type of the first bitstream may be a low-bit-rate bitstream. That is, if the type of the first bitstream is a low bitrate bitstream, the first bitstream may include an encoded low bitrate image.

If the encoded low-quality image is an encoded low-resolution image, the reconstructed high-quality image may be a reconstructed high-resolution image.

If the encoded low-quality image is an encoded low-bit-rate image, the reconstructed high-quality image may be a coding distortion correction image.

The second bitstream may be an additional bitstream.

The encoded low-quality image may be received through the first bitstream.

The encoded reconstruction information may be received through a separate second bitstream. The second bitstream may be selectively received.

In operation 2020, the processor 1710 may perform decoding on the encoded low-quality image to generate a reconstructed low-quality image.

In operation 2030, the processor 1710 may generate a reconstructed high-quality image by using the reconstructed low-quality image.

The reconstructed high-quality image may include a reconstructed high-resolution image and/or a coding distortion correction image.

For example, when the reconstructed low-quality image is a reconstructed low-resolution image (or when the type of the first bitstream is a low-resolution bitstream), the processing unit 1710 generates a reconstructed high-resolution image by using the reconstructed low-resolution image can do.

For example, when the reconstructed low-quality image is a reconstructed low-bit-rate image (or when the type of the first bitstream is a low-bit-rate bitstream), the processing unit 1710 corrects coding distortion by using the reconstructed low-bit-rate image. You can create an image.

For example, the processor 1710 may generate a reconstructed high-resolution image by performing scaling on the reconstructed low-resolution image.

In operation 2040 , the processing unit 1710 may segment the image into a plurality of regions by performing segmentation on the image. Here, the image may be at least one of an original image, a reconstructed low-quality image, a reconstructed high-quality image, a reconstructed high-resolution image, and a coding distortion correction image. Alternatively, the image may be an original image, a reconstructed low-quality image, a reconstructed high-quality image, a reconstructed high-resolution image, or a coding distortion correction image.

Here, the encoding apparatus 1600 and the decoding apparatus 1700 may divide an image into a plurality of regions in the same manner. Accordingly, the image may be equally divided into a plurality of regions in the encoding apparatus 1600 and the decoding apparatus 1700 .

Alternatively, information indicating a plurality of regions or information for dividing the regions into a plurality of regions may be signaled from the encoding apparatus 1600 to the decoding apparatus 1700 through a bitstream.

In operation 2050 , the processing unit 1710 may perform decoding on the region to generate reconstruction information. Here, the region may be at least one of the plurality of regions described above.

The processing unit 1710 may generate the reconstruction information for the region using a specific method selected in consideration of the characteristics of the region. In other words, the processing unit 1710 may perform differential decoding on a plurality of regions for each region. Through the differential decoding for each region, the processing unit 1710 may generate reconstruction information based on the characteristics of each region of the plurality of regions.

The reconstruction information may be information used to generate a reconstructed high-quality image with respect to the original image.

In operation 2060 , the processor 1710 may generate a reconstructed image using the reconstruction information.

21 is a flowchart of a method of generating decrypted information using a map according to an embodiment.

The embodiment to be described with reference to FIG. 21 may be an example or a part of the embodiment described above with reference to FIG. 20 .

Step 2040 described above with reference to FIG. 20 may include step 2110 .

In operation 2110, the processing unit 1710 may generate a map for the image by performing segmentation on the image.

The map may indicate a plurality of regions. That is, the map may divide an image into a plurality of regions and may indicate a plurality of divided regions.

For example, the map may indicate a plurality of regions with different values, respectively.

For example, the map may indicate a first area among the plurality of areas as “0” and may indicate a second area among the plurality of areas as “1”. In other words, the first area may be an area indicated by "0" in the map. The second area may be an area indicated by "1" on the map.

Alternatively, the map may indicate a first area among the plurality of areas as “1” and a second area among the plurality of areas as “0”. In other words, the first area may be an area indicated by "1" in the map. The second area may be an area indicated by "0" on the map.

Here, the map generated by the decoding apparatus 1700 and the method of generating the map may be the same as the map generated by the encoding apparatus 1600 and the method of generating the map, respectively.

Alternatively, a map, information indicating the map, or information indicating a plurality of regions may be signaled from the encoding apparatus 1600 to the decoding apparatus 1700 through a bitstream.

Step 2050 described above with reference to FIG. 20 may include steps 2120 , 2130 , 2140 , and 2150 .

In operation 2120, the processing unit 1710 may divide the image into a plurality of regions using the map, and may select a decoding method for each region of the plurality of regions.

For example, the processor 1710 may generate a reconstructed high-quality image by using “copy” for the first region of the map.

Here, the copying refers to copying the first region of the reconstructed high-quality image as it is to the first region of the reconstructed high-quality image or using the first region of the reconstructed high-quality image as the first region of the reconstructed high-quality image as it is. can mean doing Accordingly, information other than the reconstructed high-quality image and map may not be required for the first region.

For example, the processor 1710 may generate a reconstructed high-quality image using “prediction” for the second region of the map.

Here, the prediction may mean using the second region of the reconstructed high-quality image as a prediction image for the second region of the reconstructed high-quality image. In other words, the prediction may mean using the second region of the reconstructed high-quality image and the sum of the reconstructed residual image for the second region as the second region of the reconstructed high-quality image. Therefore, for the second region, a residual signal representing the result of subtracting the second region of the reconstructed high-quality image from the second region of the original image is required, and the encoded residual signal (through a conventional encoding method, etc.) It may have to be transmitted from the 1600 to the decoding device 1700 .

In an embodiment, the processing unit 1710 may generate a prediction image for the first region of the map by using a “cognitive oriented SR network”.

That is, the sum of the predicted image for the first region generated using the cognitive-oriented SR network and the reconstructed residual image for the first region may be used as the first region of the reconstructed high-quality image. Therefore, for the first region, a residual signal representing the result of subtracting the predicted image from the first region of the original image is required, and the encoded residual signal (through a conventional encoding method, etc.) is transmitted from the encoding device 1600 to the decoding device. may have to be signaled as (1700). The decoding apparatus 1700 may obtain a reconstructed residual signal by decoding the encoded residual signal.

The processor 1710 may generate a prediction image for the second region of the map using a distortion-oriented SR network.

That is, the sum of the prediction image for the second region generated using the distortion-oriented SR network and the reconstructed residual image for the second region may be used as the second region of the reconstructed high-quality image. Therefore, for the second region, a residual signal representing the result of subtracting the predicted image from the second region of the original image is required, and the encoded residual signal (through a conventional encoding method, etc.) is transmitted from the encoding device 1600 to the decoding device. may have to be signaled as (1700). The decoding apparatus 1700 may obtain a reconstructed residual signal by decoding the encoded residual signal.

In the embodiment, the image is divided into two regions, but the image may be divided into three or more regions, and different decoding methods may be used for a plurality of regions indicated by the map, respectively.

Different decoding methods may use different neural networks. That is, decoding may be performed on a plurality of regions indicated by the map using different neural networks, respectively.

In operation 2130 , the processor 1710 may extract the encoded residual signal from the bitstream.

The processor 1710 may extract the encoded residual signal from the second bitstream.

In operation 2140 , the processor 1710 may generate a (reconstructed) residual signal by performing decoding on the encoded residual signal.

Here, the decoding may include inverse quantization and inverse transformation. Also, decoding may include operations of decoding on an existing coded image or coded residual signal.

In operation 2150 , the processor 1710 may generate a reconstructed high-quality image using the (reconstructed) residual signal.

Here, the reconstructed high-quality image may mean the reconstructed image resulting from the operation 2060, and may be the final image. Also, the reconstructed high-quality image may be understood as an ultra-high-resolution image.

For example, the processing unit 1710 may generate the first region of the reconstructed high-quality image by copying the first region of the reconstructed high-quality image to the first region of the reconstructed high-quality image. That is, the first region of the reconstructed high-quality image may be the same as the first region of the reconstructed high-quality image or may be generated based on the first region of the reconstructed high-quality image.

For example, the processor 1710 may generate the second region of the reconstructed high-quality image by summing the second region of the reconstructed high-quality image and the image of the (reconstructed) residual signal. That is, the second region of the reconstructed high-quality image may be equal to the sum of the second region of the reconstructed high-quality image and the image of the (reconstructed) residual signal. Alternatively, the second region of the reconstructed high-quality image may be generated based on the sum of the second region of the reconstructed high-quality image and the image of the (reconstructed) residual signal.

For example, the processing unit 1710 generates a first region of a reconstructed high-quality image by summing the image of the (reconstructed) residual signal and the prediction image for the first region generated using the cognitive-oriented SR network. can That is, the first region of the reconstructed high-quality image may be equal to the sum of the predicted image and the (reconstructed) residual signal image for the first region generated using the cognitive-oriented SR network. Alternatively, the first region of the reconstructed high-quality image may be generated based on the sum of the predicted image for the first region and the (reconstructed) residual signal image generated using the cognitive-oriented SR network.

For example, the processing unit 1710 may generate the second region of the reconstructed high-quality image by summing the image of the (reconstructed) residual signal and the prediction image for the second region generated using the distortion-oriented SR network. can In other words, the second region of the reconstructed high-quality image may be equal to the sum of the predicted image and the (reconstructed) residual signal image for the second region generated using the distortion-oriented SR network. Alternatively, the second region of the reconstructed high-quality image may be generated based on the sum of the predicted image for the second region and the (reconstructed) residual signal image generated using the cognitive-oriented SR network.

In the following, more specific operations are described in relation to the above-described methods. Operations to be described later may be considered to be performed by the processing unit 1610 in the encoding apparatus 1600 and performed by the processing unit 1710 in the decoding apparatus 1700 .

The method of using the above-described ultra-high-resolution reconstructed image as a prediction image includes 1) signaling a bitstream including the coded low-resolution image, and 2) an coded (high-resolution) image generated using the low-resolution image as a prediction image. It may be to additionally signal the bitstream that includes it.

In this case, the image may be divided into a plurality of regions, and differential encoding/decoding may be performed on each region of the plurality of divided regions.

That is, the image may be divided into n regions, and different encoding/decoding methods may be applied to the n divided regions.

In an embodiment, the description of the reconstructed high-resolution image may also be applied to the coding distortion correction image.

In the method of using a coding distortion correction image as a prediction image, instead of using a high-resolution reconstructed image as a prediction image, a prediction image is generated by correcting encoding distortion of a low-resolution image, and an encoded bitstream is additionally transmitted using the prediction image may be doing

In this case, the image may be divided into a plurality of regions, and differential encoding/decoding may be performed on each region of the plurality of divided regions.

That is, the image may be divided into n regions, and different encoding/decoding methods may be applied to the n divided regions.

22 illustrates a processing process of an image encoding method using a reconstructed super-resolution image according to an example.

In FIG. 22 , X _org may represent an original image. X _LR may indicate a low-resolution image corresponding to the original image. Bin(X _LR ) represents a bitstream for a low-resolution image. X' _LR may represent a reconstructed low-resolution image.

In an embodiment, the SR may indicate generation of a reconstructed high-resolution image using the reconstructed low-resolution image. X' _SR may represent a reconstructed high-resolution image. Map extraction may refer to extracting a map by performing segmentation on a target image. The target image may be a reconstructed high-resolution image. M' _SR may represent a map extracted from a reconstructed high-resolution image.

In addition, since the above-mentioned copying and prediction are _{characteristic copying and characteristic prediction of the embodiment performed on the reconstructed high-resolution image M' SR} , they may be called SR copying and SR prediction, respectively. In other words, the SR copy may represent a copy for the first region of the map. The SR prediction may indicate a prediction for the second region of the map.

“SR copy” may refer to a method of generating a prediction image using an SR network and copying the prediction image as a reconstructed high-quality image.

"SR prediction" may refer to a method in which a prediction image is generated using an SR network, and a difference between the predicted image and the original image is encoded/decoded.

“AR copy” may refer to a method of generating a prediction image using an AR network and copying the prediction image as a reconstructed high-quality image.

“AR prediction” may refer to a method in which a prediction image is generated using an AR network and a difference between the prediction image and the original image is encoded/decoded.

For example, the region to which the SR copy is applied may be a region in which the predicted image can be copied if the predicted image is subjectively similar to the original image. For example, an area to which SR copy is applied may include a texture area.

For example, the region to which the SR prediction is applied may be a region in which an error (eg, a mean squared error (MSE)) with respect to a pixel of the original image should be minimized. For example, a region to which SR prediction is applied may include a region of interest (ROI) region or an edge region.

In addition, Residual(X _org -X' _SR )) may represent a difference between the original image and the reconstructed high-quality image. bin(Residual(X-X' _SR )) may indicate a bitstream for a difference between an image and a reconstructed high-quality image.

In an embodiment, the reconstructed high-resolution image may be replaced with a coding distortion correction image. For example, X' _SR may represent a coding distortion correction image.

In an embodiment, a unit of encoding, decoding, and/or prediction may be a block.

For "SR copy block", a prediction image may be generated using an SR network, and the generated prediction image may be copied. In other words, the SR copy block may be a block to which copy or SR copy is applied.

For "SR prediction block", a prediction image may be generated using an SR network, and a difference between the prediction image and the original image may be encoded. In other words, the SR prediction block may be a block to which prediction or SR prediction is applied.

For "AR copy block", a prediction image may be generated using an AR network, and the generated prediction image may be copied. In other words, the AR copy block may be a block to which copy or AR copy is applied.

For "AR prediction block", a prediction image may be generated using an AR network, and a difference between the prediction image and the original image may be encoded. In other words, the AR prediction block may be a block to which prediction or AR prediction is applied.

The above-described first region may be regarded as “a block corresponding to the first region”. The above-described second region may be regarded as a “block corresponding to the second region”. That is, the above-described description of the first area may be regarded as a description of each block corresponding to the first area. The above-described description of the second area may be regarded as a description of each block corresponding to the second area.

23 illustrates a processing process of an image decoding method using an ultra-high-resolution reconstructed image according to an example.

In FIG. 23 , Bin(X _LR ) may indicate a bitstream for a low-resolution image. X' _LR may represent a reconstructed low-resolution image. SR may indicate generation of a reconstructed high-resolution image using the reconstructed low-resolution image. X' _SR may represent a reconstructed high-resolution image.

Also, M' _SR may represent a map extracted from a reconstructed high-resolution image. bin(Residual(X-X' _SR )) may indicate a bitstream for a difference between an image and a reconstructed high-quality image. _X'prd may represent the final result, the reconstructed high-quality image.

In an embodiment, the reconstructed high-resolution image may be replaced with a coding distortion correction image. For example, X' _SR may represent a coding distortion correction image.

24 shows an original image used for semantic classification according to an example.

25 shows a result of semantic classification according to an example.

In an embodiment, an image of a map generated to classify regions may have a bit depth equal to the number of images for classifying images. In other words, a bit depth of a pixel of an image of a map may correspond to the number of divided regions represented by the map.

For example, when an image is divided into two regions, a map may be expressed as a binary image.

First Classification by Image Complexity

As an embodiment of map generation, classification may be made according to the complexity of an image.

For each region, an encoding method for the region may be divided into perception-oriented and distortion-oriented. A correlation between the characteristics of each encoding method and the complexity of the image may be analyzed.

Traditionally, it is known that people are cognitively less sensitive to errors caused by differences between pixels in a region with high complexity. According to these characteristics, a map that divides regions according to complexity may be generated.

In an embodiment, the complexity of the image may be determined based on edge energy of the image. For example, the complexity may be calculated based on the magnitude of spatial information at the position of each pixel using a horizontal kernel and a vertical kernel of a Sobel filter.

For example, an average value of spatial information may be used as the complexity of an image as shown in Equations 1 and 2 below.

[Formula 1]

[Formula 2]

In determining the encoding method based on the complexity of the image, the processing unit 1610 and the processing unit 1710 can obtain a map for the original image by deriving the complexity of the original image, and by deriving the complexity for the low-quality image It is possible to obtain a map for a low-quality image. In other words, the map described above may be a map for the original image and/or a map for a low quality image.

Alternatively, the processor 1610 and the processor 1710 may acquire a map based on complexity in consideration of both the original image and the low-quality image.

Through this method, an SR prediction image generation error due to a coding error may be reduced. That is, when the original image has a texture region having high complexity, but high-frequency information of the texture region disappears due to a coding error and only low-frequency information remains, SR Restoration by the network may have limitations. Therefore, when a region having such a high complexity is encoded by a method of copying an SR prediction image as it is, serious image quality degradation may occur in the region. Therefore, by generating a map that also considers the complexity of a low-quality image, and using SR prediction, which encodes/decodes the residual signal between the original image and the predicted image, rather than SR copy, for a region with high complexity, deterioration of image quality is prevented can be

Second Classification by Image Complexity

As an embodiment of the map generation, a machine learning-based semantic segmentation technique may be used for semantic segmentation.

For example, objects such as trees, sky, grass, grass, roads, water, and people commonly appearing in natural images may be designated as labels, and learning may be performed on the designated labels. Also, when an image after learning is input as an original image, regions may be divided in units of pixels as shown in FIG. 25 according to the meaning in the image.

According to a label designated in the semantic segmentation technique, a texture region and a non-texture region may be defined, and a map divided into a texture region and a non-texture region may be generated. In other words, the plurality of regions indicated by the map may include a textured region and a non-textured region.

According to the utilization of such technology, the category of the image may be determined according to the application program.

26 shows an original image in classification of a texture region according to an example.

27 shows a result of classification of a texture area according to an example.

26 and 27 show results of generating a map dividing regions for encoding and/or decoding of an image by the method described in the above-described embodiment. That is, the original image may be divided into a tree corresponding to the texture region and the remaining regions through a semantic classification label for the original image.

The map of the embodiment may be generated using a neural network that automatically distinguishes a textured region and a non-textured region. Alternatively, the map of the embodiment may be generated using a neural network that automatically classifies regions corresponding to features according to encoding/decoding methods.

As described above, a map for dividing regions by semantic classification of the target image may be extracted. The target image may be an original image or a reconstructed high-quality image generated by converting the reconstructed low-quality image. Also, the target image may be both an original image and a reconstructed high-quality image.

In addition, in order to improve image quality degradation that occurs when encoding/decoding and SR restoration or encoding/decoding and AR restoration are applied, semantic segmentation and edge power may be utilized in generating a map.

For example, the edge power of the image may be calculated, and the map may be modified based on the edge power of the image.

The utilization of edge power is described in more detail below.

As illustrated, an image may be analyzed through the application of a semantic segmentation technique. According to this analysis, regions to which radiation and prediction are applied may be divided.

28 illustrates generation of a map using a neural network according to an example.

A map for dividing regions according to the above-described encoding/decoding method may be generated using a neural network that automatically divides a texture region and a non-texture region with respect to an actual input image.

In an embodiment, the neural network may perform learning to classify regions of real textures in units of blocks, and a map may be automatically extracted from an image using the learned neural network.

28 , a label and a labeled image may be input to a texture detector (ie, a neural network). The texture detector may output an estimated label.

Here, the loss may be the cross entropy between the label and the estimated label.

A map for dividing regions according to the encoding/decoding method of the embodiment may be generated by reflecting characteristics according to the encoding/decoding method. For example, the neural network may perform learning to classify an image by reflecting the conspicuous degree of distortion caused by encoding/decoding, and the neural network on which this learning is performed may generate a map.

29 illustrates generation of a region differential super-resolution reconstructed image according to an exemplary embodiment.

In an embodiment, the reconstructed image generated by decoding may be referred to as an ultra-high-resolution reconstructed image.

In generating the reconstructed super-resolution image, the low-resolution image input from the encoding apparatus 1600 and the decoding apparatus 1700 may be reconstructed into a high-resolution image using the SR network. Such a reconstructed high-resolution image may be used as a prediction image in encoding and decoding.

As described above in the embodiments, in generating the reconstructed high-resolution image, a method of generating the reconstructed high-resolution image may be determined according to an encoding method of the original image or a region of the original image. That is, a method of generating a reconstructed high-resolution image corresponding to the used encoding method may be selected according to which encoding method is used for the original image or a region of the original image.

29 shows that SR is separately applied to a region to which SR radiation is applied and a region to which SR prediction is applied.

For example, compared to an image with a small error in terms of mean squared error (MSE), an image that has little effect on subjective image quality even if copied as it is is used as a copy area for a reconstructed image. can be advantageous Accordingly, a reconstructed high-resolution image may be generated in a manner in which artifacts due to encoding are less generated in the copy area.

For example, when encoding a difference from an original image according to an existing method (eg, DCT transformation that converts a signal in a frequency domain), an image in which energy compaction can easily occur is a reconstructed image. It may be more advantageous to be used as a prediction region for Accordingly, a high-resolution image reconstructed by the conventional method may be generated for the copy area.

As illustrated, the encoding method can be differentially applied to each region according to the characteristics of each region of the plurality of regions, and the reconstructed high-quality image used as a prediction image is also differentially applied by this differential application. can be created Alternatively, the reconstructed high-quality image may also be divided into a plurality of regions, and in generating the reconstructed high-quality image, different restoration methods may be applied to the plurality of regions, respectively.

For example, if the encoded low-quality image is an encoded low-resolution image, differential reconstruction from a low resolution to a high resolution may be performed on the image based on the map, and a prediction image may be generated by the differential reconstruction. Here, the differential reconstruction may mean that different reconstructions are performed on a plurality of regions of the map, respectively.

For example, if the encoded low-quality image is an encoded low-bit rate image, differential correction may be performed on coding distortion generated in the image based on the map, and a predicted image may be generated by the differential correction. Here, the differential correction may mean that different corrections are performed on a plurality of regions of the map, respectively.

For example, when generating a reconstructed high-quality image using the reconstructed low-quality image, a generative adversarial network (GAN) method may be used. By using the GAN, a natural image can be generated through the learning of the neural network rather than the aspect of the MSE. This reconstructed high-quality image may be used as a prediction image for the radiation area. In other words, the copy area of the reconstructed high-quality image may be generated through the GAN using the reconstructed low-quality image.

For example, in a prediction region that encodes a difference from an original image, a reconstructed high-quality image may be generated using an existing interpolation method so that the residual signal can be better compressed, and such a reconstructed high-quality image may be used as a prediction image for the prediction region. That is, the prediction region of the reconstructed high-quality image may be generated through interpolation using the reconstructed low-quality image.

As illustrated in FIG. 29 , a plurality of prediction images for a plurality of regions generated by classification may be synthesized based on a map.

In FIG. 29 , X' _LR may represent a reconstructed low-resolution image.

For example, X' _{SR_t} may represent a reconstructed high-resolution image generated by a reconstruction method corresponding to SR copy. For example, the reconstruction method corresponding to the SR copy may be a GAN generation method using the _{reconstructed low-resolution image X' LR.} X' _{SR_t} may be a reconstructed high-resolution image for a region representing a texture.

For example, X' _{SR_nt} may indicate a reconstructed high-resolution image generated by a reconstruction method corresponding to SR prediction. For example, the reconstruction method corresponding to the SR prediction may be interpolation using the _{reconstructed low-resolution image X' LR.} X' _{SR_t} may be a reconstructed high-resolution image for a non-texture region.

Can be a selected area of the texture by being a final result of a high quality reconstructed image X _'prd produced - by the map, "the area of the texture selected and, in the X _{SR_t'} ratio in the X _{SR_nt.} For example, the region indicated by the map as “0” _{may be generated using X' SR_t} , and the region indicated by the map as “1” may be created using _{X' SR_nt.}

30 and 31 illustrate degradation of image quality in a low bit rate environment according to an example.

30 shows a reconstructed image of HEVC qp37 according to an example.

31 shows a reconstructed high-quality image of a low-resolution HEVC qp37 according to an example.

30 and 31 , qp37 may represent a quantization parameter in encoding and/or decoding. In addition, in FIG. 30 , the illustrated circle may indicate an area in which blur appears heavily.

In the division of regions by semantic segmentation, since the encoding apparatus 1600 can access the original image itself, it is possible to classify the original image by applying the semantic segmentation technique to the original image, and A region and a region for SR prediction can be distinguished.

However, the target of the SR copy may be a reconstructed high-quality image generated using the reconstructed low-quality image rather than the original image. Therefore, it may be undesirable to copy the reconstructed high-quality image as it is.

In order to solve the problem of copying, in an embodiment, the complexity of an image may be derived using edge power as a path for adjusting image quality.

In the encoding and/or decoding method of the embodiment, the adjustment between the bit rate and the image quality may be achieved through adjustment of a quantization parameter. For example, in an environment of a low bit rate, the quantization parameter may be set high.

As a bitstream of a low-quality image encoded using a high quantization parameter is input, an image used to generate a reconstructed high-quality image may be an image that has already been encoded, or an image to which encoding artifacts according to encoding are added. have.

Even if a reconstructed high-quality image is generated using an image to which such encoding artifacts are added, the quality of the reconstructed high-quality image used as the prediction image may have already deteriorated. In particular, in the reconstructed high-quality image, blur may occur severely. When the reconstructed high-quality image in this state is copied as it is by using the SR radiation as a prediction image, the quality of the reconstructed high-quality image generated by the SR radiation may be seriously deteriorated. Therefore, the method to be described below can be used to prevent such phenomena such as deterioration.

In the encoding and/or decoding method of the embodiment, a low-resolution reconstructed image given as encoding is performed in a low-bit-rate environment may also be assumed to be a bitstream generated by encoding in a low-bit-rate environment.

When an image for a low bit rate is reconstructed with a high resolution, an image generated by such reconstruction may be significantly different from an image reconstructed in a high bit rate environment. Since image quality degradation due to encoding of a low-resolution image has already occurred, even if the low-resolution image is restored to a high resolution, especially in a high-frequency region such as a texture region, blur may be severe.

32 illustrates a region in which edge power exceeding a threshold value occurs when a reconstructed high-quality image is generated using a low-resolution image with a low bit rate according to an example.

32 shows a reconstructed high-quality image generated using a low-resolution image with a low bit rate according to an example. In FIG. 31 , a circle may indicate a region in which an edge power greater than or equal to a threshold has occurred.

The area indicated by the blur may be excluded from the area to which the copy is applied. In order not to designate the area where the blur appears as the area of the SR copy, the area where the blur appears needs to be filtered out of the map. In other words, the map may indicate that the area where the blur appears is not the area to which the copy is applied.

For this exclusion, in the embodiment, after generating a reconstructed high-quality image using the reconstructed low-quality image, for each block of the blocks in the reconstructed high-quality image, the edge power of each block may be calculated.

Only if the edge power of the block is greater than a specified threshold can SR copy be selected for a region of the block. For example, if the edge power of the block is greater than a specified threshold, then SR copy may be selected for the region of the block. If the edge power of the block is below a specified threshold, then SR prediction may be selected for the region of the block.

The edge power may be a value generated by dividing the number of edges in the block by the size of the block. The edge power may be generated as in Equation 3 below.

[Equation 3]

In Equation 3, N may represent the width of the block or the height of the block.

Number of Edge may indicate the number of edges in a block.

NxN may be replaced by the product of the width of the block and the height of the block.

As shown in FIG. 31 , after a reconstructed high-quality image is generated using the reconstructed low-quality image of a low bit rate, edge power is calculated on the reconstructed high-quality image, and only a region having an edge power higher than a threshold value can be displayed. A region having an edge power higher than this threshold may be considered and extracted as a texture region in which blur occurs after SR restoration.

33, 34, 35 and 36 show images for classifying a texture region according to an example.

33 shows an original image according to an example.

34 shows a high-resolution image reconstructed by SR reconstruction according to an example.

35 illustrates edge power of a reconstructed high-resolution image according to an example.

36 illustrates a result of division of regions for a reconstructed high-resolution image according to an example.

A map for performing differential encoding for each region according to an embodiment may be generated by synthesizing a plurality of different maps. In other words, the map may be generated by synthesizing 1) the above-described map by semantic segmentation of an image, and 2) an edge power map for improving image quality degradation due to encoding and SR reconstruction. The map generated by the synthesis can be used to generate the reconstructed image.

Here, the synthesis of the plurality of maps may be determined based on values indicated by the plurality of maps with respect to a specified area. For example, in a map generated by synthesizing a plurality of maps, a value for a specified region of the map generated by synthesizing uses values for the specified region of the plurality of maps used for compositing. It can be determined by a specified calculation formula. For example, the specified formula may be a formula for deriving an average value, a median value, a minimum value, a maximum value, a mode, and the like.

For example, the SR copy may be applied in a map generated by synthesis to an area to which the SR copy is commonly applied in all of the plurality of maps. In a map generated by synthesis, SR prediction may be applied to a region to which SR copy is not applied in any one of the plurality of maps.

Alternatively, SR prediction may be applied to a region to which SR prediction is commonly applied in all of the plurality of maps in a map generated by synthesis. In an area to which SR prediction is not applied in any one of the plurality of maps, SR copy may be applied in a map generated by synthesis.

For example, an area to which AR radiation is commonly applied in all of the plurality of maps may be applied to AR radiation in a map generated by synthesis. AR prediction may be applied to an area to which AR radiation is not applied in any one of the plurality of maps in a map generated by synthesis.

Alternatively, AR prediction may be applied to a region to which AR prediction is commonly applied in all of the plurality of maps in a map generated by synthesis. AR radiation may be applied to an area to which AR prediction is not applied in any one of the plurality of maps in a map generated by synthesis.

As another embodiment that utilizes semantic segmentation information, a smooth area such as sky may be additionally included as an area to which SR radiation is applied. However, in a flat area within the reconstructed high-quality image, a phenomenon in which the blocking artifact becomes more prominent may occur. In particular, in a low bit rate environment, blocking artifacts may be severely represented in a flat area, and blocking artifacts may be alleviated when SR copy from a reconstructed high-quality image is applied.

According to this characteristic, the map may be modified based on the semantic classification in the image.

For example, an area to which an SR copy of a map is applied may be increased using semantic classification according to semantic segmentation, and the bit amount for an image may be reduced through this increase.

37 is a flowchart of generation and use of a map for classification of regions according to an embodiment.

In operation 3710, the context of the image may be identified by applying semantic segmentation to the original image.

Step 3710 may include steps 3711 , 3712 , and 3713 .

In operation 3711, a high level context for the image may be identified.

In operation 3712, the depth of the map may be determined so that an encoding method suitable for the identified image region may be selected.

In step 3713, an SR scheme suitable for the encoding characteristic of the region may be selected.

In operation 3720, SR may be performed according to the selected encoding method. An SR image may be generated by SR.

Step 3720 may include steps 3721 and 3722 .

In step 3721, an SR suitable for SR prediction may be performed. That is, SR suitable for use as a prediction image of DCT-based transform encoding may be performed.

At step 3722, an SR suitable for SR copy may be performed.

In operation 3730 , synthesis may be performed on the SR image in operation 3620 .

In operation 3731 , a high-level context for an image may be identified.

In operation 3732, the depth of the map may be determined so that an encoding method suitable for the identified image region may be selected.

In step 3733, an SR scheme suitable for the encoding characteristics of the region may be selected, and synthesis using the selected SR scheme may be performed.

In operation 3740, encoding may be performed on a region of the image using the SR image to which the synthesis is applied.

Step 3740 may include steps 3741 and 3742 . Steps 3741 and 3742 may be performed for each block in the SR image.

In step 3741, if the SR copy is applied to the block, it may be indicated that the copy is performed for the block through a flag for the block.

At step 3742 , if SR prediction is applied to the block, a residual signal may be generated for the block, and coded information may be generated by encoding the residual signal. The residual signal may be a result of subtracting the SR image from the original image for the block region. Encoding may include transform and quantization using frequency domain functions.

The above-described description of encoding may be applied to the encoding apparatus 1600 , and in the decoding apparatus 1700 , the above-described description of encoding may be replaced with a description of decoding.

38 shows syntax for signaling of a type of a base bitstream according to an example.

In an embodiment, one of the inputs to the encoding apparatus 1600 and/or the decoding apparatus 1700 may be a primary encoded bitstream for reconstructing a high-quality image.

The base bitstream may mean a primary encoded bitstream for reconstructing a high-quality image.

The type of the base bitstream may indicate whether the base bitstream is a low-resolution bitstream for a low-resolution image and a low-bit-rate bitstream for a low-bitrate image.

The type of the base bitstream may indicate whether the base bitstream includes an encoded low-resolution image and an encoded low-bitrate image.

38 may indicate syntax for signaling the type of the base bitstream. In FIG. 38 , the type of the base bitstream is included as a part of the sequence parameter set Raw Byte Sequence Payload (RBSP) syntax.

A syntax element (eg, base_bitstream_type) indicating the type of the base bitstream may be included in the sequence parameter.

For example, the type of the base bitstream may be signaled from the encoding apparatus 1600 to the decoding apparatus 1700 as a flag. For interworking with the existing encoding/decoding method, the syntax of the coding unit may be modified as shown in FIG. 38 .

The flag may correspond to the above-described type information of the first bitstream.

When base_bitstream_type is “1”, a prediction image may be generated using the SR network, and the prediction image may be encoded/decoded using the SR network. Using the SR network may mean that SR copy and SR prediction are used.

When base_bitstream_type is “0”, a prediction image may be generated using the AR network, and the prediction image may be encoded/decoded using the AR network. Using an AR network may mean that AR copy and AR prediction are used.

39 shows signaling information for coding of SR copy and SR prediction according to an example.

After the image is divided into semantic regions, signaling using a bitstream from the encoding apparatus 1600 to the decoding apparatus 1700 may be used for encoding and/or decoding using a method suitable for the characteristics of each region. .

For example, when copying is used, signaling through a flag may be used, and when prediction is used, the syntax for the target unit may be modified as illustrated in FIG. 39 so that it can be linked with an existing encoding and/or decoding method. have. Here, the target unit may be a CU as shown. Also, elements of this modified syntax may be applied for other units such as CTUs, PUs and TUs.

The elements of the syntax illustrated in FIG. 39 may have the following meanings. x0 and y0 may be unit coordinates. In other words, x0 and y0 may specify the unit to which the elements apply.

- SRARcopy_flag[x0][y0]: SRARcopy_flag may be a flag indicating whether copying is applied to a unit.

For example, when the value of SRARcopy_flag for a unit is “1”, SR copy may be applied to the unit. In other words, a unit in the reconstructed high-quality image can be generated by SR copying a block of the same position in the reference image. The reference image may be a reconstructed high-resolution image. For example, when the value of SRARcopy_flag for a unit is “0”, SR prediction may be applied to the unit. In other words, the unit in the reconstructed high-quality image may be generated through decoding on the encoded residual signal. The encoded residual signal may be generated by encoding a difference between the predicted image and the original image.

- intra_chroma_pred_mode[x0][y0]: intra_chroma_pred_mode may indicate a color coding mode when SRAR prediction is applied to a unit.

- rqt_root_cbf: rqt_root_cbf may indicate whether a transform_tree syntax exists for a unit. For example, when the value of rqt_root_cbf is “1”, the transform_tree syntax for the unit may exist.

40 shows the structure of a decoding apparatus according to an embodiment.

The type of the base bitstream may be a low-resolution bitstream or a low-bitrate bitstream.

The low-resolution bitstream may be a bitstream generated by encoding an image having a lower resolution than that of the original image.

The low bit rate bitstream may be a bitstream generated by encoding an original image at a low bit rate.

The type of the base bitstream may be identified by a flag in the base bitstream.

A flag indicating the type of the base bitstream may be signaled for each specified unit. The specified unit may be a sequence and/or a frame.

The additional bitstream may include encoded additional information about the reconstructed image generated using the base bitstream.

Additional bitstreams may optionally be used. Whether an additional bitstream is used may be determined according to a type of the base bitstream and a neural network used for the base bitstream.

41 is a flowchart of a decoding method using a base bitstream according to an example.

In operation 4110, the processing unit 1710 may identify the type of the base bitstream.

For example, the processing unit 1710 may identify a flag for a specified unit, and may identify a base bitstream type according to a value of the flag.

In operation 4120, the processing unit 1710 may perform decoding on the image according to the type of the base bitstream.

Decoding may be performed for each block of an image. That is, the processing unit 1710 may perform decoding on a block of an image according to the type of the base bitstream.

42 is a flowchart of a decoding method for a low-resolution bitstream according to an example.

Step 4120 described above with reference to FIG. 41 may include steps 4210 , 4220 , 4230 , 4240 , and 4250 .

In one embodiment, steps 4210 , 4220 , 4230 , 4240 and 4250 may be performed when the base bitstream is a low resolution bitstream.

In operation 4210, the processing unit 1710 may perform decoding on the base bitstream to obtain a reconstructed low-resolution image. The resolution of the reconstructed low-resolution image may be smaller than the resolution of the original image.

In operation 4220, the processor 1710 may determine a reconstruction type of a block of a reconstructed low-resolution image. A reconstruction type may be determined for each block of a plurality of blocks of an image.

For example, the reconstruction type of the block may indicate whether the block is an MSE SR block or a texture SR block.

The MSE SR block may be a block to which SR prediction is applied. The texture SR block may be a block to which SR copy is applied.

In operation 4230, the processing unit 1710 may select an SR network according to a reconstruction type of the block, and may generate an SR image by performing SR reconstruction on the block according to the selected SR network. SR restoration may refer to restoration/reconstruction using SR copy and/or SR prediction.

If the block is an MSE SR block, an MSE SR network may be selected.

A reconstructed low-resolution image may be input to the MSE SR network. When a reconstructed low-resolution image (or a block of the reconstructed low-resolution image) is input, the MSE SR network may output an SR image (or a block of the SR image) having the same resolution as that of the original image. The SR image may be the reconstructed high-resolution image described above.

If the block is a texture SR block, a GAN SR network may be selected.

A reconstructed low-resolution image may be input to the GAN SR network. When a reconstructed low-resolution image (or a block of the reconstructed low-resolution image) is input, the GAN SR network may output an SR image (or a block of the SR image) having the same resolution as that of the original image. The SR image may be the reconstructed high-resolution image described above.

If the block is an MSE SR block, steps 4240 and 4250 may be performed.

In operation 4240, the processing unit 1710 may perform decoding on the additional bitstream. The processor 1710 may perform decoding on the additional bitstream to generate a residual image (or a block of the residual image).

In operation 4250 , the processor 1710 may generate a reconstructed high-quality image by adding the residual image to the SR image.

43 is a flowchart of a decoding method for a low bit rate bitstream according to an example.

Step 4120 described above with reference to FIG. 41 may include steps 4310 , 4320 , 4330 , 4340 , and 4350 .

In one embodiment, steps 4310 , 4320 , 4330 , 4340 and 4350 may be performed when the base bitstream is a low bitrate bitstream.

In operation 4310, the processor 1710 may perform decoding on the base bitstream to obtain a reconstructed low-bit-rate image. The bit rate of the reconstructed low bit rate image may be smaller than the bit rate of the original image.

In operation 4320 , the processor 1710 may determine a reconstruction type of a block of a reconstructed low bit rate image. A reconstruction type may be determined for each block of a plurality of blocks of an image.

For example, the reconstruction type of the block may indicate whether the block is an MSE AR block or a texture AR block.

The MSE AR block may be a block to which AR prediction is applied. The texture AR block may be a block to which AR copy is applied.

In operation 4230, the processing unit 1710 may select an AR network according to the reconstruction type of the block, and may generate an AR image by performing AR reconstruction on the block according to the selected AR network. AR restoration may refer to restoration/reconstruction using AR copy and/or AR prediction.

If the block is an MSE AR block, an MSE AR network may be selected.

A reconstructed low-resolution image may be input to the MSE AR network. When a reconstructed low-resolution image (or a block of the reconstructed low-resolution image) is input, the MSE AR network may output an AR image (or a block of the AR image) from which coding artifacts are removed. The AR image may be the above-described coding distortion correction image.

If the block is a texture AR block, a GAN AR network may be selected.

A reconstructed low-resolution image may be input to the GAN AR network. When a reconstructed low-resolution image (or a block of a reconstructed low-resolution image) is input, the GAN AR network removes the coding artifacts, block) can be printed. The image in which the coding artifact is removed and the high frequency region is reconstructed may be the above-described coding distortion correction image.

If the block is an MSE AR block, steps 4340 and 4350 may be performed.

In operation 4340, the processing unit 1710 may perform decoding on the additional bitstream. The processor 1710 may perform decoding on the additional bitstream to generate a residual image (or a block of the residual image).

In operation 4350 , the processor 1710 may generate a reconstructed high-quality image by adding the residual image to the AR image from which the coding artifact has been removed.

The above embodiments may be performed in the same and/or corresponding manner in the encoding apparatus 1600 and the decoding apparatus 1700 . Also, a combination of one or more of the above embodiments may be used in encoding and/or decoding an image.

The order in which the above embodiments are applied may be different from each other in the encoding apparatus 1600 and the decoding apparatus 1700 . Alternatively, the order in which the above embodiments are applied may be (at least partially) the same in the encoding apparatus 1600 and the decoding apparatus 1700 .

The order of applying the embodiment may be different in the encoding apparatus 1600 and the decoding apparatus 1700 , and the order of applying the embodiment may be the same in the encoding apparatus 1600 and the decoding apparatus 1700 .

The above embodiments may be performed for each of the luma signal and the chroma signal. The above embodiments may be performed in the same manner for the luma signal and the chroma signal.

The shape of the block to which the embodiments of the present invention are applied may have a square shape or a non-square shape.

The above embodiments of the present invention may be applied according to the size of at least one of a target block, a coding block, a prediction block, a transform block, a block, a current block, a coding unit, a prediction unit, a transform unit, a unit, and a current unit. Here, the size may be defined as a minimum size and/or a maximum size to which the embodiments are applied, or may be defined as a fixed size to which the embodiments are applied. Also, in the above embodiments, the first embodiment may be applied to the first size, and the second embodiment may be applied to the second size. That is, the regular embodiments may be complexly applied according to the size. In addition, the above embodiments of the present invention may be applied only when the minimum size or more and the maximum size or less. That is, the above embodiments may be applied only when the block size is included within a certain range.

In addition, the above embodiments of the present invention can be applied only when the condition greater than or equal to the minimum size and the condition less than or equal to the maximum size are satisfied, where the minimum size and the maximum size are the blocks described above in the embodiment and the above in the embodiment, respectively. It can be the size of one of the units. That is, the block to be the target of the minimum size and the block to be the target of the maximum size may be different from each other. For example, the above embodiments of the present invention can be applied only when the size of the target block is greater than or equal to the minimum size of the block and less than or equal to the maximum size of the block.

For example, the above embodiments can be applied only when the size of the target block is 8x8 or more. For example, the above embodiments can be applied only when the size of the target block is 16x16 or more. For example, the above embodiments can be applied only when the size of the target block is 32x32 or more. For example, the above embodiments may be applied only when the size of the target block is 64x64 or more. For example, the above embodiments can be applied only when the size of the target block is 128x128 or more. For example, the above embodiments can be applied only when the size of the target block is 4x4. For example, the above embodiments can be applied only when the size of the target block is 8x8 or less. For example, the above embodiments can be applied only when the size of the target block is 16x16 or less. For example, the above embodiments may be applied only when the size of the target block is 8x8 or more and 16x16 or less. For example, the above embodiments may be applied only when the size of the target block is 16x16 or more and 64x64 or less.

The above embodiments of the present invention may be applied according to a temporal layer. A separate identifier may be signaled to identify a temporal layer to which the embodiments are applicable, and the embodiments may be applied to a temporal layer specified by the corresponding identifier. The identifier herein may be defined as the lowest layer and/or the highest layer to which the embodiment is applicable, or may be defined to indicate a specific layer to which the embodiment is applied. In addition, a fixed temporal layer to which the above embodiment is applied may be defined.

For example, the above embodiments may be applied only when the temporal layer of the target image is the lowest layer. For example, the above embodiments may be applied only when the temporal layer identifier of the target image is 1 or more. For example, the above embodiments may be applied only when the temporal layer of the target image is the highest layer.

A slice type or tile group type to which the embodiments of the present invention are applied may be defined, and the embodiments of the present invention may be applied according to the corresponding slice type or tile group type.

In the above-described embodiments, in applying a specified processing to a specified object, a specified condition may be required, and when the specified processing is described as being processed under a specified determination, the specified coding parameter is If it has been described that whether or not a specified condition is satisfied based on it is determined, or that a specified determination is made based on a specified coding parameter, it may be construed that the specified coding parameter may be replaced with another coding parameter. In other words, it is to be understood that the specified conditions or coding parameters influencing the specified determination may only be regarded as exemplary, and that in addition to the specified coding parameters, a combination of one or more other coding parameters performs the role of the specified coding parameters. can

In the above-described embodiments, the methods are described on the basis of a flowchart as a series of steps or units, but the present invention is not limited to the order of steps, and some steps may occur in a different order or concurrently with other steps as described above. can In addition, those of ordinary skill in the art will recognize that the steps shown in the flowchart are not exclusive, other steps may be included, or that one or more steps in the flowchart may be deleted without affecting the scope of the present invention. You will understand.

The above-described embodiments include examples of various aspects. Although not every possible combination for representing the various aspects may be described, one of ordinary skill in the art will recognize that combinations other than those explicitly described are possible. Accordingly, it is intended that the present invention cover all other substitutions, modifications and variations falling within the scope of the following claims.

The embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded in a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the computer readable recording medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the computer software field.

The computer-readable recording medium may contain information used in the embodiments according to the present invention. For example, the computer-readable recording medium may include a bitstream, and the bitstream may include the information described in the embodiments according to the present invention.

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

Examples of the computer-readable recording medium include a hard disk, a magnetic medium such as a floppy disk and a magnetic tape, an optical recording medium such as a CD-ROM, a DVD, and a magneto-optical medium such as a floppy disk. media), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules to perform processing according to the present invention, and vice versa.

In the above, the present invention has been described with specific matters such as specific components and limited embodiments and drawings, but these are provided to help a more general understanding of the present invention, and the present invention is not limited to the above embodiments. , those of ordinary skill in the art to which the present invention pertains can make various modifications and variations from these descriptions.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments, and not only the claims described below, but also all modifications equivalently or equivalently to the claims described below belong to the scope of the spirit of the present invention. will do it

Claims (20)

generating a reconstructed low-quality image by encoding the original image;
generating a reconstructed high-quality image by using the reconstructed low-quality image;
dividing the image into a plurality of regions;
generating encoded reconstruction information by performing encoding on the region; and
Generating a bitstream including encoded reconstruction information
including,
the region is at least one of the plurality of regions;
wherein the image is the original image, the restored low-quality image, or the restored high-quality image. According to claim 1,
The bitstream includes an encoded low-quality image,
The encoded low-quality image is an encoded low-resolution image or an encoded low-bit rate image,
The encoding method, wherein the bitstream includes type information indicating which image of the encoded low-resolution image and the encoded low-bit-rate image is included in the bitstream. 3. The method of claim 2,
If the encoded low-quality image is the encoded low-resolution image, the reconstructed high-quality image is a reconstructed high-resolution image,
If the encoded low-quality image is the encoded low-bit rate image, the reconstructed high-quality image is a coding distortion correction image. According to claim 1,
By performing classification on the image, a map for the image is generated,
The map indicates a plurality of areas,
An encoding method, wherein different encoding methods are respectively used for a plurality of regions. 5. The method of claim 4,
an edge power of the image is calculated, and the map is modified based on the edge power. 5. The method of claim 4,
wherein the map is modified based on a semantic classification within the image. 5. The method of claim 4,
If the encoded low-quality image is an encoded low-resolution image, differential reconstruction from a low-resolution to a high-resolution image is performed based on the map. 5. The method of claim 4,
If the encoded low-quality image is an encoded low-bit rate image, differential correction is performed on coding distortion generated in the image based on the map. 5. The method of claim 4,
The encoding method, wherein each encoding is performed using different neural networks for the plurality of regions. A recording medium for recording the bitstream generated by the encoding method according to claim 1. Receiving a bitstream including an encoded low-quality image;
generating a reconstructed low-quality image by performing decoding on the encoded low-quality image;
generating a reconstructed high-quality image by using the reconstructed low-quality image;
dividing the image into a plurality of regions;
generating reconstruction information by performing decoding on the region; and
generating a reconstructed image using the reconstruction information
including,
the region is at least one of the plurality of regions;
The image is the reconstructed low-quality image or the reconstructed high-quality image. 12. The method of claim 11,
The encoded low-quality image is an encoded low-resolution image or an encoded low-bit rate image,
The bitstream includes type information indicating which image of the encoded low-resolution image and the encoded low-bit-rate image is included in the bitstream. 13. The method of claim 12,
If the encoded low-quality image is the encoded low-resolution image, the reconstructed high-quality image is a reconstructed high-resolution image,
If the encoded low-quality image is the encoded low-bit rate image, the reconstructed high-quality image is a coding distortion correction image. 12. The method of claim 11,
By performing classification on the image, a map for the image is generated,
The map indicates a plurality of areas,
A decoding method, wherein different decoding methods are respectively used for a plurality of regions. 15. The method of claim 14,
and an edge power of the image is calculated, and the map is modified based on the edge power. 15. The method of claim 14,
The decoding method, wherein the map is modified based on a semantic classification in the image. 15. The method of claim 14,
If the encoded low-quality image is an encoded low-resolution image, differential restoration from a low resolution to a high resolution is performed on the image based on the map. 15. The method of claim 14,
If the encoded low-quality image is an encoded low-bit rate image, differential correction is performed on coding distortion generated in the image based on the map. 15. The method of claim 14,
The decoding method, wherein each decoding is performed using different neural networks for the plurality of regions. A computer-readable recording medium for storing a bitstream, the bitstream comprising:
Encoded low quality video
including,
A reconstructed low-quality image is generated by performing decoding on the encoded low-quality image,
A reconstructed high-quality image is generated using the reconstructed low-quality image,
The image is divided into a plurality of regions,
Reconstruction information is generated by performing decryption on the area,
A reconstructed image is generated using the reconstruction information,
the region is at least one of the plurality of regions;
wherein the image is the restored low-quality image or the restored high-quality image.

Images