KR20230074682A - Method for encoding/decoding video and apparatus thereof - Google Patents

Abstract

The present invention relates to video encoding and decoding methods. The video decoding method comprises the steps of: deriving a spatial merge candidate from at least one spatial neighboring block of a current block; deriving a temporal merge candidate from the corresponding position block of the current block; and generating a prediction block of the current block based on at least one of the derived spatial merge candidate and the derived temporal merge candidate. A reference image for the temporal merge candidate may be selected based on a reference image list of a current image including the current block and a reference image list of a corresponding position image including the corresponding position block. Accordingly, the present invention provides a method and device for performing motion compensation using combined merge candidates to improve video encoding/decoding efficiency.

Description

Video encoding/decoding method and apparatus {METHOD FOR ENCODING/DECODING VIDEO AND APPARATUS THEREOF}

The present invention relates to a method and apparatus for encoding/decoding an image, and more particularly, to a method and apparatus for performing motion compensation using a merge mode.

Recently, demand for high-resolution and high-quality images such as high definition (HD) images and ultra high definition (UHD) images is increasing in various application fields. As image data becomes higher resolution and higher quality, the amount of data increases relatively compared to existing image data. Therefore, when image data is transmitted using a medium such as an existing wired/wireless broadband line or stored using an existing storage medium, transmission cost and Storage costs increase. In order to solve these problems caused by high-resolution and high-quality video data, high-efficiency video encoding/decoding technology for video with higher resolution and quality is required.

An inter-prediction technique for predicting pixel values included in the current picture from pictures before or after the current picture as an image compression technique, an intra-prediction technique for predicting pixel values included in the current picture using pixel information within the current picture, There are various technologies such as transformation and quantization technology for compressing the energy of the residual signal, and entropy encoding technology that assigns short codes to values with high frequency of occurrence and long codes to values with low frequency of occurrence. Image data can be effectively compressed and transmitted or stored.

In the conventional motion compensation using merge mode, only spatial merge candidates, temporal merge candidates, bi-predictive merge candidates, and zero merge candidates are added to the merge candidate list and used. Since only unidirectional prediction and bidirectional prediction are used, there is a limit to improving the coding efficiency. .

In the conventional motion compensation using merge mode, there is a dependency between the temporal merge candidate derivation process and the bi-predictive merge candidate derivation process, so the throughput of the merge mode is limited, and it is difficult to perform each merge candidate derivation process in parallel. there is.

In the motion compensation using the conventional merge mode, since a bi-predictive merge candidate is generated using a bi-predictive merge candidate derivation process and used as motion information, memory access bandwidth increases during motion compensation compared to a uni-predictive merge candidate.

In the conventional motion compensation using merge mode, since the derivation of zero merge candidates is performed differently depending on the slice type, the hardware logic becomes complicated. As a result, the memory access bandwidth increases.

The present invention may provide a method and apparatus for performing motion compensation using a combined merge candidate to improve video encoding/decoding efficiency.

The present invention may provide a method and apparatus for performing motion compensation using unidirectional prediction, bidirectional prediction, 3-way prediction, and 4-way prediction to improve video encoding/decoding efficiency.

The present invention parallelizes each merge candidate derivation process, removes dependencies between merge candidate derivation processes, divides bi-prediction merge candidates, and deduces uni-predictive zero-merge candidates to increase merge mode throughput and simplify hardware logic. Provides a method and apparatus for determining

In the present invention, when a temporal merge candidate is derived from a corresponding position block in a corresponding position image (call picture) corresponding to a current block, a reference picture related to a motion vector derived from the corresponding position block is used as a reference picture for the temporal merge candidate. It provides a method and apparatus for doing so.

According to the present invention, a video decoding method includes deriving a spatial merge candidate from at least one spatial neighboring block of a current block, deriving a temporal merge candidate from a block corresponding to the current block, and the derived spatial merge candidate. and generating a prediction block of the current block based on at least one of the temporal merge candidates, wherein reference images for the temporal merge candidate include a reference image list of a current image including the current block and the corresponding position block. It may be selected based on a reference image list of corresponding location images including .

In the image decoding method, a reference image for the temporal merge candidate may be selected based on whether a reference image list of the current image and a reference image list of the corresponding position image are identical.

In the image decoding method, when the reference image list of the current image and the reference image list of the corresponding location image are identical to each other, the reference image for the temporal merge candidate is a reference used by a motion vector derived from the corresponding location block. It can be selected as an image.

In the image decoding method, when at least one reference image is the same in the reference image list of the current image and the reference image list of the corresponding location image, the reference image for the temporal merge candidate is selected from among the at least one identical reference image. Either one can be selected.

In the video decoding method, each reference video for the temporal merge candidate may be selected according to an inter-prediction direction.

In the video decoding method, the spatial merge candidate and the temporal merge candidate of the current block may be derived in units of sub-blocks of the current block.

In the video decoding method, a temporal merge candidate of a sub-block of the current block may be derived from a sub-block at the same position as a sub-block of the current block included in the corresponding block.

In the video decoding method, when the co-located sub-block is unavailable, the temporal merge candidate of the sub-block of the current block is a sub-block at the center of the corresponding-located block and a left sub-block of the co-located sub-block. , It can be derived from any one of the upper sub-blocks of the co-located sub-block.

In the image decoding method, the step of deriving the temporal merge candidate may include scaling a motion vector of the corresponding block based on each of the reference images of the reference image list of the current block, and the plurality of scaled motion vectors. A temporal merge candidate including can be derived.

In the video decoding method, a prediction block of the current block may be generated using a motion vector generated based on a weighted sum of the plurality of scaled motion vectors.

In the video decoding method, a plurality of temporary prediction blocks may be generated using each of the plurality of scaled motion vectors, and a prediction block of the current block may be generated based on a weighted sum of the generated plurality of temporary prediction blocks. there is.

In the video decoding method, the deriving of the temporal merge candidate may include deriving the temporal merge candidate by scaling motion information of the corresponding location block based on a reference video for the temporal merge candidate.

In the video decoding method, deriving the temporal merge candidate by scaling motion information of the corresponding position block comprises: an image order count value between a current video including the current block and a reference image of the current block and the corresponding position It may be selectively performed based on an image order count value between the corresponding location image including the block and the reference image of the corresponding location block.

According to the present invention, an image encoding method includes deriving a spatial merge candidate from at least one spatial neighboring block of a current block, deriving a temporal merge candidate from a block corresponding to the current block, and the derived spatial merge candidate. and generating a prediction block of the current block based on at least one of the temporal merge candidates, wherein reference images for the temporal merge candidate include a reference image list of a current image including the current block and the corresponding position block. It may be selected based on a reference image list of corresponding location images including .

According to the present invention, an image decoding apparatus derives a spatial merge candidate from at least one spatial neighboring block of a current block, and derives a temporal merge candidate from a block corresponding to the current block, and uses the derived spatial merge candidate and the temporal merge candidate. An inter-prediction unit generating a prediction block of the current block based on at least one of merge candidates, wherein the inter-prediction unit converts a reference image for the temporal merge candidate to a reference image of a current image including the current block. Selection can be made based on the list and the reference image list of the corresponding position image including the corresponding position block.

According to the present invention, an image encoding apparatus derives a spatial merge candidate from at least one spatial neighboring block of a current block, and derives a temporal merge candidate from a corresponding block of the current block, and derives a spatial merge candidate from a corresponding block of the current block and a temporal merge candidate. An inter-prediction unit generating a prediction block of the current block based on at least one of merge candidates, wherein the inter-prediction unit converts a reference image for the temporal merge candidate to a reference image of a current image including the current block. Selection can be made based on the list and the reference image list of the corresponding position image including the corresponding position block.

According to the present invention, a recording medium storing a bitstream includes deriving a spatial merge candidate from at least one spatial neighboring block of a current block, deriving a temporal merge candidate from a block corresponding to the current block, and the derivation generating a prediction block of the current block based on at least one of a spatial merge candidate and a temporal merge candidate, wherein a reference image for the temporal merge candidate is a reference image list of a current image including the current block, and A bitstream generated by an image encoding method characterized in that the selection is made based on a reference image list of the corresponding location image including the corresponding location block may be stored.

In the present invention, a method and apparatus for performing motion compensation using merge candidates combined to improve video encoding/decoding efficiency are provided.

In the present invention, a method and apparatus for performing motion compensation using unidirectional prediction, bidirectional prediction, 3-way prediction, and 4-way prediction are provided to improve the encoding/decoding efficiency of an image.

In the present invention, in order to increase the throughput of the merge mode and simplify the hardware logic, motion compensation is performed by parallelizing each merge candidate derivation process, removing dependencies between merge candidate derivation processes, bi-predictive merge candidate division, and uni-predictive zero-merge candidate derivation. A method and apparatus for performing are provided.

In the present invention, when deriving a temporal merge candidate from a corresponding position block in a collocated picture corresponding to a current block, a method and apparatus for using a reference picture related to a motion vector derived from a corresponding position block as a reference picture for a temporal merge candidate Provided.

1 is a block diagram showing a configuration according to an embodiment of an encoding device to which the present invention is applied.
2 is a block diagram showing a configuration according to an embodiment of a decoding device to which the present invention is applied.
3 is a diagram schematically illustrating a division structure of an image when encoding and decoding an image.
4 is a diagram illustrating a form of a prediction unit (PU) that may be included in a coding unit (CU).
5 is a diagram illustrating a form of a transform unit (TU) that may be included in a coding unit (CU).
6 is a diagram for explaining an embodiment of an inter-screen prediction process.
7 is a flowchart illustrating a method of encoding an image in merge mode according to the present invention.
8 is a flowchart illustrating a video decoding method in merge mode according to the present invention.
9 is a diagram for explaining an example of deriving a spatial merge candidate and a temporal merge candidate of a current block according to the present invention.
10 is a diagram for explaining an example in which a temporal merge candidate is added to a merge candidate list according to the present invention.
11 shows an example of scaling a motion vector of a corresponding position block in order to derive a temporal merge candidate of a current block according to the present invention.
12 and 13 are diagrams for explaining an embodiment of a method of adaptively selecting a reference picture for a temporal merge candidate and a method of generating a prediction block using a temporal merge candidate according to the present invention.
14 and 15 are diagrams for explaining an embodiment of a method of adaptively selecting a reference image for a temporal merge candidate in a low delay encoding/decoding environment according to the present invention.
16 and 17 are diagrams for explaining lower blocks of the corresponding location block according to the present invention.
18 is a diagram for explaining an embodiment of performing motion compensation in sub-block units according to the present invention.
19 is a flowchart illustrating an image decoding method according to an embodiment of the present invention.

Since the present invention can make various changes and have various embodiments, specific embodiments will be illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention. Like reference numbers in the drawings indicate the same or similar function throughout the various aspects. The shapes and sizes of elements in the drawings may be exaggerated for clarity. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS For detailed descriptions of exemplary embodiments described below, reference is made to the accompanying drawings, which illustrate specific embodiments by way of example. These embodiments are described in sufficient detail to enable those skilled in the art to practice the embodiments. It should be understood that the various embodiments are different, but need not be mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be implemented in one embodiment in another embodiment without departing from the spirit and scope of the invention. Additionally, it should be understood that the location or arrangement of individual components within each disclosed embodiment may be changed without departing from the spirit and scope of the embodiment. Accordingly, the detailed description set forth below is not to be taken in a limiting sense, and the scope of the exemplary embodiments, if properly described, is limited only by the appended claims, along with all equivalents as claimed by those claims.

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

When a component of the present invention is referred to as “connected” or “connected” to another component, it may be directly connected or connected to the other component, but there may be other components in the middle. It should be understood that it may be On the other hand, when an element is referred to as “directly connected” or “directly connected” to another element, it should be understood that no other element exists in the middle.

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

Terms used in the present invention are only used to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly dictates otherwise. In the present invention, terms such as "comprise" or "having" are intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, but one or more other features It should be understood that the presence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded. That is, the description of "including" a specific configuration in the present invention does not exclude configurations other than the corresponding configuration, and means that additional configurations may be included in the practice of the present invention or the scope of the technical spirit of the present invention.

Some of the components of the present invention are not essential components that perform essential functions in the present invention, but may be optional components for improving performance. The present invention can be implemented by including only components essential to implement the essence of the present invention, excluding components used for performance improvement, and a structure including only essential components excluding optional components used for performance improvement. Also included in the scope of the present invention.

EMBODIMENT OF THE INVENTION Hereinafter, embodiment of this invention is described concretely with reference to drawings. In describing the embodiments of this specification, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present specification, the detailed description will be omitted, and the same reference numerals will be used for the same components in the drawings. and duplicate descriptions of the same components are omitted.

In addition, hereinafter, an image may mean one picture constituting a video, and may also indicate a video itself. For example, "encoding and/or decoding an image" may mean "encoding and/or decoding a video", and may mean "encoding and/or decoding one of images constituting a video". may be Here, a picture may have the same meaning as an image.

Glossary of Terms

Encoder: May mean a device that performs encoding.

Decoder: This may mean a device that performs decryption.

Parsing: Determining the value of a syntax element by entropy decoding, or entropy decoding itself.

Block: An MxN array of samples, where M and N represent positive integer values, and a block can often refer to a two-dimensional array of samples.

Sample: This is the basic unit constituting a block, and can represent a value from 0 to 2 ^Bd - 1 according to the bit depth (B _d ). In the present invention, pixel and pixel may be used in the same meaning as sample.

Unit: May mean a unit of video encoding and decoding. In encoding and decoding of an image, a unit may be a region created by dividing one image. Also, a unit may refer to a divided unit when encoding or decoding an image is divided into subdivided units. In encoding and decoding of an image, a predefined process may be performed for each unit. One unit can be further divided into sub-units having a smaller size relative to the unit. According to function, units are Block, Macroblock, Coding Tree Unit, Coding Tree Block, Coding Unit, Coding Block, and Prediction. It may mean a prediction unit, a prediction block, a transform unit, a transform block, and the like. In addition, a unit may mean that it includes a Luma component block, a Chroma component block corresponding thereto, and a syntax element for each block to be referred to as a block. The unit may have various sizes and shapes, and in particular, the shape of the unit may include not only a rectangle but also a geometric figure that can be expressed in two dimensions, such as a square, a trapezoid, a triangle, and a pentagon. Also, the unit information may include at least one of a unit type indicating a coding unit, a prediction unit, a transform unit, and the like, a size of the unit, a depth of the unit, and an encoding and decoding order of the unit.

Reconstructed Neighbor Unit: This may refer to a unit that has already been spatially/temporally encoded or decoded around a unit to be encoded/decoded and reconstructed. In this case, the reconstructed neighboring unit may mean a reconstructed neighboring block.

Neighbor block: This may mean a block adjacent to an encoding/decoding target block. A block adjacent to an encoding/decoding object block may mean a block whose boundary abuts with the encoding/decoding object block. A neighboring block may refer to a block located at an adjacent vertex of an encoding/decoding target block. A neighboring block may mean a restored neighboring block.

Unit Depth: It means the degree to which a unit is divided. In a tree structure, the root node has the shallowest depth and the leaf node has the deepest depth.

Symbol: It may mean a unit syntax element to be coded/decoded, a coding parameter, a value of a transform coefficient, and the like.

Parameter Set: Among structures within a bitstream, it may correspond to header information, and includes a video parameter set, a sequence parameter set, a picture parameter set, and an adaptation parameter set. At least one of (adaptation parameter sets) may be included in the parameter set. In addition, the parameter set may have a meaning including slice header and tile header information.

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

Prediction Unit: It is a basic unit when inter-prediction or intra-prediction and compensation are performed, and one prediction unit may be divided into a plurality of small-sized partitions. In this case, each of a plurality of partitions becomes a basic unit for performing the prediction and compensation, and a partition into which a prediction unit is divided may also be referred to as a prediction unit. The prediction unit may have various sizes and shapes, and in particular, the shape of the prediction unit may include not only a rectangle but also a geometric figure that can be expressed in two dimensions, such as a square, a trapezoid, a triangle, and a pentagon.

Prediction Unit Partition: This may mean a form in which a prediction unit is divided.

* Reference Picture List: This may mean a list including one or more reference pictures used for inter-picture prediction or motion compensation. The type of reference image list may include LC (List Combined), L0 (List 0), L1 (List 1), L2 (List 2), L3 (List 3), etc. A list may be used.

Inter Prediction Indicator: In inter-prediction, it can mean the inter-prediction direction (uni-directional prediction, bi-directional prediction, etc.) It may mean the number of reference images used when performing inter-prediction or motion compensation on a block to be encoded/decoded.

Reference Picture Index: This may mean an index for a specific reference picture in a reference picture list. Here, the index may mean an index.

Reference Picture: This may refer to a picture referenced by a specific unit for inter-picture prediction or motion compensation, and a reference picture may also be referred to as a reference picture.

Motion Vector: A 2D vector used for inter-screen prediction or motion compensation, and may mean an offset between an encoding/decoding target image and a reference image. For example, (mvX, mvY) may represent a motion vector, mvX may represent a horizontal component, and mvY may represent a vertical component.

Motion Vector Candidate: When predicting a motion vector, it may mean a unit that becomes a prediction candidate or a motion vector of the unit.

Motion vector candidate list: This may mean a list constructed using motion vector candidates.

Motion vector candidate index: An indicator indicating a motion vector candidate in the motion vector candidate list, which may also be referred to as an index of a motion vector predictor.

Motion information: information including at least one of reference image list information, reference image, motion vector candidate, motion vector candidate index, as well as a motion vector, reference image index, and inter prediction indicator can mean

* Merge Candidate List: This may mean a list constructed using merge candidates.

Merge Candidate: Spatial merge candidate, temporal merge candidate, combined merge candidate, combined predictive merge candidate, zero merge candidate, etc. may be included, and the merge candidate includes prediction type information, each list It may include motion information such as a reference picture index and a motion vector for .

Merge Index: This may mean information indicating a merge candidate in a merge candidate list. Also, the merge index may indicate a block from which a merge candidate is derived, among blocks reconstructed that are spatially/temporally adjacent to the current block. Also, the merge index may indicate at least one piece of motion information of a merge candidate.

Transform Unit: May mean a basic unit when encoding/decoding a residual signal such as transform, inverse transform, quantization, inverse quantization, and transform coefficient encoding/decoding, and one transform unit It can be divided into a plurality of small-sized transform units. The conversion unit may have various sizes and shapes, and in particular, the shape of the conversion unit may include not only a rectangle but also a geometric figure that can be expressed in two dimensions, such as a square, a trapezoid, a triangle, and a pentagon.

Scaling: This may refer to a process of multiplying a transform coefficient level by a factor, and a transform coefficient may be generated as a result. Scaling can also be called dequantization.

Quantization Parameter: May mean a value used when scaling a transform coefficient level in quantization and inverse quantization. In this case, the quantization parameter may be a value mapped to a quantization step size.

Residual quantization parameter (Delta Quantization Parameter): This may mean a difference between a predicted quantization parameter and a quantization parameter of a unit to be encoded/decoded.

Scan: It can mean a method of arranging the order of coefficients in a block or matrix. For example, arranging a 2D array into a 1D array is called a scan, and a 1D array is converted into a 2D array. Alignment can also be called a scan or an inverse scan.

Transform Coefficient: A coefficient value generated after performing a transform. In the present invention, a quantized transform coefficient level obtained by applying quantization to the transform coefficient may also be included in the meaning of the transform coefficient.

Non-zero transform coefficient: This may mean a transform coefficient whose magnitude is not 0 or a transform coefficient level whose magnitude is not 0.

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

Quantization Matrix Coefficient: This may mean each element in a quantization matrix. Quantization matrix coefficients may also be referred to as matrix coefficients.

Default Matrix: This may mean a predetermined quantization matrix predefined in an encoder and a decoder.

Non-default matrix: This may mean a quantization matrix that is not predefined in an encoder and a decoder and is signaled by a user.

Coding Tree Unit: It may be composed of two chrominance component (Cb, Cr) coding tree blocks related to one luminance component (Y) coding tree block. Each coding tree unit may be split using one or more division schemes such as a quad tree or a binary tree to form subunits such as a coding unit, a prediction unit, and a transform unit. It can be used as a term to refer to a pixel block that is a processing unit in the decoding/coding process of an image, like segmentation of an input image.

Coding tree block: It 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.

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

The encoding device 100 may be a video encoding device or an image encoding device. A video may include one or more images. The encoding apparatus 100 may sequentially encode one or more images of a video according to time.

Referring to FIG. 1 , the encoding apparatus 100 includes a motion estimation unit 111, a motion compensation unit 112, an intra prediction unit 120, a switch 115, a subtractor 125, a transform unit 130, and a quantization unit. It may include a unit 140, an entropy encoding unit 150, an inverse quantization unit 160, an inverse transform unit 170, an adder 175, a filter unit 180, and a reference picture buffer 190.

The encoding apparatus 100 may perform encoding on an input image in an intra mode and/or an inter mode. Also, the encoding device 100 may generate a bitstream through encoding of an input image and output the generated bitstream. When the intra mode is used as the prediction mode, the switch 115 may be switched to intra, and when the inter mode is used as the prediction mode, the switch 115 may be switched to inter. Here, the intra mode may mean an intra prediction mode, and the inter mode may mean an inter prediction mode. The encoding apparatus 100 may generate a prediction block for an input block of an input image. Also, the encoding apparatus 100 may encode a residual between the input block and the prediction block after the prediction block is generated. The input image may be referred to as a current image that is a current encoding target. The input block may be referred to as a current block that is a current encoding target or an encoding target block.

When the prediction mode is the intra mode, the intra predictor 120 may use pixel values of blocks already coded adjacent to the current block as reference pixels. The intra predictor 120 may perform spatial prediction using a reference pixel and generate prediction samples for an input block through spatial prediction. Here, intra prediction may mean prediction within a screen.

When the prediction mode is the inter mode, the motion predictor 111 may search for a region that best matches the input block from the reference image in the motion prediction process, and derive a motion vector using the searched region. . A reference image may be stored in the reference picture buffer 190 .

The motion compensation unit 112 may generate a prediction block by performing motion compensation using a motion vector. Here, the motion vector may be a 2D vector used for inter prediction. Also, the motion vector may indicate an offset between the current image and the reference image. Here, inter prediction may mean inter prediction.

When the motion vector value does not have an integer value, the motion estimation unit 111 and the motion compensation unit 112 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, the motion prediction and motion compensation methods of the prediction unit included in the corresponding coding unit based on the coding unit are skip mode, merge mode, and AMVP mode (AMVP mode). ), it is possible to determine which of the current picture reference modes is used, and inter-picture prediction or motion compensation can be performed according to each mode. Here, the current picture reference mode may mean a prediction mode using a pre-reconstructed region in the current picture to which the encoding target block belongs. A motion vector for a current picture reference mode may be defined to specify the pre-reconstructed region. Whether the encoding object block is encoded in the current picture reference mode can be encoded using the reference image index of the encoding object block.

The subtractor 125 may generate a residual block using a difference between the input block and the prediction block. A residual block may be referred to as a residual signal.

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

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

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

The entropy encoding unit 150 may generate a bitstream by performing entropy encoding according to a probability distribution on the values calculated by the quantization unit 140 or the values of coding parameters calculated in the encoding process. Yes, and can output bitstreams. The entropy encoding unit 150 may perform entropy encoding on information for decoding an image in addition to pixel information of the image. For example, information for decoding an image may include a syntax element and the like.

When entropy encoding is applied, a small number of bits are allocated to a symbol with a high probability of occurrence and a large number of bits are allocated to a symbol with a low probability of occurrence to represent the symbol, thereby reducing the number of bits for symbols to be coded. The size of the column may be reduced. Accordingly, compression efficiency of image encoding may be increased through entropy encoding. The entropy encoding unit 150 may use an encoding method such as exponential Golomb, CAVLC (Context-Adaptive Variable Length Coding), or CABAC (Context-Adaptive Binary Arithmetic Coding) for entropy encoding. For example, the entropy encoding unit 150 may perform entropy encoding using a variable length coding/code (VLC) table. In addition, the entropy encoding unit 150 derives a binarization method of the target symbol and a probability model of the target symbol/bin, and then performs arithmetic encoding using the derived binarization method or probability model. You may.

The entropy encoding unit 150 may change a 2D block-type coefficient into a 1-D vector form through a transform coefficient scanning method in order to encode a transform coefficient level. For example, by scanning the coefficients of blocks using up right scanning, they can be changed into a one-dimensional vector form. Depending on the size of the transform unit and the intra-prediction mode, instead of upright scan, vertical scan for scanning 2-dimensional block-type coefficients in a column direction and horizontal scan for scanning 2-dimensional block-type coefficients in a row direction may be used. That is, it is possible to determine which scan method among upright scan, vertical scan, and horizontal scan is used according to the size of the transform unit and the intra-prediction mode.

The coding parameter may include not only information coded in the encoder and signaled to the decoder, such as a syntax element, but also information derived from an encoding or decoding process, and may mean information necessary for encoding or decoding an image. there is. For example, block size, block depth, block division information, unit size, unit depth, unit division information, quad tree division flag, binary tree division flag, binary tree division direction, intra-prediction mode, Intra-prediction direction, reference sample filtering method, prediction block boundary filtering method, filter tap, filter coefficient, inter-prediction mode, motion information, motion vector, reference image index, inter-prediction direction, inter-prediction indicator, reference image list , motion vector predictor, motion vector candidate list, whether to use motion merge mode, motion merge candidate, motion merge candidate list, whether to use skip mode, interpolation filter type, motion vector size, motion vector expression accuracy , transform type, transform size, additional (quadratic) transform information, residual signal presence information, coded block pattern, coded block flag, quantization parameter, quantization matrix, in-loop filter Information, filter application information in loop, filter coefficient in loop, binarization/inverse binarization method, context model, context bin, bypass bin, transform coefficient, transform coefficient level, transform coefficient level scanning method, image display/output order, slice At least one value or a combination of identification information, slice type, slice division information, tile identification information, tile type, tile division information, picture type, bit depth, luminance signal, or color difference signal information may be included in the coding parameter. there is.

The residual signal may mean a difference between the original signal and the prediction signal. Alternatively, the residual signal may be a signal generated by transforming a difference between the original signal and the prediction signal. Alternatively, the residual signal may be a signal generated by transforming and quantizing the difference between the original signal and the prediction signal. The residual block may be a residual signal in units of blocks.

When the encoding apparatus 100 performs encoding through inter prediction, the encoded current image may be used as a reference image for other image(s) to be processed later. Accordingly, the encoding device 100 may decode the encoded current image again and store the decoded image as a reference image. Inverse quantization and inverse transformation of the current image encoded for decoding may be processed.

The quantized coefficient may be dequantized by the inverse quantization unit 160. It may be inversely transformed in the inverse transform unit 170. The inverse quantized and inverse transformed coefficients may be combined with the prediction block through the adder 175. A reconstructed block may be generated by summing the inverse quantized and inverse transformed coefficients and the prediction block.

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

The deblocking filter may remove block distortion generated at a boundary between blocks. In order to determine whether to perform the deblocking filter, it may be determined whether to apply the deblocking filter to the current block based on pixels included in several columns or rows included in the block. When a deblocking filter is applied to a block, a strong filter or a weak filter may be applied according to the required deblocking filtering strength. In addition, in applying the deblocking filter, when vertical filtering and horizontal filtering are performed, horizontal filtering and vertical filtering may be processed in parallel.

The sample adaptive offset may add an appropriate offset value to a pixel value to compensate for a coding error. The sample-adaptive offset may correct an offset from an original image on a pixel-by-pixel basis for an image on which deblocking is performed. In order to perform offset correction for a specific picture, after dividing the pixels included in the image into a certain number of areas, determining the area to be offset and applying the offset to the area, or offset by considering the edge information of each pixel method can be used.

The adaptive loop filter may perform filtering based on a value obtained by comparing a reconstructed image and an original image. After dividing the pixels included in the image into predetermined groups, filtering may be performed differentially for each group by determining one filter to be applied to the corresponding group. Information related to whether or not to apply the adaptive loop filter may be signaled for each coding unit (CU) of the luminance signal, and the shape and filter coefficients of the adaptive loop filter to be applied may vary according to each block. In addition, the same type (fixed type) of the adaptive loop filter may be applied regardless of the characteristics of the block to be applied.

A reconstructed block that has passed through the filter unit 180 may be stored in the reference picture buffer 190 .

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

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

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

The decoding device 200 may receive the bitstream output from the encoding device 100. The decoding apparatus 200 may perform decoding on a bitstream in an intra mode or an inter mode. Also, the decoding apparatus 200 may generate a restored image through decoding and output the restored image.

When the prediction mode used for decoding is the intra mode, the switch may be switched to the intra mode. When the prediction mode used for decoding is the inter mode, the switch may be switched to inter mode.

The decoding apparatus 200 may obtain a reconstructed residual block from the input bitstream and generate a prediction block. When the reconstructed residual block and the prediction block are obtained, the decoding apparatus 200 may generate a reconstructed block that is a decoding target block by adding the reconstructed residual block and the prediction block. A block to be decoded may be referred to as a current block.

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

The entropy decoding unit 210 may change a 1-dimensional vector form coefficient into a 2-dimensional block form through a transform coefficient scanning method in order to decode a transform coefficient level. For example, it can be changed into a 2D block form by scanning coefficients of a block using up right scanning. Depending on the size of the conversion unit and the intra-prediction mode, vertical scan or horizontal scan may be used instead of upright scan. That is, it is possible to determine which scan method among upright scan, vertical scan, and horizontal scan is used according to the size of the transform unit and the intra-prediction mode.

The quantized transform coefficient level may be inversely quantized in the inverse quantization unit 220 and inversely transformed in the inverse transformation unit 230 . As a result of inverse quantization and inverse transformation of the quantized transform coefficient level, a reconstructed residual block may be generated. At this time, the inverse quantization unit 220 may apply a quantization matrix to the quantized transform coefficient level.

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

When the inter mode is used, the motion compensation unit 250 may generate a prediction block by performing motion compensation using a motion vector and a reference image stored in the reference picture buffer 270 . When the motion vector value does not have an integer value, the motion compensator 250 may generate a prediction block by applying an interpolation filter to a partial region in the reference image. In order to perform motion compensation, the motion compensation method of the prediction unit included in the corresponding coding unit based on the coding unit is skip mode, merge mode, AMVP mode, and current picture reference mode. It is possible to determine which method is used, and motion compensation can be performed according to each mode. Here, the current picture reference mode may mean a prediction mode using a pre-reconstructed region in the current picture to which the decoding target block belongs. A motion vector for a current picture reference mode may be used to specify the pre-reconstructed region. A flag or index indicating whether the decoding object block is a block coded in the current picture reference mode may be signaled or may be inferred through a reference picture index of the decoding object block. The current picture in the current picture reference mode may exist at a fixed position (eg, a position with a reference image index of 0 or the last position) within the reference picture list for the decoding target block. Alternatively, it may be variably positioned in the reference picture list, and for this purpose, a separate reference picture index indicating the location of the current picture may be signaled. Here, signaling the flag or index may mean that the encoder entropy-encodes the corresponding flag or index and includes it in a bitstream, and the decoder entropy-decodes the corresponding flag or index from the bitstream ( entropy decoding).

The reconstructed residual block and the predicted block may be added through an adder 255. A block generated by adding the reconstructed residual block and the predicted block may pass through the filter unit 260 . The filter unit 260 may apply at least one of a deblocking filter, a sample adaptive offset, and an adaptive loop filter to a reconstructed block or a reconstructed image. The filter unit 260 may output a reconstructed image. The reconstructed image may be stored in the reference picture buffer 270 and used for inter prediction.

3 is a diagram schematically illustrating a division structure of an image when encoding and decoding an image. 3 schematically illustrates an embodiment 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. Here, the coding unit may mean a coding unit. A unit may be a term collectively referring to 1) a syntax element and 2) a block including image samples. For example, “division of a unit” may mean “division of a block corresponding to a unit”. Block division information may include information about the depth of a unit. Depth information may indicate the number and/or degree of division of a unit.

Referring to FIG. 3 , an image 300 is sequentially divided into units of largest coding units (LCUs), and a division structure is determined in units of LCUs. Here, LCU may be used in the same meaning as a coding tree unit (CTU). One unit may be hierarchically divided with depth information based on a tree structure. Each divided sub-unit may have depth information. Since the depth information indicates the number and/or degree of division of the unit, it may include information about the size of the sub-unit.

The division structure may refer to a distribution of Coding Units (CUs) within the LCU 310 . A CU may be a unit for efficiently encoding/decoding an image. This distribution can be determined according to whether to divide one CU into a plurality of (positive integers of 2 or more including 2, 4, 8, 16, etc.) CUs. The horizontal size and vertical size of the CU created by splitting are half the horizontal size and half the vertical size of the CU before splitting, respectively, or a size smaller than the horizontal size and vertical size of the CU before splitting according to the number of splits. can have The divided CU may be recursively divided into a plurality of CUs having reduced horizontal and vertical sizes in the same manner.

In this case, the division of the CU may be recursively performed up to a predefined depth. Depth information may be information representing the size of a CU and may be stored for each CU. For example, the depth of the LCU may be 0, and the depth of the smallest coding unit (SCU) may be a predefined maximum depth. Here, as described above, the LCU may be a coding unit having the largest coding unit size, and the SCU may be a coding unit having the smallest coding unit size.

The division starts from the LCU 310, and the depth of the CU increases by 1 whenever the horizontal and vertical sizes of the CU are reduced by division. For each depth, a CU that is not split may have a size of 2Nx2N. In the case of a CU to be divided, a 2Nx2N CU may be divided into a plurality of NxN CUs. The size of N is halved for each increase in depth by one.

For example, when one coding unit is divided into 4 coding units, the horizontal and vertical sizes of the divided 4 coding units may be halved compared to the horizontal and vertical sizes of the coding unit before division. there is. For example, when a 32x32 coding unit is divided into 4 coding units, each of the divided 4 coding units may have a size of 16x16. When one coding unit is divided into four coding units, it can be said that the coding units are divided in a quad-tree form.

For example, when one coding unit is divided into two coding units, the horizontal or vertical size of the two divided coding units may be half of the horizontal or vertical size of the coding unit before being divided. . For example, when a coding unit having a size of 32x32 is vertically divided into two coding units, each of the two divided coding units may have a size of 16x32. For example, when a coding unit having a size of 32x32 is horizontally divided into two coding units, each of the two divided coding units may have a size of 32x16. When one coding unit is divided into two coding units, it can be said that the coding unit is divided in the form of a binary tree.

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

Also, information on whether the CU is split may be expressed through split information of the CU. The division information may be 1 bit of information. All CUs except for the SCU may include partition information. For example, if the value of the split information is 0, the CU may not be split, and if the value of the split information is 1, the CU may be split.

4 is a diagram illustrating a form of a prediction unit (PU) that may be included in a coding unit (CU).

Among the CUs divided from the LCU, a CU that is not further divided may be divided into one or more prediction units (PUs). This process may also be referred to as segmentation.

A PU may be a basic unit for prediction. A PU may be encoded and decoded in any one of a skip mode, an inter-screen mode, and an intra-screen mode. A PU may be divided into various types according to modes.

Also, a coding unit is not divided into prediction units, and a coding unit and a prediction unit may have the same size.

* As shown in FIG. 4, in skip mode, division may not exist within the CU. In the skip mode, the 2Nx2N mode 410 having the same size as the CU without division may be supported.

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

One coding unit may be divided into one or more prediction units, and one prediction unit may be divided into one or more prediction units.

For example, when one prediction unit is divided into four prediction units, the horizontal and vertical sizes of the four prediction units may each have half the size compared to the horizontal and vertical sizes of the prediction unit before being divided. there is. For example, when a prediction unit having a size of 32x32 is divided into 4 prediction units, each of the divided 4 prediction units may have a size of 16x16. When one prediction unit is divided into four prediction units, it can be said that the prediction unit is divided in the form of a quad-tree.

For example, when one prediction unit is divided into two prediction units, the horizontal or vertical size of the two prediction units may be half the size compared to the horizontal or vertical size of the prediction unit before being divided. . For example, when a prediction unit having a size of 32x32 is vertically divided into two prediction units, each of the two prediction units may have a size of 16x32. For example, when a prediction unit having a size of 32x32 is horizontally divided into two prediction units, each of the two prediction units may have a size of 32x16. When one prediction unit is split into two prediction units, it can be said that the prediction unit is split in the form of a binary-tree.

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

A transform unit (TU) may be a basic unit used for transform, quantization, inverse transform, and inverse quantization processes within a CU. A TU may have a square shape or a rectangular shape. The TU may be determined dependent on the size and/or shape of the CU.

Among the CUs split from the LCU, a CU that is no longer split into CUs may be split into one or more TUs. In this case, the division structure of the TU may be a quad-tree structure. For example, as shown in FIG. 5 , one CU 510 may be divided one or more times according to a quadtree structure. When one CU is split more than once, it can be said to be split recursively. Through division, one CU 510 may be composed of TUs of various sizes. Alternatively, the CU may be divided into one or more TUs based on the number of vertical lines and/or horizontal lines dividing the CU. A CU may be divided into symmetric TUs or asymmetric TUs. For division into asymmetric TUs, information on the size/shape of TUs may be signaled or may be derived from information on the size/shape of CUs.

Also, the coding unit is not divided into transform units, and the coding unit and the transform unit may have the same size.

One coding unit may be divided into one or more transform units, and one transform unit may be divided into one or more transform units.

For example, when one conversion unit is divided into four conversion units, the horizontal and vertical sizes of the four divided conversion units may each have half the horizontal and vertical sizes of the conversion unit before being divided. there is. For example, when a 32x32 transform unit is divided into 4 transform units, each of the divided 4 transform units may have a size of 16x16. When one transform unit is divided into four transform units, it can be said that the transform unit is divided in a quad-tree form.

For example, when one transform unit is divided into two transform units, the horizontal or vertical size of the two divided transform units may have a half size compared to the horizontal or vertical size of the transform unit before being divided. . For example, when a transform unit having a size of 32x32 is vertically divided into two transform units, each of the two divided transform units may have a size of 16x32. For example, when a transform unit having a size of 32x32 is horizontally divided into two transform units, each of the two divided transform units may have a size of 32x16. When one conversion unit is divided into two conversion units, it can be said that the conversion unit is divided in the form of a binary tree.

When performing transformation, the residual block may be transformed using at least one of a plurality of pre-defined transformation methods. For example, discrete cosine transform (DCT), discrete sine transform (DST), or KLT may be used as a plurality of pre-defined transform methods. Which transform method is applied to transform the residual block may be determined using at least one of inter-prediction mode information of the prediction unit, intra-prediction mode information, and size/shape of the transform block, and in certain cases, the transform method is indicated. Information may be signaled.

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

The intra-prediction mode may be a non-directional mode or a directional mode. The non-directional mode may be a DC mode or a planar mode, and the directional mode may be a prediction mode having a specific direction or angle, and the number may be one or more M. The directional mode may be expressed as at least one of a mode number, a mode value, a mode number, and a mode angle.

The number of intra-prediction modes may be one or more N including the non-directional and directional modes.

The number of intra-prediction modes may vary according to the size of a block. For example, the block size may be 67 in the case of 4x4 or 8x8, 35 in the case of 16x16, 19 in the case of 32x32, and 7 in the case of 64x64.

The number of intra-prediction modes may be fixed to N regardless of the block size. For example, at least one of 35 and 67 may be fixed regardless of the size of the block.

The number of intra-prediction modes may be different according to the type of color component. For example, the number of prediction modes may be different depending on whether the color component is a luma signal or a chroma signal.

In-picture encoding and/or decoding may be performed using sample values or encoding parameters included in adjacent reconstructed blocks.

In order to encode/decode the current block through intra-prediction, a step of checking whether or not samples included in neighboring reconstructed blocks are available as reference samples of the block to be encoded/decoded may be performed. If there are samples that cannot be used as reference samples of a block to be encoded/decoded, copy sample values to samples that cannot be used as reference samples by using at least one of samples included in adjacent reconstructed blocks and/or It may be interpolated and used as a reference sample of a block to be encoded/decoded.

During intra-prediction, a filter may be applied to at least one of reference samples and prediction samples based on at least one of an intra-prediction mode and a size of an encoding/decoding target block. In this case, the encoding/decoding target block may mean a current block, and may mean at least one of a coding block, a prediction block, and a transform block. The type of filter applied to the reference sample or prediction sample may be different according to at least one of the intra prediction mode or size/shape of the current block. The type of filter may vary according to at least one of the number of filter taps, filter coefficient values, and filter strength.

Among the intra-prediction modes, the non-directional planar mode, when generating a prediction block of a target encoding/decoding block, sets the sample value in the prediction block according to the sample position: the top reference sample of the current sample, the left reference sample of the current sample, It can be generated as a weighted sum of the top-right reference sample of the current block and the bottom-left reference sample of the current block.

Among the intra-prediction modes, the non-directional DC mode can be generated as an average value of top reference samples of the current block and left reference samples of the current block when generating a prediction block of a target encoding/decoding block. In addition, filtering may be performed using reference sample values for one or more top rows and one or more left columns adjacent to the reference sample in the encoding/decoding block.

In the case of a plurality of angular modes among intra-prediction modes, prediction blocks may be generated using the upper right and/or lower left reference samples, and the directional modes may have different directivity. Interpolation in units of real numbers may be performed to generate predicted sample values.

To perform the intra-prediction method, the intra-prediction mode of the current prediction block can be predicted from the intra-prediction modes of prediction blocks existing around the current prediction block. When predicting the intra-prediction mode of the current prediction block using the mode information predicted from the prediction mode in the neighboring picture, if the intra-prediction mode of the current prediction block and the neighboring prediction block are the same, the current prediction is made using predetermined flag information Information that the intra-prediction mode of the block and the neighboring prediction block are identical may be signaled, and if the intra-prediction modes of the current prediction block and the neighboring prediction block are different, entropy encoding is performed to intra-prediction of the block to be encoded/decoded. Mode information can be encoded.

7 is a diagram for explaining an embodiment of an inter-screen prediction process.

The rectangle shown in FIG. 7 may represent an image (or picture). Also, arrows in FIG. 7 may indicate prediction directions. That is, an image may be encoded and/or decoded according to a prediction direction. Each image may be classified into an I-picture (Intra Picture), a P-picture (Uni-predictive Picture), and a B-picture (Bi-predictive Picture) according to an encoding type. Each picture may be coded and decoded according to the coding type of each picture.

When an image to be encoded is an I-picture, the image itself may be intra-encoded without inter-prediction. When an image to be encoded is a P picture, the image may be encoded through inter prediction or motion compensation using a reference image only in a forward direction. If an image to be encoded is a B picture, it can be coded through inter prediction or motion compensation using reference pictures in both forward and backward directions, and inter prediction or motion using reference pictures in either forward or backward direction. It can be encoded through compensation. Here, when the inter-prediction mode is used, the encoder may perform inter-prediction or motion compensation, and the decoder may perform motion compensation corresponding thereto. Pictures of P pictures and B pictures that are encoded and/or decoded using reference pictures may be regarded as pictures in which inter prediction is used.

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

Inter-picture prediction or motion compensation may be performed using a reference picture and motion information. Also, inter-picture prediction may use the above-described skip mode.

A reference picture may be at least one of a picture before the current picture or a picture after the current picture. In this case, inter-prediction may perform prediction on a block of the current picture based on the reference picture. Here, the reference picture may mean an image used for prediction of a block. In this case, the region within the reference picture may be specified by using a reference picture index (refIdx) indicating the reference picture and a motion vector to be described later.

Inter-prediction may select a reference picture and a reference block corresponding to a current block within the reference picture, and generate a prediction block for the current block using the selected reference block. The current block may be a block currently being encoded or decoded among blocks of the current picture.

Motion information may be derived during inter prediction by each of the encoding device 100 and the decoding device 200 . Also, the derived motion information can be used to perform inter-picture prediction. At this time, the encoding device 100 and the decoding device 200 use motion information of a reconstructed neighboring block and/or motion information of a collocated block (col block) to achieve encoding and/or decoding efficiency can improve A collocated block may be a block corresponding to a spatial position of a block to be encoded/decoded in an already reconstructed collocated picture (col picture). The reconstructed neighboring block may be a block in the current picture and a block already reconstructed through encoding and/or decoding. In addition, the reconstruction block may be a block adjacent to the encoding/decoding object block and/or a block located at an outer corner of the encoding/decoding object block. Here, the block located at the outer corner of the encoding/decoding target block is a block vertically adjacent to a neighboring block horizontally adjacent to the encoding/decoding target block or a block horizontally adjacent to a neighboring block vertically adjacent to the encoding/decoding target block. can

Each of the encoding apparatus 100 and the decoding apparatus 200 may determine a block spatially located at a position corresponding to a block to be encoded/decoded in a collocated picture, and determine a predefined relative position with respect to the determined block. can The predefined relative position may be a position inside and/or outside a block spatially corresponding to the encoding/decoding target block. Also, each of the encoding apparatus 100 and the decoding apparatus 200 may derive a collocated block based on the determined relative position. Here, the collocated picture may be one picture among at least one reference picture included in the reference picture list.

A method of deriving motion information may change according to a prediction mode of a block to be encoded/decoded. For example, prediction modes applied for inter-prediction may include advanced motion vector prediction (AMVP) and merge mode. Here, the merge mode may be referred to as a motion merge mode.

For example, when AMVP is applied as a prediction mode, each of the encoding apparatus 100 and the decoding apparatus 200 uses a motion vector of a reconstructed neighboring block and/or a motion vector of a collocated block to list motion vector candidates. (motion vector candidate list) can be created. A motion vector of a reconstructed neighboring block and/or a motion vector of a collocated block may be used as a motion vector candidate. Here, a motion vector of a collocated block may be referred to as a temporal motion vector candidate, and a motion vector of a reconstructed neighboring block may be referred to as a spatial motion vector candidate.

A bitstream generated by the encoding device 100 may include a motion vector candidate index. That is, the encoding apparatus 100 may entropy-encode the motion vector candidate index to generate a bitstream. The motion vector candidate index may indicate an optimal motion vector candidate selected from motion vector candidates included in the motion vector candidate list. The motion vector candidate index may be signaled from the encoding apparatus 100 to the decoding apparatus 200 through a bitstream.

The decoding apparatus 200 may entropy-decode a motion vector candidate index from a bitstream and select a motion vector candidate of a decoding target block from motion vector candidates included in a motion vector candidate list by using the entropy-decoded motion vector candidate index. .

The encoding apparatus 100 may calculate a motion vector difference (MVD) between a motion vector of a block to be encoded and a motion vector candidate, and entropy-encode the MVD. The bitstream may include an entropy-encoded MVD. The MVD may be signaled from the encoding device 100 to the decoding device 200 through a bitstream. At this time, the decoding apparatus 200 may entropy-decode the received MVD from the bitstream. The decoding apparatus 200 may derive the motion vector of the decoding object block through the sum of the decoded MVD and the motion vector candidate.

The bitstream may include a reference picture index indicating a reference picture. The reference image index may be entropy-encoded and signaled from the encoding apparatus 100 to the decoding apparatus 200 through a bitstream. The decoding apparatus 200 may predict the motion vector of the decoding object block using motion information of neighboring blocks, and may derive the motion vector of the decoding object block using the predicted motion vector and the motion vector difference. The decoding apparatus 200 may generate a prediction block for a decoding target block based on the derived motion vector and reference image index information.

Another example of a motion information derivation method is a merge mode. Merge mode may mean merging of motions of a plurality of blocks. Merge mode may mean applying motion information of one block to another block as well. When merge mode is applied, each of the encoding device 100 and the decoding device 200 may generate a merge candidate list using motion information of a reconstructed neighboring block and/or motion information of a collocated block. can The motion information may include at least one of 1) a motion vector, 2) a reference image index, and 3) an inter-picture prediction indicator. Prediction indicators can be unidirectional (L0 prediction, L1 prediction) or bidirectional.

In this case, the merge mode may be applied in units of CUs or units of PUs. When the merge mode is performed in units of CUs or PUs, the encoding device 100 may entropy-encode predefined information to generate a bitstream and then signal it to the decoding device 200. A bitstream may include predefined information. The predefined information includes 1) a merge flag, which indicates whether merge mode is to be performed for each block partition, and 2) which block to merge with among neighboring blocks adjacent to the encoding target block. It may include a merge index, which is information about. For example, neighboring blocks of the encoding object block may include a left adjacent block of the encoding object block, an upper adjacent block of the encoding object block, and a temporal adjacent block of the encoding object block.

The merge candidate list may indicate a list in which motion information is stored. Also, the merge candidate list may be created before the merge mode is performed. The motion information stored in the merge candidate list includes motion information of neighboring blocks adjacent to the encoding/decoding object block, motion information of blocks collocated with the encoding/decoding object block in the reference image, and motion already existing in the merge candidate list. It may be at least one of new motion information generated by a combination of pieces of information and a zero merge candidate. Here, motion information of neighboring blocks adjacent to the encoding/decoding target block is a spatial merge candidate, and motion information of blocks collocated with the encoding/decoding target block in the reference image is a temporal merge candidate. ) can be referred to as

The skip mode may be a mode in which motion information of neighboring blocks is applied to an encoding/decoding target block as it is. A skip mode may be one of modes used for inter prediction. When the skip mode is used, the encoding apparatus 100 may entropy-encode information on which block motion information is to be used as the motion information of an encoding target block and signal the decoding apparatus 200 through a bitstream. The encoding device 100 may not signal other information to the decoding device 200. For example, the other information may be syntax element information. Syntax element information may include at least one of motion vector difference information, a coded block flag, and a transform coefficient level.

Based on the foregoing, the video encoding/decoding method according to the present invention will be described in detail.

7 is a flowchart illustrating an image encoding method using merge mode according to the present invention, and FIG. 8 is a flowchart illustrating an image decoding method using merge mode according to the present invention.

Referring to FIG. 7 , the encoding device may derive a merge candidate (S701) and generate a merge candidate list based on the derived merge candidate. When the merge candidate list is generated, motion information may be determined using the generated merge candidate list (S702), and motion compensation of the current block may be performed using the determined motion information (S703). After that, the encoding device may entropy-encode information about motion compensation (S704).

Referring to FIG. 8 , the decoding device may entropy-decode information about motion compensation received from the encoding device (S801), derive a merge candidate (S802), and generate a merge candidate list based on the derived merge candidate. there is. When the merge candidate list is generated, motion information of the current block may be determined using the generated merge candidate list (S803). Thereafter, the decoding apparatus may perform motion compensation using the motion information (S804).

Hereinafter, each step shown in FIGS. 7 and 8 will be described in detail.

First, the step of deriving a merge candidate will be described in detail (S701 and S802).

A merge candidate for the current block may include at least one of a spatial merge candidate, a temporal merge candidate, and an additional merge candidate.

A spatial merge candidate of the current block may be derived from reconstruction blocks adjacent to the current block. For example, motion information of a reconstruction block adjacent to the current block may be determined as a spatial merge candidate for the current block. Here, the motion information may include at least one of a motion vector, a reference image index, or a prediction list utilization flag.

In this case, the motion information of the spatial merge candidate is L0, L1, . . . as well as motion information corresponding to L0 and L1. , may have motion information corresponding to LX. Here, X may be a positive integer including 0. Therefore, the reference image list is L0, L1, . . . , LX, etc. may include at least one or more.

9 is a diagram for explaining an example of deriving a spatial merge candidate and a temporal merge candidate of a current block.

Referring to FIG. 9 , spatial merge candidates of the current block may be derived from neighboring blocks adjacent to the current block (X). Neighboring blocks adjacent to the current block are: the block adjacent to the top of the current block (B1), the block adjacent to the left side of the current block (A1), the block adjacent to the upper right corner of the current block (B0), and the upper left corner of the current block It may include at least one of the adjacent block B2 and the block A0 adjacent to the lower left corner of the current block.

In order to derive a spatial merge candidate of the current block, it may be determined whether neighboring blocks adjacent to the current block can be used to derive a spatial merge candidate of the current block. In this case, whether neighboring blocks adjacent to the current block can be used for deriving a spatial merge candidate of the current block may be determined according to a predetermined priority order. For example, in the example shown in FIG. 14 , spatial merge candidate derivation availability may be determined in the order of blocks of locations A1, B1, B0, A0, and B2. Spatial merge candidates determined based on the availability determination order may be sequentially added to the merge candidate list of the current block.

Next, a method of deriving a temporal merge candidate of the current block will be described.

A temporal merge candidate of the current block may be derived from a reconstructed block included in a co-located picture of the current picture. Here, the corresponding position image is an image for which encoding/decoding has been completed prior to the current image, and may be an image having a temporal order (or image order count, POC) different from that of the current image.

Meanwhile, the temporal merge candidate of the current block may be derived from one or more blocks in the corresponding location image, or may be derived from a plurality of blocks belonging to a plurality of corresponding location images.

The information on the corresponding position image may be transmitted to decoding by the encoder or implicitly derived according to the order of encoding/decoding in the encoder/decoder. Here, the information on the corresponding position image may be at least one of an inter prediction indicator, a reference image index, and motion vector information.

Deriving a temporal merge candidate may mean deriving temporal merge candidate information from a corresponding position block in a corresponding position image and adding it to the merge candidate list of the current block. Temporal merge candidate information may include at least one of a motion vector, a reference picture index, an inter prediction indicator, or a picture order count (POC).

Referring to FIG. 9 , in a collocated picture of a current image, a block including an upper position of a block corresponding to a position spatially identical to the current block X or a position spatially identical to the current block X A temporal merge candidate of the current block may be derived from the block including the internal position of the block corresponding to . Here, the temporal merge candidate may mean motion information of a corresponding location block. For example, the temporal merge candidate of the current block (X) is from a block (H) adjacent to the lower left corner of the block (C) corresponding to the same spatial position as the current block or a block (C3) including the center point of the block C. can be induced. A block H or a block C3 used to derive a temporal merge candidate of the current block may be referred to as a 'collocated block'.

On the other hand, the block corresponding to the current block or the current block may have a square shape or a non-square shape.

When a temporal merge candidate of the current block can be derived from block H including a location outside block C, block H may be set as a block corresponding to the current block. In this case, the temporal merge candidate of the current block may be derived based on the motion information of block H. On the other hand, when a temporal merge candidate of the current block cannot be derived from block H, block C3 including an internal location of block C may be set as a block corresponding to the current block. In this case, a temporal merge candidate of the current block may be derived based on motion information of block C3. If the temporal merge of the current block cannot be derived from block H and block C3 (eg, block H and block C3 are both intra-encoded), a temporal merge candidate for the current block is not derived or the block It may be derived from a block in a different location from H and block C3.

As another example, the temporal merge candidate of the current block may be derived from a plurality of blocks in the corresponding location image. For example, a plurality of temporal merge candidates for the current block may be derived from block H and block C3.

As another example, a temporal merge candidate may be derived from a corresponding position block corresponding to a position moved to the current block X according to arbitrary motion information. In this case, the arbitrary motion information may be derived from the motion information of the adjacent block on which encoding/decoding has been completed.

Meanwhile, the corresponding position block may mean a block at a predetermined position belonging to an image (corresponding position image) used to derive motion information of the current block. Here, the predetermined position may mean at least one of the same position as the current block, a position adjacent to the current block, or a position separated by a predetermined distance from the current block in the corresponding position image. The aforementioned constant distance may be a fixed distance pre-promised to the encoder/decoder, or may be derived based on a predetermined vector component (including at least one of an x component and a y component).

10 is a diagram for explaining an example in which a temporal merge candidate is added to a merge candidate list.

Referring to FIG. 10 , the encoder/decoder may add the derived temporal merge candidate to the merge candidate list when one temporal merge candidate is derived from the corresponding position block at the H1 position.

A block corresponding to the current block may be divided into sub-block units. In this case, the encoder/decoder may determine motion information of one sub-block among sub-blocks of a block corresponding to the current block as a temporal merge candidate of the current block. Also, the encoder/decoder may determine a temporal merge candidate of the current block based on at least one of motion information of lower blocks of a block corresponding to the current block. Here, a lower block may mean a block having a smaller size, shape, or depth than the current block.

For example, when the encoder/decoder is divided into sub-block units and at least one temporal merge candidate is derived, a smaller size, shape, or depth than corresponding blocks corresponding to blocks H and C3 of FIG. 9 described above. A temporal merge candidate can be derived from a sub-block.

As another example, the encoder/decoder may derive at least one temporal merge candidate from motion information of each sub-block unit of a block corresponding to a position to which the current block (X) has moved according to arbitrary motion information.

Even here, the encoder/decoder determines whether motion information of a lower block of the corresponding block exists or whether motion information of a lower block of the corresponding block is available as a temporal merge candidate of the current block, and is a temporal merge candidate of the current block. can be determined by

In addition, the encoder/decoder uses any one of a median value, an average value, a minimum value, a maximum value, a weighted average value, or a mode value of at least one (eg, motion vector) of motion information of sub-blocks of the corresponding location block as a temporal merge candidate of the current block. can be determined by

9 shows that the temporal merge candidate of the current block can be derived from a block adjacent to the lower left corner of the corresponding block or a block including the center point of the corresponding block. However, the position of a block for deriving a temporal merge candidate of the current block is not limited to the example shown in FIG. 9 . For example, the temporal merge candidate of the current block may be derived from a block adjacent to an upper/lower boundary, a left/right boundary, or a corner of the corresponding block, and may be derived from a block including a specific position within the corresponding block (e.g., the corresponding position blocks adjacent to the block's corner boundary).

The temporal merge candidate of the current block may be determined by considering the reference image list (or prediction direction) of the current block and the corresponding block. Meanwhile, the motion information of the temporal merge candidate includes motion information corresponding to L0 and L1 as well as L0, L1, . . . , may have motion information corresponding to LX. Here, X may be a positive integer including 0.

For example, when the reference picture list available for the current block is L0 (ie, when the inter-prediction indicator indicates PRED_L0), motion information corresponding to L0 in the corresponding block is induced as a temporal merge candidate for the current block. can do. That is, when the reference image list available for the current block is LX (where X is an integer representing the index of the reference image list, such as 0, 1, 2, or 3), motion information corresponding to LX of the corresponding position block (hereinafter , referred to as 'LX motion information') can be derived as a temporal merge candidate of the current block.

Even when the current block uses a plurality of reference image lists, a temporal merge candidate of the current block may be determined in consideration of the reference image lists of the current block and the corresponding block.

For example, when the current block performs bidirectional prediction (ie, when the inter-prediction indicator is PRED_BI), L0 motion information, L1 motion information, L2 motion information, . . . , LX motion information can be derived as a temporal merge candidate. When the current block performs three-way prediction (ie, when the inter-prediction indicator is PRED_TRI), L0 motion information, L1 motion information, L2 motion information, . . . , LX motion information can be derived as a temporal merge candidate. When the current block performs four-way prediction (ie, when the inter-prediction indicator is PRED_QUAD), the L0 motion information, L1 motion information, L2 motion information, . . . , At least four or more of the LX motion information may be derived as temporal merge candidates.

In addition, at least one of a temporal merge candidate, a corresponding position image, a corresponding position block, a prediction list utilization flag, and a reference image index may be derived based on at least one of the coding parameters of the current block, the neighboring block, or the corresponding position block.

A temporal merge candidate may be preliminarily derived when the number of derived spatial merge candidates is smaller than the maximum number of merge candidates. Accordingly, when the number of derived spatial merge candidates reaches the maximum number of merge candidates, the process of deriving temporal merge candidates can be omitted.

For example, when the maximum number of merge candidates, maxNumMergeCand, is two and the two derived spatial merge candidates have different values, the process of deriving the temporal merge candidates may be omitted.

Meanwhile, within the maximum number of merge candidates, maxNumMergeCand, at least one temporal merge candidate must be included. In this case, the number of spatial merge candidates can be reduced by excluding at least one or more of the derived spatial merge candidates or by merging the derived spatial merge candidates to necessarily include one or more temporal merge candidates. Here, the merging of spatial merge candidates can be calculated using any one of the average value, maximum value, minimum value, median value, weighted average value, or mode value of motion information between spatial merge candidates having the same inter-prediction indicator and/or reference image.

Also, the encoder/decoder may change the maximum number of merge candidates so that at least one temporal merge candidate is necessarily included.

As another example, the temporal merge candidate of the current block may be derived based on the maximum number of temporal merge candidates. Here, the maximum number of temporal merge candidates may be preset so that the same value is used in the encoding device and the decoding device. Alternatively, information representing the maximum number of temporal merge candidates of the current block may be coded through a bitstream and signaled to a decoder. For example, the encoder may encode maxNumTemporalMergeCand indicating the maximum number of temporal merge candidates of the current block and signal the decoder through a bitstream. At this time, maxNumTemporalMergeCand may be set to a positive integer including 0. For example, maxNumTemporalMergeCand may be set to 1. The value of maxNumTemporalMergeCand may be variably derived based on information about the number of temporal merge candidates signaled, or may be a fixed value preset in an encoder/decoder.

Meanwhile, the encoder/decoder performs a redundancy check on whether the derived merge candidate has motion information different from that of the merge candidate already added to the merge candidate list, and only when the derived merge candidate has motion information different from the merge candidate merge candidate can be added to the list.

If the distance between the current image including the current block and the reference image of the current block is different from the distance between the corresponding location image including the corresponding location block and the reference image of the corresponding location block, the motion vector of the temporal merge candidate of the current block is It can be obtained by scaling the motion vector of the corresponding position block.

Alternatively, even when the reference for inter-prediction of the image including the current block and the reference for inter-prediction of the image including the corresponding block are different from each other, the motion vector of the temporal merge candidate of the current block is It can be obtained by scaling the motion vector.

Here, scaling may be performed based on at least one of a distance between a current image and a reference image referred to by the current block and a distance between a corresponding location image and a reference image referred to by the corresponding location block. For example, a temporal merge candidate of the current block is performed by scaling the motion vector of the corresponding location block according to the ratio of the distance between the current image and the reference image referred to by the current block and the distance between the corresponding location image and the reference image referred to by the corresponding location block. A motion vector of can be derived.

Based on the size (hereinafter referred to as 'first block size') or block depth (hereinafter referred to as 'first block depth') of a block in which motion compensation information is entropy encoded/decoded, the first block size or Temporal merge candidates may be shared among blocks having a size smaller than the first block depth or a depth greater than the first block depth. Here, the motion compensation information may be at least one of skip mode use information, merge mode use information, and merge index information.

A block in which information about motion compensation is entropy-encoded/decoded may be a CTU or a sub-unit of the CTU, a CU or a PU.

Specifically, when the size of the current block is smaller than the size of the first block, the encoder/decoder may derive a temporal merge candidate of the current block from a corresponding position block of an upper block having the first block size. Also, blocks included in the upper block may share the derived temporal merge candidate.

Also, when the block depth of the current block is deeper than the first block depth, the encoder/decoder may derive a temporal merge candidate from a block corresponding to an upper block having the first block depth. Also, blocks included in the upper block may share the derived temporal merge candidate.

Here, sharing a temporal merge candidate may mean that each merge candidate list of shared blocks may be generated based on the same temporal merge candidate.

Also, sharing temporal merge candidates may mean that shared blocks may perform motion compensation using one merge candidate list. Here, the shared merge candidate list may include a temporal merge candidate derived based on an upper block in which motion compensation information is entropy-encoded/decoded.

11 illustrates an example of scaling a motion vector among motion information of a corresponding block in order to derive a temporal merge candidate of a current block.

The motion vector of the corresponding position block is a difference value (td) between a picture order count (POC) representing the display order of the corresponding position image and the POC of the reference image of the corresponding position block, and a POC of the current image and the reference image of the current block. It may be scaled based on at least one of the difference values (tb) between POCs.

Prior to scaling, td or tb may be adjusted so that td or tb exists within a predetermined range. For example, when the predetermined range indicates -128 to 127 and td or tb is smaller than -128, td or tb may be adjusted to -128. When td or tb is greater than 127, td or tb may be adjusted to 127. If td or tb is within the range of -128 to 127, do not adjust td or tb.

The scaling factor DistScaleFactor can be calculated based on td or tb. In this case, the scaling factor may be calculated based on Equation 1 below.

In Equation 1, Abs() represents an absolute value function, and the output value of the function becomes the absolute value of the input value.

The value of the scaling factor DistScaleFactor calculated based on Equation 1 may be adjusted within a predetermined range. For example, DistScaleFactor may be adjusted to exist within the range of -1024 to 1023.

The motion vector of the temporal merge candidate of the current block may be determined by scaling the motion vector of the corresponding position block using a scaling factor. For example, the motion vector of the temporal merge candidate of the current block may be determined by Equation 2 below.

In Equation 2 above, Sign() is a function that outputs sign information of a value included in (). For example, if Sign(-1), - is output. In Equation 2 above, mvCol may mean a motion vector of a corresponding position block.

According to the above-described method, a motion vector scaled according to an arbitrary reference image of the current block is used as a motion vector of a temporal merge candidate of the current block, or scaling information of at least one block among blocks that have been encoded/decoded nearby is used. Thus, a motion vector scaled as a temporal merge candidate of the current block may be corrected at least once and used as a motion vector of the temporal merge candidate of the current block.

*Next, a method of deriving an additional merge candidate of the current block will be described.

The additional merge candidate means at least one of a modified spatial merge candidate, a modified temporal merge candidate, a combined merge candidate, and a merge candidate having a predetermined motion information value. can do. Here, deriving additional merge candidates may mean deriving additional merge candidates and adding them to the merge candidate list.

The changed spatial merge candidate may refer to a merge candidate obtained by changing at least one of motion information of the derived spatial merge candidate.

The changed temporal merge candidate may refer to a merge candidate obtained by changing at least one of motion information of the derived temporal merge candidate.

The combined merge candidate is at least one of a spatial merge candidate, a temporal merge candidate, a modified spatial merge candidate, a modified temporal merge candidate, a combined merge candidate, and motion information of merge candidates having a predetermined motion information value existing in the merge candidate list. It may mean a merge candidate derived by combining motion information. Alternatively, the combined merge candidate is a spatial merge candidate derived from a block that does not exist in the merge candidate list but can derive at least one or more of a spatial merge candidate and a temporal merge candidate, and a modified temporal merge candidate generated based on the spatial merge candidate and the derived temporal merge candidate. It may refer to a merge candidate derived by combining at least one motion information among a spatial merge candidate, a modified temporal merge candidate, a combined merge candidate, and a merge candidate having a predetermined motion information value.

Meanwhile, a merge candidate having a predetermined motion information value may mean a zero merge candidate having a motion vector of (0, 0).

* The zero merge candidate may refer to a merge candidate having a motion vector of (0, 0) in at least one motion information among L0 motion information, L1 motion information, L2 motion information, and L3 motion information.

Also, the zero merge candidate may be at least one of two types. The first zero merge candidate may refer to a merge candidate in which a motion vector is (0, 0) and a reference image index may have a value greater than or equal to 0. Also, the second zero merge candidate may refer to a merge candidate having a motion vector of (0, 0) and a reference image index of 0.

Next, the step of determining the motion information of the current block using the generated merge candidate list will be described in detail (S702 and S803).

The encoder may determine a merge candidate used for motion compensation among merge candidates in the merge candidate list through motion estimation, and encode a merge candidate index (merge_idx) indicating the determined merge candidate in a bitstream.

Meanwhile, the encoder may determine motion information of the current block by selecting a merge candidate from the merge candidate list based on the above-described merge candidate index to generate a prediction block. Here, a prediction block of the current block may be generated by performing motion compensation based on the determined motion information.

For example, when merge candidate index 3 is selected, a merge candidate indicated by merge candidate index 3 in the merge candidate list may be determined as motion information and used for motion compensation of an encoding object block.

The decoder may decode a merge candidate index in the bitstream to determine a merge candidate in the merge candidate list indicated by the merge candidate index. The determined merge candidate may be determined using motion information of the current block. The determined motion information is used for motion compensation of the current block. In this case, motion compensation may have the same meaning as inter prediction.

For example, when the merge candidate index is 2, the merge candidate indicated by the merge candidate index 2 in the merge candidate list may be determined as motion information and used for motion compensation of the decoding target block.

In addition, the motion information may be used for inter-picture prediction or motion compensation of the current block by changing at least one value of information corresponding to the motion information of the current block. Here, the value to be changed among information corresponding to motion information may be at least one of an x component of a motion vector, a y component of a motion vector, and a reference image index.

Next, a step of performing motion compensation of the current block using the determined motion information will be described in detail (S703 and S804).

The encoder and decoder may perform inter prediction or motion compensation using the motion information of the determined merge candidate. Here, the current block (block to be encoded/decoded) may have motion information of the determined merge candidate.

The current block may have a minimum of 1 to a maximum of N pieces of motion information according to prediction directions. A final prediction block of the current block may be derived by generating a minimum of 1 to a maximum of N prediction blocks using motion information.

For example, when the current block has one piece of motion information, a prediction block generated using the motion information may be determined as the final prediction block of the current block.

On the other hand, when the current block has a plurality of motion information, a plurality of prediction blocks may be generated using the plurality of motion information, and a final prediction block of the current block may be determined based on a weighted sum of the plurality of prediction blocks. Reference images including each of a plurality of prediction blocks indicated by a plurality of pieces of motion information may be included in different reference image lists or in the same reference image list. Also, when the current block has a plurality of pieces of motion information, a plurality of reference images among the pieces of motion information may indicate the same reference image.

For example, a plurality of prediction blocks are generated based on at least one of a spatial merge candidate, a temporal merge candidate, a modified spatial merge candidate, a modified temporal merge candidate, a merge candidate or combined merge candidate having a predetermined motion information value, and an additional merge candidate. and a final prediction block of the current block may be determined based on a weighted sum of a plurality of prediction blocks.

As another example, a plurality of prediction blocks may be generated based on merge candidates indicated by a preset merge candidate index, and a final prediction block of the current block may be determined based on a weighted sum of the plurality of prediction blocks. In addition, a plurality of prediction blocks may be generated based on merge candidates existing in a preset merge candidate index range, and a final prediction block of the current block may be determined based on a weighted sum of the plurality of prediction blocks.

The weight applied to each prediction block may have an equal value of 1/N (where N is the number of generated prediction blocks). For example, when 2 prediction blocks are generated, the weight applied to each prediction block is 1/2, and when 3 prediction blocks are generated, the weight applied to each prediction block is 1/3, and 4 prediction blocks are generated. When a block is generated, a weight applied to each prediction block may be 1/4. Alternatively, a final prediction block of the current block may be determined by assigning a different weight to each prediction block.

The weight does not have to have a fixed value for each prediction block, and may have a variable value for each prediction block. In this case, the weights applied to each prediction block may be the same or different. For example, when two prediction blocks are generated, the weight applied to the two prediction blocks is not only (1/2, 1/2), but also (1/3, 2/3), (1/4, 3 /4), (2/5, 3/5), (3/8, 5/8) may be a variable value for each block. Meanwhile, the weight may be a positive real value and a negative real value. For example, it may include negative real values such as (-1/2, 3/2), (-1/3, 4/3), (-1/4, 5/4), and the like.

Meanwhile, in order to apply a variable weight, one or more weight information for the current block may be signaled through a bitstream. Weight information may be signaled for each prediction block or for each reference picture. It is also possible for a plurality of prediction blocks to share one weight information.

The encoder and decoder may determine whether motion information of the merge candidate is used based on the prediction block list utilization flag. For example, when the prediction block list utilization flag for each reference picture list indicates the first value of 1, an encoder and a decoder can use motion information of a merge candidate of a current block to perform inter prediction or motion compensation. and when indicating the second value of 0, it may indicate that the encoder and decoder do not perform inter-prediction or motion compensation by using the motion information of the merge candidate of the current block. Meanwhile, the first value of the prediction block list utilization flag may be set to 0 and the second value may be set to 1.

In the following Equations 3 to 5, the inter-prediction indicators of the current block are, respectively, PRED_BI (or when the current block can use 2 motion information), PRED_TRI (or when the current block can use 3 motion information) ) and PRED_QUAD (or when the current block can use 4 pieces of motion information), and when the prediction direction for each reference picture list is unidirectional, an example of generating a final prediction block of the current block is shown.

In Equations 3 to 5, P_BI, P_TRI, and P_QUAD may indicate final prediction blocks of the current block, and LX (X = 0, 1, 2, 3) may indicate a reference image list. WF_LX may represent a weight value of a prediction block generated using LX, and OFFSET_LX may represent an offset value for a prediction block generated using LX. P_LX means a prediction block generated using motion information on LX of the current block. RF means a rounding factor and can be set to 0, a positive number, or a negative number. The LX reference image list includes a long-term reference image, a reference image without performing a deblocking filter, a reference image without performing a sample adaptive offset, and an adaptive loop filter (adaptive loop filter). loop filter), reference image with only deblocking filter and adaptive offset, reference image with only deblocking filter and adaptive loop filter, reference image with sample adaptive offset and adaptive loop filter only It may include at least one of reference images on which all of the image, deblocking filter, sample adaptive offset, and adaptive loop filter are performed. In this case, the LX reference picture list may be at least one of an L0 reference picture list, an L1 reference picture list, an L2 reference picture list, and an L3 reference picture list.

Even when prediction directions for a predetermined reference image list are multi-directional, a final prediction block for a current block may be obtained based on a weighted sum of prediction blocks. In this case, weights applied to prediction blocks derived from the same reference image list may have the same value or different values.

At least one of a weight (WF_LX) and an offset (OFFSET_LX) for a plurality of prediction blocks may be an encoding parameter for entropy encoding/decoding.

As another example, the weight and offset may be derived from encoded/decoded neighboring blocks around the current block. Here, neighboring blocks of the current block may include at least one of a block used to derive a spatial merge candidate of the current block and a block used to derive a temporal merge candidate of the current block.

As another example, the weight and offset may be determined based on the display order (POC) of the current image and each reference image. In this case, the weight or offset may be set to a smaller value as the distance between the current image and the reference image increases, and the weight or offset may be set to a larger value as the distance between the current image and the reference image decreases. For example, when the POC difference between the current image and the L0 reference image is 2, the weight value applied to the prediction block generated with reference to the L0 reference image is set to 1/3, whereas the POC difference between the current image and the L0 reference image is set. When is 1, a weight value applied to a prediction block generated with reference to the L0 reference picture may be set to 2/3. As exemplified above, the weight or offset value may have an inverse relationship with the display order difference between the current image and the reference image. As another example, the weight or offset value may have a proportional relationship with a difference in display order between the current image and the reference image.

As another example, entropy encoding/decoding may be performed on at least one of weights and offsets based on at least one or more of encoding parameters of a current block, a neighboring block, or a corresponding block. In addition, a weighted sum of prediction blocks may be calculated based on at least one of the coding parameters of the current block, neighboring blocks, and corresponding blocks.

A weighted sum of a plurality of prediction blocks may be applied only to some regions within a prediction block. Here, the partial region may be a region corresponding to a boundary within a prediction block. As above, in order to apply weighted sum only to some regions, weighted sum may be performed in units of sub-blocks of prediction blocks.

In addition, when calculating the weighted sum of prediction blocks using the merge mode, the weighted sum can be calculated using at least one merge candidate existing in the merge candidate list and used as the final prediction block of the encoding/decoding object block.

For example, prediction blocks may be generated using only spatial merge candidates, a weighted sum of the prediction blocks may be calculated, and the calculated weighted sum may be used as a final prediction block of an encoding/decoding target block.

For example, prediction blocks may be generated using spatial merge candidates and temporal merge candidates, a weighted sum of the prediction blocks may be calculated, and the calculated weighted sum may be used as a final prediction block of an encoding/decoding target block.

For example, prediction blocks may be generated using only combined merge candidates, a weighted sum of the prediction blocks may be calculated, and the calculated weighted sum may be used as a final prediction block of an encoding/decoding object block.

* For example, it is possible to generate prediction blocks only with merge candidates having specific merge candidate indices, calculate a weighted sum of the prediction blocks, and use the calculated weighted sum as a final prediction block of an encoding/decoding target block.

For example, prediction blocks may be generated only with merge candidates existing within a specific merge candidate index range, a weighted sum of the prediction blocks may be calculated, and the calculated weighted sum may be used as a final prediction block of an encoding/decoding target block.

Next, the process of entropy encoding/decoding information related to motion compensation will be described in detail (S704 and S801).

The encoding device may entropy-encode motion compensation information through a bitstream, and the decoding device may entropy-decode motion compensation information included in the bitstream. Here, the information on motion compensation to be entropy encoded/decoded includes skip mode use/non-use information (cu_skip_flag), merge mode use/non-use information (merge_flag), merge index information (merge_index), inter prediction indicator (inter prediction indicator) (inter_pred_idc ), weight values (wf_l0, wf_l1, wf_l2, wf_l3), and offset values (offset_l0, offset_l1, offset_l2, offset_l3). The motion compensation information may be entropy encoded/decoded in at least one unit of a CTU, a coding block, and a prediction block.

Skip mode usage availability information (cu_skip_flag) may indicate use of skip mode when it has a first value of 1, and may not indicate use of skip mode when it has a second value of 0. Motion compensation of the current block may be performed by using the skip mode based on information on whether the skip mode is used.

Merge mode use availability information (merge_flag) may indicate use of merge mode when it has a first value of 1, and may not indicate use of merge mode when it has a second value of 0. Motion compensation of the current block may be performed using merge mode based on information on whether merge mode is used or not.

The merge index information (merge_index) may refer to information indicating a merge candidate in a merge candidate list.

Also, merge index information may refer to information about a merge index.

In addition, the merge index information may indicate a block from which a merge candidate is derived among blocks reconstructed that are spatially/temporally adjacent to the current block.

Also, the merge index information may indicate at least one piece of motion information of a merge candidate. For example, if the merge index information has a first value of 0, it may indicate the first merge candidate in the merge candidate list, and if it has a second value of 1, it may indicate the second merge candidate in the merge candidate list, If it has the third value of 2, it may indicate the third merge candidate in the merge candidate list. Similarly, when the values have the 4th to Nth values, merge candidates corresponding to the values may be indicated according to the order in the merge candidate list. Here, N may mean a positive integer including 0.

Motion compensation of the current block may be performed using merge mode based on merge mode index information.

The inter prediction indicator may mean at least one of an inter prediction direction of the current block or the number of prediction directions when encoding/decoding is performed by inter prediction of the current block. For example, the inter-prediction indicator may indicate unidirectional prediction or multidirectional prediction such as bidirectional prediction, three-way prediction, or four-way prediction. The inter-prediction indicator may indicate the number of reference images used when the current block generates a prediction block. Alternatively, one reference image may be used for prediction of a plurality of directions. In this case, N (N>M) direction prediction may be performed using M reference images. The inter-prediction indicator may mean the number of prediction blocks used when performing inter-prediction or motion compensation for a current block.

In this way, according to the inter-prediction indicator, the number of reference images used when generating the prediction block of the current block, the number of prediction blocks used when inter-prediction or motion compensation of the current block is performed, or the current block can be used The number of reference image lists present may be determined. Here, the number N of the reference image list is a positive integer and may have a value of 1, 2, 3, 4 or more. For example, the reference image list may include L0, L1, L2, and L3. The current block may perform motion compensation using one or more reference picture lists.

For example, the current block may generate at least one prediction block using at least one reference image list, and perform motion compensation on the current block. For example, motion compensation is performed by generating one or more prediction blocks using the reference image list L0, or motion compensation is performed by generating one or more prediction blocks using the reference image lists L0 and L1. can be done Alternatively, motion compensation may be performed by generating one, one or more prediction blocks, or up to N prediction blocks (where N is a positive integer of 3 or 2 or more) using the reference image lists L0, L1, and L2, or reference image lists L0, L1, and L2. Using the image lists L0, L1, L2, and L3, motion compensation for the current block is performed using one, one or more prediction blocks, or up to N prediction blocks (where N is 4 or a positive integer greater than or equal to 2). can be done

The reference image indicator may indicate unidirectional (PRED_LX), bidirectional (PRED_BI), three-directional (PRED_TRI), four-directional (PRED_QUAD) or more directions according to the number of prediction directions of the current block.

For example, assuming that unidirectional prediction is performed for each reference image list, the inter-prediction indicator PRED_LX uses the reference image list LX (X is an integer such as 0, 1, 2, or 3) to form one prediction block. It may mean generating and performing inter-prediction or motion compensation using one generated prediction block. The inter prediction indicator PRED_BI indicates that two prediction blocks are generated using at least one of L0, L1, L2, and L3 reference picture lists, and inter prediction or motion compensation is performed using the generated two prediction blocks. can mean The inter prediction indicator PRED_TRI indicates that three prediction blocks are generated using at least one of L0, L1, L2, and L3 reference picture lists, and inter prediction or motion compensation is performed using the generated three prediction blocks. can mean The inter-prediction indicator PRED_QUAD indicates that 4 prediction blocks are generated using at least one of L0, L1, L2, and L3 reference picture lists, and inter-prediction or motion compensation is performed using the generated 4 prediction blocks. can mean That is, the sum of the numbers of prediction blocks used to perform inter prediction of the current block may be set as an inter prediction indicator.

When multi-directional prediction is performed on the reference image list, the inter-prediction indicator PRED_BI means that bi-directional prediction is performed on the L0 reference image list, and the inter-prediction indicator PRED_TRI indicates that three-way prediction is performed on the L0 reference image list. performed, uni-prediction is performed on the L0 reference picture list and bi-prediction is performed on the L1 reference picture list, or bi-directional prediction is performed on the L0 reference picture list, and uni-directional prediction is performed on the L1 reference picture list. can mean becoming

As such, the inter-prediction indicator performs motion compensation by generating a minimum of 1 to a maximum of N (where N is the number of prediction directions indicated by the inter-prediction indicator) from at least one reference picture list. or generating at least one to a maximum of N prediction blocks from N reference images, and performing motion compensation on the current block using the generated prediction blocks.

For example, the inter-prediction indicator PRED_TRI means that three prediction blocks are generated using at least one of L0, L1, L2, and L3 reference image lists to perform inter-prediction or motion compensation of the current block, or L0 , L1, L2, and L3 reference image lists to generate three prediction blocks using at least three of the reference image lists to perform inter-prediction or motion compensation of the current block. In addition, PRED_QUAD means to perform inter-prediction or motion compensation of the current block by generating 4 prediction blocks using at least one of the L0, L1, L2, and L3 reference picture lists, or L0, L1, and L2. , It may mean an inter-prediction indicator that performs inter-prediction or motion compensation of the current block by generating four prediction blocks using at least four of the L3 reference image lists.

Available inter prediction directions may be determined according to an inter prediction indicator, and all or some of the available inter prediction directions may be selectively used based on the size and/or shape of a current block.

A prediction list utilization flag indicates whether a prediction block is generated using a corresponding reference image list.

For example, when the prediction list utilization flag indicates a first value of 1, it indicates that a prediction block can be generated using the corresponding reference image list, and when indicating a second value of 0, using the corresponding reference image list It may indicate not generating a prediction block. Here, the first value of the prediction list utilization flag may be set to 0 and the second value may be set to 1.

That is, when the prediction list utilization flag indicates a first value, a prediction block of the current block may be generated using motion information corresponding to the corresponding reference image list.

Meanwhile, the prediction list utilization flag may be set based on the inter prediction indicator. For example, when the inter-prediction indicator indicates PRED_LX, PRED_BI, PRED_TRI, or PRED_QUAD, the prediction list utilization flag predFlagLX may be set to a first value of 1. If the inter-prediction indicator is PRED_LN (N is a positive integer other than X), the prediction list utilization flag predFlagLX may be set to a second value of 0.

Also, the inter-picture prediction indicator may be set based on the prediction list utilization flag. For example, when the prediction list utilization flags predFlagL0 and predFlagL1 indicate the first value of 1, the inter prediction indicator may be set to PRED_BI. For example, when only the prediction list utilization flag predFlagL0 indicates the first value of 1, the inter prediction indicator may be set to PRED_L0.

At least one or more of the above-described information on motion compensation may be entropy encoded/decoded in at least one unit of the CTU or a sub-CTU of the CTU. Here, the lower unit of the CTU may include at least one unit of CU and PU. A block of a subunit of a CTU may have a square shape or a non-square shape. Information on motion compensation, which will be described later, may refer to at least one of motion compensation information for convenience.

When information on motion compensation is entropy encoded/decoded in the CTU, motion compensation can be performed using the information on motion compensation in all or some blocks existing in the CTU according to the value of the information on motion compensation. .

When entropy encoding/decoding of motion compensation information is performed in the CTU or a subunit of the CTU, the motion compensation information may be entropy encoded/decoded based on at least one of a predetermined block size and a predetermined block depth.

Here, information on a predetermined block size or predetermined block depth may be additionally entropy encoded/decoded. Alternatively, the information on the predetermined block size or predetermined block depth may be determined based on at least one of values preset in the encoder and decoder, encoding parameters, or other syntax element values.

Information on motion compensation can be entropy encoded/decoded only in blocks having a block size greater than or equal to the predetermined block size, and information on motion compensation can be entropy encoded/decoded in blocks having a block size smaller than the predetermined block size. It may not be. In this case, sub-blocks within a block having a block size greater than or equal to a predetermined block size are subjected to motion compensation based on entropy-encoded/decoded motion compensation information in a block having a block size greater than or equal to the predetermined block size. can be done That is, information about motion compensation including a motion vector candidate, a motion vector candidate list, a merge candidate, a merge candidate list, and the like can be shared with sub-blocks in a block having a block size greater than or equal to a predetermined block size.

Information on motion compensation can be entropy encoded/decoded only in blocks having a block depth smaller than or equal to the predetermined block depth, and information on motion compensation can be entropy encoded/decoded in blocks having a block depth deeper than the predetermined block depth. may not be decrypted. In this case, sub-blocks within a block having a block depth equal to or less than a predetermined block depth perform motion compensation based on information about motion compensation that is entropy-encoded/decoded in a block having a block depth equal to or less than the predetermined block depth. can be done That is, motion compensation information including a motion vector candidate, a motion vector candidate list, a merge candidate, a merge candidate list, and the like can be shared with sub-blocks in a block having a block depth smaller than or equal to a predetermined block depth.

For example, when information about motion compensation is entropy encoded/decoded in a CTU subunit in which the block size of the CTU is 64 x 64 and 32 x 32, in a block belonging to a 32 x 32 block and smaller than the 32 x 32 block unit, Motion compensation may be performed based on information about motion compensation that is entropy-encoded/decoded in units of 32x32 blocks.

As another example, if the block size of the CTU is 128 x 128 and information about motion compensation is entropy encoded/decoded in a CTU subunit of 16 x 16, it belongs to a 16 x 16 block and has a size smaller than or equal to the 16 x 16 block unit. In a block, motion compensation may be performed based on information about motion compensation that is entropy-encoded/decoded in units of 16x16 blocks.

As another example, when information about motion compensation is entropy encoded/decoded in a CTU subunit having a block depth of 0 and a block depth of 1 of a CTU, a block belonging to block depth 1 and having a block depth deeper than block depth 1 has a block depth of 1. Motion compensation may be performed based on information about motion compensation that is entropy encoded/decoded in 1.

For example, if at least one piece of information about motion compensation is entropy encoded/decoded in a CTU subunit having a block depth of 0 and a block depth of 2, it belongs to block depth 2 and is equal to or equal to block depth 2. In a block having a block depth deeper than 2, motion compensation may be performed based on information about motion compensation that is entropy-encoded/decoded at a block depth of 2.

Here, the value of the block depth may have a positive integer including 0. As the value of the block depth increases, it may mean that the block depth is deep, and as the value of the block depth decreases, it may mean that the block depth is shallow. Accordingly, the block size may decrease as the block depth value increases, and the block size may increase as the block depth value decreases. Further, the lower part of the predetermined block depth may mean a deeper depth than the predetermined block depth, and the lower part of the predetermined block depth may mean a deeper depth within a block corresponding to the predetermined block depth.

Information on motion compensation may be entropy-encoded/decoded on a block-by-block basis or entropy-encoded/decoded at a higher level. For example, motion compensation information is entropy encoded/decoded in units of blocks such as CTU, CU, or PU, or a video parameter set, a sequence parameter set, or a picture parameter set. ), entropy encoding/decoding may be performed at a higher level such as an Adaptation Parameter Set or a Slice Header.

Information on motion compensation may be entropy-encoded/decoded based on a difference value between information on motion compensation and a predicted value of information on motion compensation. Taking an inter-prediction indicator, which is one piece of motion compensation information, as an example, entropy encoding/decoding may be performed on a prediction value of the inter-prediction indicator and a difference value of the inter-prediction indicator. In this case, the inter-prediction indicator difference value may be entropy-encoded/decoded in units of blocks, and the predictive value of the inter-prediction indicator may be entropy-encoded/decoded at a higher level. When information prediction values related to motion compensation, such as inter-prediction indicator prediction values, are entropy-encoded/decoded in units of pictures or slices, blocks included in a picture or slice may use a common information prediction value related to motion compensation.

Information prediction values related to motion compensation may be derived through a specific region within a video, slice or tile, or a specific region within a CTU or CU. For example, an inter prediction indicator of a specific region within an image, slice, tile, CTU, or CU may be used as an inter prediction indicator predicted value. In this case, entropy encoding/decoding of predicted information related to motion compensation may be omitted, and entropy encoding/decoding of only information difference values related to motion compensation may be performed.

Alternatively, information prediction values related to motion compensation may be derived from adjacent blocks encoded/decoded around the current block. For example, an inter prediction indicator of an encoded/decoded neighboring block around the current block may be set as a predictive value of the inter prediction indicator of the current block. Here, neighboring blocks of the current block may include at least one of a block used to derive a spatial merge candidate and a block used to derive a temporal merge candidate. Also, the neighboring block may have a depth equal to or smaller than that of the current block. When there are a plurality of neighboring blocks, one may be selectively used according to a predetermined priority order. Neighboring blocks used to predict motion compensation information may have a fixed position relative to the current block or may have a variable position according to the position of the current block. Here, the location of the current block may be a location based on a picture or slice to which the current block belongs, or a location based on the location of a CTU, CU, or PU to which the current block belongs.

The merge index information may be calculated using index information within predetermined sets in an encoder and a decoder.

In the case of using an information prediction value related to motion compensation and an information difference value related to motion compensation, the decoding apparatus calculates an information value related to motion compensation for a prediction block by summing the information prediction value related to motion compensation and the information difference value related to motion compensation. can

Information about motion compensation or a difference value of information about motion compensation may be entropy encoded/decoded based on at least one of encoding parameters of a current block, a neighboring block, or a corresponding block.

Based on at least one of the current block, the neighboring block, or the encoding parameter of the corresponding block, information on motion compensation, a predicted value of the information on motion compensation, or a difference between the information on motion compensation is converted into information on motion compensation of the current block. , can be derived as a predicted value of motion compensation-related information or a difference value of motion compensation-related information.

Instead of entropy encoding/decoding information on motion compensation of the current block, information on motion compensation of blocks encoded/decoded around the current block may be used as information on motion compensation of the current block. For example, an inter prediction indicator of the current block may be set to the same value as an inter prediction indicator of an encoded/decoded neighboring block of the current block.

Also, at least one of the motion compensation information may have a predetermined fixed value in the encoder and decoder. The predetermined fixed value may be determined as a value for at least one of motion compensation information, and in blocks having a smaller block size inside a specific block size, motion compensation having the predetermined fixed value At least one of the information may be shared. Similarly, blocks having a deeper block depth below a specific block size may share at least one of the motion compensation information having a predetermined fixed value. Here, the fixed value may be a positive integer value including 0 or an integer vector value including (0, 0).

Here, the meaning of sharing at least one piece of information about motion compensation means that at least one piece of information about motion compensation in blocks may have the same value or at least one piece of information about motion compensation in blocks. This may mean performing motion compensation using the same values.

Motion compensation information includes a motion vector, a motion vector candidate, a motion vector candidate index, a motion vector difference value, a motion vector prediction value, skip mode use information (skip_flag), merge mode use information (merge_flag), merge index information (merge_index) ), motion vector resolution information, overlapped block motion compensation information, local illumination compensation information, affine motion compensation information, decoder motion vector It may further include at least one of decoder-side motion vector derivation information and bi-directional optical flow information.

The motion vector resolution information may be information indicating whether a specific resolution is used for at least one of the motion vector and the motion vector difference value. Here, resolution may mean precision. In addition, specific resolutions are integer-pixel (integer-pel) units, 1/2-pixel (1/2-pel) units, 1/4-pixel (1/4-pel) units, 1/8-pixel (1 /8-pel) unit, 1/16-pixel (1/16-pel) unit, 1/32-pixel (1/32-pel) unit, and 1/64-pixel (1/64-pel) unit at least can be set to one.

The overlapped block motion compensation information may be information indicating whether a weighted sum of prediction blocks of the current block is calculated by additionally using motion vectors of neighboring blocks spatially adjacent to the current block during motion compensation of the current block.

Area illumination compensation information may be information indicating whether at least one of a weight value and an offset value is applied when generating a prediction block of a current block. Here, at least one of the weight value and the offset value may be a value calculated based on the reference block.

The affine motion compensation information may be information indicating whether an affine motion model is used when motion compensation for the current block is performed. Here, the affine motion model may be a model that divides one block into multiple sub-blocks using a plurality of parameters and calculates motion vectors of the divided sub-blocks using representative motion vectors.

The decoder motion vector derivation information may be information indicating whether the decoder derives and uses a motion vector necessary for motion compensation. Based on the decoder motion vector derivation information, motion vector information may not be entropy encoded/decoded. And, when the decoder motion vector derivation information indicates that the decoder derives and uses the motion vector, information on the merge mode may be entropy encoded/decoded. That is, the decoder motion vector derivation information may indicate whether the decoder uses merge mode.

The bidirectional optical flow information may be information indicating whether motion compensation is performed by calibrating a motion vector in units of pixels or units of sub-blocks. Based on the bidirectional optical flow information, a motion vector in units of pixels or units of sub-blocks may not be entropy-encoded/decoded. Here, the motion vector correction may be to change a motion vector value of a motion vector in units of blocks or in units of pixels or sub-blocks.

The current block may perform motion compensation using at least one piece of motion compensation information, and entropy-encode/decode at least one piece of motion compensation information.

In the case of entropy encoding/decoding information related to motion compensation, a truncated rice binarization method, a K-th order Exp_Golomb binarization method, and a restricted K-th order Exp_Golomb binarization method ) Binarization method, such as a fixed-length binarization method, a unary binarization method, or a truncated unary binarization method, may be used.

When entropy encoding/decoding information about motion compensation, information about motion compensation of neighboring blocks around the current block or area information of neighboring blocks, information about previously encoded/decoded motion compensation, or information about previously encoded/decoded motion compensation A context model may be determined using at least one of region information, information about the depth of the current block, and information about the size of the current block.

In addition, when entropy encoding/decoding information about motion compensation, information about motion compensation of neighboring blocks, information about previously encoded/decoded motion compensation, information about the depth of the current block, and information about the size of the current block Entropy encoding/decoding may be performed by using at least one of the pieces of information as a predicted value for motion compensation information of the current block.

The image encoding/decoding method using the merge mode has been described with reference to FIGS. 7 to 11 above. Hereinafter, a method for deriving a temporal merge candidate will be described in more detail with reference to FIGS. 12 to 15 .

The encoder/decoder may derive a temporal merge candidate from at least one block reconstructed from a reference picture that is a temporal neighbor of the current block. Here, the reference picture, which is a temporal neighbor of the current block, may be a co-located picture.

Information on the corresponding position image (eg, at least one of an inter-prediction indicator, a reference image index, motion vector information, and a picture order count (POC)) is decoded by the encoder to obtain a sequence, picture, slice, or tile , CTU, CU, or PU may be transmitted in at least one unit.

In addition, the information on the corresponding position image is a hierarchy according to the encoding/decoding order in the encoder/decoder, motion information of temporal neighboring blocks/spatial neighboring blocks that have been encoded/decoded, or sequences, pictures, slices, tiles, etc. It can be implicitly induced in an encoder/decoder using at least one of an inter-prediction indicator for a corresponding video of a level or reference picture index information.

Here, deriving a temporal merge candidate derives temporal merge candidate information (eg, at least one of a motion vector, a reference image index, an inter prediction indicator, and an image order count) from a corresponding position block in a corresponding position image to obtain a current block It may mean adding to the merge candidate list of .

The position of the corresponding position block may be the same position in the corresponding position image based on the current block position. Also, the position of the corresponding position block may be a position moved using at least one of motion vectors of temporal/spatial neighboring blocks on which coding/decoding has been completed based on the current block position.

For example, assuming that the corresponding position image of the image (or slice) level including the current block is set as the first reference image in the L1 direction, motion information in the L1 direction while spatial neighboring blocks are scanned in an arbitrary order. After scanning whether there is motion information in the L1 direction, a block corresponding to the current block may be derived using the motion information. Here, the spatial neighboring block may be a neighboring block of the current block for deriving the spatial merge candidate described above with reference to FIG. 9 .

As another example, assuming that the corresponding position image of the image (or slice) level including the current block is set as the first reference image in the L1 direction, motion information in the L1 direction while spatial neighboring blocks are scanned in an arbitrary order. If there is no motion information in the L1 direction after scanning to see if there is, it may be similarly scanned to see if there is motion information in the L0 direction. Here, when there is motion information for the L0 direction, a corresponding position block for the current block can be derived using the corresponding motion information.

In deriving a temporal merge candidate for the current block, as described with reference to FIG. 11 , motion vectors for each reference picture list obtained from the corresponding location block may be scaled according to an arbitrary reference picture of the current block.

In addition, after a plurality of scaled motion vectors are derived by performing motion vector scaling so that the motion vector obtained from the corresponding position block corresponds to at least one reference image among all reference images that can be referred to in the current block, a plurality of scaled motion vectors are derived. A plurality of prediction blocks may be generated using the scaled motion vector. Here, a final prediction block of the current block may be generated using a weighted sum of a plurality of generated prediction blocks.

Meanwhile, a temporal merge candidate for a sub-block of the current block may be derived using the above method, and a final prediction block of the sub-block of the current block may be generated.

For each reference image list, the distance (difference value of POC) between the image including the current block (or sub-block of the current block) and the reference image of the current block (or sub-block of the current block) is the corresponding position image and the corresponding position block ( Alternatively, if the distance (the difference value of POC) of the corresponding position block is different from the reference image, the temporal merge candidate is performed by scaling the motion vector derived from the corresponding position block (or the corresponding position block's lower block). can induce

And, if the reference image of the current block (or lower block of the current block) and the reference image of the corresponding position block (or lower block of the corresponding position block) are different for each reference image list, the corresponding position block (or lower block of the corresponding position block) is different. A temporal merge candidate may be derived by scaling the motion vector derived from block).

In addition, for each reference image list, the distance between the image (current picture) including the current block (or sub-block of the current block) and the reference image of the current block (or sub-block of the current block) (difference value of POC) is the corresponding position. The distance (POC difference value) between the image and the reference image of the corresponding position block (or sub-block of the corresponding position block) is different, and the reference image and corresponding position of the current block (or sub-block of the current block) are different for each reference image list. When reference images of blocks (or lower blocks of the corresponding position block) are different from each other, a temporal merge candidate may be derived by performing scaling on a motion vector derived from the corresponding position block (or lower block of the corresponding position block). Here, the reference image list is L0, L1, . . . according to inter-screen prediction indicators. , LX (X is a positive integer), and the like.

At this time, in performing scaling on the motion vector derived from the corresponding position block (or lower block of the corresponding position block), a reference image is determined for each reference image list for the current block (or lower block of the current block) , the motion vector may be scaled according to the determined reference image.

Here, in order to determine a reference image used for motion compensation using a temporal merge candidate (hereinafter referred to as “reference image for a temporal merge candidate”) or a reference image required for motion vector scaling, 1) N of the reference image list A method for selecting a th reference image, 2) a method for selecting a reference image most frequently selected as a reference image from spatial/temporal neighboring blocks where encoding/decoding is completed, 3) an RD cost function among spatial/temporal merge candidates of the current block A method of selecting a reference image of a merge candidate with a minimum value, 4) a method of selecting a reference image with a minimum RD cost function after performing motion prediction of the current block, and the like may be used.

1) How to select the Nth reference image in the reference image list

An N-th reference image of the reference image list of the current block may be determined as a reference image for a temporal merge candidate of the current block or a reference image necessary for motion vector scaling.

In this case, the N value may be a positive integer greater than 0, signaled from an encoder to a decoder, or may be a predetermined fixed value for the encoder/decoder. Also, the N value may be derived from the N value of the spatio-temporal neighboring block. The maximum value of N may mean the maximum number of reference images included in the reference image list.

For example, when N = 1, the corresponding position block (or corresponding position The motion information derived from the lower block of the block) may be scaled according to the first reference image of the reference image list of the current block. Here, the encoding parameter may include at least one of a motion vector, a reference picture index, an inter prediction indicator, or a picture order count.

2) A method of selecting the most selected reference image as a reference image among spatial/temporal neighboring blocks that have been encoded/decoded

A reference image most frequently selected as a reference image among spatial/temporal neighboring blocks that have been encoded/decoded may be determined as a reference image for a temporal merge candidate of the current block or a reference image necessary for motion vector scaling.

For example, in a corresponding location block (or a sub-block of a corresponding location block) based on an encoding parameter obtained from at least one block among a current block, a sub-block of the current block, or spatial/temporal blocks where encoding/decoding is completed. The derived motion information may be scaled according to a reference image selected the most as a reference image among spatial/temporal neighboring blocks that have been encoded/decoded.

3) A method of selecting a reference image of a merge candidate having a minimum RD cost function among spatial/temporal merge candidates of the current block

Among the spatial/temporal merge candidates of the current block, a reference image of a merge candidate having a minimum RD cost function may be determined as a reference image for a temporal merge candidate of the current block or a reference image necessary for motion vector scaling.

In this case, among the spatial/temporal merge candidates of the current block, index information of a merge candidate having a minimum RD cost function may be signaled from an encoder to a decoder or may be a preset fixed value in the encoder/decoder. Also, the index information may be derived from index information of a spatio-temporal neighboring block.

For example, in a corresponding location block (or a sub-block of a corresponding location block) based on an encoding parameter obtained from at least one block among a current block, a sub-block of the current block, or spatial/temporal blocks where encoding/decoding is completed. The derived motion information may be scaled according to a reference image of a merge candidate having a minimum RD cost function among spatial/temporal merge candidates of the current block.

4) A method of selecting a reference image having a minimum RD cost function after motion prediction of the current block is performed

After motion prediction of the current block is performed, a reference image having a minimum RD cost function may be determined as a reference image for a temporal merge candidate of the current block or a reference image necessary for motion vector scaling.

In this case, index information on the reference picture for which the RD cost function is minimized after motion prediction of the current block is signaled from the encoder to the decoder or may be a preset fixed value for the encoder/decoder. Also, the index information may be derived from index information of a spatio-temporal neighboring block.

For example, in a corresponding location block (or a sub-block of a corresponding location block) based on an encoding parameter obtained from at least one block among a current block, a sub-block of the current block, or spatial/temporal blocks where encoding/decoding is completed. The derived motion information may be scaled according to a reference picture having a minimum RD cost function after motion prediction of the current block is performed.

The distance (difference value of POC) between the image (current picture) including the current block (or sub-block of the current block) and the reference image of the current block (or sub-block of the current block) is the corresponding position image and the corresponding position block ( Or, although the distance (the difference value of POC) of the corresponding position block to the reference image is different, the reference image of the current block (or sub-block of the current block) and the corresponding position block (or corresponding position block) are different for each reference image list. When at least one reference image of a sub-block of ) is identical to each other, a reference image for a temporal merge candidate of the current block may be adaptively selected. In this case, after scaling a motion vector derived from a corresponding position block (or a lower block of the corresponding position block) according to the selected reference image, it may be used as a temporal merge candidate.

12 and 13 are diagrams for explaining an embodiment of a method of adaptively selecting a reference picture for a temporal merge candidate and a method of generating a prediction block using a temporal merge candidate according to the present invention.

12 and 13 assume a random access encoding/decoding environment in a group of picture (GOP) size of 16, a random square box indicates one picture, and the number in the box indicates the picture. Indicates a picture order count (POC), and when the inter prediction indicators for each picture are L0 and L1, a reference picture can be configured according to the respective inter prediction indicators. In the encoding/decoding environment of FIGS. 12 and 13, 0, 16, 8, 4, 2, 1, 3, 6, 5, 7, 12, 10, 9, 11, 14, 13, 15, 32, 24, 20, 18, 17, 19, 22, 21, 23, 28, 26, 25, 27, 30, 29, and 31 may be encoded/decoded.

In FIG. 12, when the POC of the currently encoded/decoded picture/slice is 20, POC 20 constructs reference images for POC 16 and POC 8 in the L0 direction and references POCs 24 and 32 in the L1 direction. It is assumed that the image is composed. In this case, it can be expressed as “L0: 16, 8” and “L1: 24, 32”, where POC 16 is the first reference image in the L0 reference image list, and POC 8 is the second reference image in the L0 reference image list. POC 24 may indicate the first reference image in the L1 reference image list, and POC 32 may indicate the second reference image in the L1 reference image list.

In FIG. 13, when the POC of the currently encoded/decoded picture/slice is 22, POC 22 constructs reference images for POC 20 and POC 16 in the L0 direction, and references POCs 24 and 32 in the L1 direction. It is assumed that the image is composed. In this case, it can be expressed as “L0: 20, 16” and “L1: 24, 32”, where POC 20 is the first reference image in the L0 reference image list, and POC 16 is the second reference image in the L0 reference image list. POC 24 may indicate the first reference image in the L1 reference image list, and POC 32 may indicate the second reference image in the L1 reference image list.

When the LX reference image list of the image (current picture) including the current block is the same as the LX reference image list of the co-located picture including the corresponding location block, a temporal merge candidate for the current block is selected. The reference image in the LX direction for 1) is selected as a reference image used by the LX-direction motion vector derived from the corresponding position block, or 2) selected as a reference image not used by the LX-direction motion vector derived from the corresponding position block. It can be. Here, X is an integer including 0, and LX can be expressed as L0, L1, L2, and the like.

As shown in FIG. 12, when it is assumed that the current picture is POC 20 and the co-located picture is POC 24, the current picture POC 20 and the co-located picture POC 24 are the same reference picture in the LO direction. In the case of having a list (L0: 16, 8), the reference image in the LO direction for the temporal merge candidate of the current block is 1) selected as a reference image used by the motion vector in the L0 direction derived from the corresponding block, or 2) Any one of the reference images that do not use the L0-direction motion vector derived from the location block may be selected.

For example, when the motion vector in the L0 direction derived from the corresponding location block in FIG. 12 uses POC 16 as the L0 reference image of the corresponding location block, POC 16 can be used as the LO reference image for the temporal merge candidate of the current block. , the corresponding L0 motion vector may be scaled according to the image of POC 16.

As another example, if the motion vector in the L0 direction derived from the corresponding location block in FIG. 12 uses POC 16 as the L0 reference image of the corresponding location block, POC 8 can be used as the LO reference image for the temporal merge candidate of the current block, , the corresponding L0 motion vector may be scaled according to the image of POC 8.

As shown in FIG. 13, when it is assumed that the current picture is POC 22 and the co-located picture is POC 20, the current picture POC 22 and the co-located picture POC 20 are the same reference picture in the L1 direction. In the case of having a list (L1: 24, 32), the L1-direction reference image for the temporal merge candidate of the current block is 1) selected as a reference image used by the L1-direction motion vector derived from the corresponding position block, or 2) corresponding One of the reference images that do not use the L1-direction motion vector derived from the location block may be selected.

For example, when the motion vector in the L1 direction derived from the corresponding location block in FIG. 13 uses POC 24 as the L1 reference image of the corresponding location block, POC 24 can be used as the L1 reference image for the temporal merge candidate of the current block. , the corresponding L0 motion vector may be scaled according to the image of POC 24.

As another example, when the motion vector in the L1 direction derived from the corresponding location block in FIG. 13 uses POC 24 as the L1 reference image of the corresponding location block, POC 32 can be used as the L1 reference image for the temporal merge candidate of the current block, , the corresponding L1 motion vector may be scaled according to the image of POC 32.

On the other hand, the LX reference image list of the image (current picture) including the current block and the LX reference image list of the co-located picture including the corresponding location block are not identical to each other, and only some reference images In the same case, a reference image in the LX direction for the temporal merge candidate of the current block may be selected as a reference image used by a motion vector in the LX direction derived from a corresponding block among some identical reference images.

Here, if the reference picture used by the LX-direction motion vector derived from the corresponding block is not included in the LX reference picture list of the current picture, the LX-direction reference picture for the temporal merge candidate of the current block is the LX reference of the current block. Among the images, a reference image closest to the current image may be selected.

12, when the motion vector in the L1 direction derived from the corresponding location block uses POC 32 as the L1 reference image of the corresponding location block, POC 32 can be used as the L1 reference image for the temporal merge candidate of the current block. , the L1 motion vector corresponding to this may be scaled according to the image of POC 32.

On the other hand, in the environment as shown in FIG. 12, when the motion vector in the L1 direction derived from the corresponding corresponding location block uses POC 16 as the L1 reference image of the corresponding location block, POC 24 is used as the L1 reference image for the temporal merge candidate of the current block. It can be selected and used, and the corresponding L1 motion vector can be scaled according to the selected reference image.

In the environment as shown in FIG. 13, when the motion vector in the L0 direction derived from the corresponding location block uses POC 16 as the L0 reference image of the corresponding location block, POC 16 can be used as the L0 reference image for the temporal merge candidate of the current block. , the corresponding L0 motion vector may be scaled according to the image of POC 16.

On the other hand, in the environment as shown in FIG. 13, when the motion vector in the L0 direction derived from the corresponding location block uses POC 8 as the L0 reference image of the corresponding location block, POC 20 is selected as the L0 reference image for the temporal merge candidate of the current block. and can be used, and the corresponding L0 motion vector can be scaled according to the selected reference image.

Meanwhile, in generating an inter-prediction block in the LX direction using a temporal merge candidate, the LX motion vector derived from the corresponding position block is scaled based on each reference image of the LX reference image list of the current block, and the scaled at least An inter prediction block may be generated using one LX motion vector. Here, at least one LX motion vector scaled based on each reference image of the LX reference image list of the current block may be derived as many as the number of LX reference images of the LX reference image list of the current block.

Accordingly, when there are a plurality of LX reference images of the LX reference image list of the current block, a plurality of scaled LX motion vectors can be derived.

As described above, when a plurality of scaled LX motion vectors are derived, LX of one temporal merge candidate using any one of the median value, average value, minimum value, maximum value, weighted average value, or mode among the plurality of scaled LX motion vectors A motion vector can be derived.

Alternatively, when a plurality of scaled LX motion vectors are derived, a plurality of prediction blocks may be generated using a plurality of scaled LX motion vectors corresponding to different LX reference images, respectively, and a A prediction block of one current block may be derived using any one of a median value, an average value, a minimum value, a maximum value, a weighted average value, or a mode value. Here, the weighted average value may be calculated based on a proportion according to a POC difference value between the current image and the reference image.

For example, in the environment shown in FIG. 12, when the motion vector in the L0 direction derived from the corresponding location block uses POC 16 as the L0 reference image of the corresponding location block, the L0 motion vector derived from the corresponding location block refers to L0 of the current block. Two scaled L0 motion vectors are derived based on each of the reference images (POC 16 and POC 8) in the image list, and based on the derived two L0 motion vectors, an inter-prediction block in the L0 direction using a temporal merge candidate is generated. can create

In this case, the L0 motion vector of one temporal merge candidate can be derived by using the average value between the two derived L0 motion vectors, and the L0 motion vector of one temporal merge candidate is the L0 reference image (POC 16, 8) of the current block. ) to generate a prediction block. Here, the L0 reference image of the current block applied to generate the prediction block may be determined by the above-described method (for example, the L0 reference image of the temporal merge candidate is the L0 reference image used by the L0 motion vector derived from the corresponding position block). It can be determined by video.)

Alternatively, two prediction blocks may be generated using two L0 motion vectors corresponding to different L0 reference images (POCs 16 and 8), respectively, and a weighted average value (or weighted sum) between the two generated prediction blocks is A prediction block of one current block can be derived by using. Specifically, in FIG. 12, since the POC difference between the reference image of POC 16 and the reference image of POC 8 with respect to the current image (POC 20) corresponds to 4 and 12, respectively, the prediction block corresponding to the reference image of POC 16 and POC 8 A prediction block may be generated by calculating a weighted sum to which a ratio of 3:1 is applied to the prediction block corresponding to the reference image of .

The method of generating the L1-direction inter-prediction block using the temporal merge candidate in FIG. 13 may be performed in the same manner as the method of generating the L0-direction inter-prediction block of FIG. 12 described above.

As another example, when the motion vector in the L1 direction derived from the corresponding location block in FIG. 12 uses POC 32 as the L1 reference image of the corresponding location block, the L1 motion vector derived from the corresponding location block is included in the L1 reference image list of the current block. Two scaled L1 motion vectors are derived based on each of the reference images (POC 24 and POC 32), and an L1-direction inter-prediction block can be generated using a temporal merge candidate based on the derived two L1 motion vectors. there is.

In this case, the L1 motion vector of one temporal merge candidate can be derived using the average value between the two derived L1 motion vectors, and the L1 motion vector of one temporal merge candidate is the L1 reference image (POC 24, 32 ) to generate a prediction block. Here, the L1 reference image of the current block applied to generate the prediction block may be determined by the method described above.

Alternatively, two prediction blocks may be generated using two L1 motion vectors corresponding to different L1 reference images (POCs 24 and 32), respectively, and a weighted average value (or weighted sum) between the generated two prediction blocks is A prediction block of one current block can be derived by using. Specifically, in FIG. 12, since the POC difference between the reference image of POC 24 and the reference image of POC 32 with respect to the current image (POC 20) corresponds to 4 and 12, respectively, the prediction block corresponding to the reference image of POC 24 and POC 32 A prediction block may be generated by calculating a weighted sum to which a ratio of 3:1 is applied to the prediction block corresponding to the reference image of .

The method of generating an L0-direction inter-prediction block using the temporal merge candidate in FIG. 13 may be performed in the same manner as the method of generating the L1-direction inter-prediction block of FIG. 12 described above.

14 and 15 are views for explaining an embodiment of a method of adaptively selecting a reference image for a temporal merge candidate in a low delay encoding/decoding environment (low delay configuration). It is assumed that FIG. 14 is a unidirectional predictive encoding/decoding environment using only the L0 reference picture list, and FIG. 15 is a bidirectional predictive encoding/decoding environment using the L0 reference picture list and the L1 reference picture list.

Even in a low-latency encoding/decoding environment, as in the random access encoding/decoding environment described above, the LX reference image list of the image (current picture) including the current block and the corresponding location image including the corresponding location block (Co-located If the LX reference image lists of picture) are not identical to each other and only some reference images are identical, the LX-direction motion vector derived from the corresponding block among the same partial reference images is used as the LX-direction reference image for the temporal merge candidate of the current block. can be selected as a reference image to That is, when the reference image used by the LX direction motion vector derived from the corresponding position block is included in the LX reference image list of the current image, the corresponding reference image may be selected as a reference image for the temporal merge candidate of the current block.

On the other hand, if the reference picture used by the LX-direction motion vector derived from the corresponding block is not included in the LX reference picture list of the current picture, the LX-direction reference picture for the temporal merge candidate of the current block is the LX reference of the current block. Among the images, a reference image closest to the current image may be selected.

For example, in the environment shown in FIG. 14, when a motion vector in the L0 direction derived from a corresponding location block uses POC 8 as the L0 reference image of the corresponding location block, POC 8 is used as the L0 reference image for the temporal merge candidate of the current block. It can be used, and the corresponding L0 motion vector can be scaled according to the image of POC 8.

On the other hand, in the environment as shown in FIG. 14, when the motion vector in the L0 direction derived from the corresponding corresponding location block uses POC 9 as the L0 reference image of the corresponding location block, POC 11 is used as the L0 reference image for the temporal merge candidate of the current block. It can be selected and used, and the corresponding L0 motion vector can be scaled according to the selected reference image.

In FIG. 15 , the method of generating inter prediction blocks in directions L0 and L1 using temporal merge candidates may be performed in the same manner as the method of generating inter prediction blocks in directions L0 in FIG. 14 described above.

Hereinafter, a method of deriving a temporal merge candidate for each divided sub-block by dividing the current block into sub-blocks will be described.

A temporal merge candidate for a sub-block of the current block may be derived from motion information of a sub-block of the corresponding position block (referred to as “corresponding position sub-block”). Here, ''' is a sub-block at a specific position within the corresponding position block, and may be a sub-block (for example, a center block or a corner block) at a fixed position pre-promised to an encoder/decoder. Alternatively, the corresponding position sub-block may be variably determined at the position of the current encoding/decoding target sub-block (referred to as “current sub-block”). For example, it may be a sub-block at the same location as the current sub-block or a sub-block corresponding to a location adjacent to the current sub-block. The corresponding location sub-block may be a sub-block divided into the same size as the current sub-block within the corresponding location block.

On the other hand, when the motion information of the corresponding position sub-block is unavailable (for example, when the corresponding position sub-block is intra-encoded or its use is restricted in deriving the motion information of the current sub-block), in the following way Temporal merge candidates for sub-blocks of the current block may be derived.

1) Temporal merge candidates for sub-blocks of the current block may be derived using motion information of spatial neighboring blocks of the current block that has been encoded/decoded. At this time, the first spatial neighboring block may be selected according to the spatial merge candidate search order, or the spatial neighboring block may be selected based on a predetermined priority.

Alternatively, a temporal merge candidate for a lower block of the current block may be derived using any one of a median value, an average value, a minimum value, a maximum value, a weighted average value, or a mode value of motion information of a plurality of spatial neighboring blocks of the current block.

2) A temporal merge candidate for a lower block of the current block may be derived using motion information of spatial neighboring blocks of the corresponding block. At this time, the first spatial neighboring block may be selected according to the spatial merge candidate search order, or the spatial neighboring block may be selected based on a predetermined priority.

Alternatively, a temporal merge candidate for a lower block of the current block may be derived using any one of a median value, an average value, a minimum value, a maximum value, a weighted average value, and a mode value among motion information of a plurality of spatial neighboring blocks of the corresponding position block. there is.

3) A temporal merge candidate of the current sub-block may be derived using motion information having a predetermined fixed value. For example, a zero motion vector having a magnitude of 0 in horizontal and vertical directions may be used as a temporal merge candidate for a current sub-block.

4) A temporal merge candidate of the current sub-block may be derived using motion information of other corresponding sub-blocks (left, top, right, bottom, diagonal, center).

For example, as shown in FIG. 16 , when motion information of a sub-block of a current corresponding position is unavailable, motion information of a sub-block of a center position in a corresponding position block may be derived as a temporal merge candidate of a current sub-block.

As another example, as shown in FIG. 17, when the motion information of the current corresponding position sub-block is unavailable, motion information of the left sub-block of the current corresponding-position sub-block and/or the top sub-block of the current corresponding-position sub-block within the corresponding position block is obtained. can be used to derive a temporal merge candidate of the current sub-block.

18 is a diagram for explaining an embodiment in which motion compensation is performed in sub-block units.

Referring to FIG. 18, a current block may be divided into four sub-blocks (A, B, C, and D), and a spatial neighboring block (or a sub-block of a spatial neighboring block) and a corresponding Motion compensation may be performed using at least one of motion information of a location block (or a lower block of a corresponding location block). Here, the motion information may include at least one of a motion vector, an inter-picture prediction indicator, a reference picture index, and a picture order count.

A process of performing motion compensation of sub-block A of the current block will be described in detail.

1) Obtaining motion information of spatial neighboring blocks and/or corresponding blocks

First, a plurality of pieces of motion information may be obtained by scanning spatial neighboring blocks of a lower block A of the current block in a predetermined order. Here, the predetermined order may be a pre-defined order in the encoder/decoder, or may be an order determined by information signaled from the encoder to the decoder.

For example, motion information may be obtained while scanning from left to right with sub-block a at the top of sub-block A as a starting point. If the upper sub-block a is unavailable, motion information may be acquired by scanning the upper sub-block b. That is, scanning may be performed until an upper sub-block having available motion information is found.

After obtaining motion information from the upper sub-block, motion information can be acquired while scanning from top to bottom with the viewpoint of sub-block c on the left side of sub-block A. If the left sub-block c is unavailable, motion information may be obtained by scanning the left sub-block d.

After obtaining motion information from the top sub-block and/or the left sub-block, motion information may be obtained from a corresponding block of sub-block A. In this case, the corresponding location block of sub-block A may be a corresponding location block of the current block including sub-block A or a sub-block of a corresponding location block directly corresponding to sub-block A. A sub-block of the corresponding location block that directly corresponds to sub-block A can be selected as described above. Alternatively, motion information may be obtained from a corresponding position block (or a lower block of the corresponding position block) at a position compensated by a motion vector of a spatial neighboring block of the current block.

In the above method, up to three pieces of motion information can be obtained from the top sub-block, left sub-block, and corresponding position block, and up to N pieces of motion information can be obtained according to an embodiment. Here, N is a positive integer greater than or equal to 2.

2-1) Selection of reference images for motion compensation and generation of motion vectors of sub-blocks

The reference image for motion compensation of the sub-block A is a) the first reference image in the reference image list, and b) the reference image when motion vectors obtained from the spatially neighboring sub-blocks among the three pieces of acquired motion information have the same reference image. or c) If at least one of reference images of motion vectors obtained from a spatial neighboring sub-block among the obtained three pieces of motion information is identical to a reference image of a motion vector obtained from a corresponding location block, a corresponding reference image may be selected.

Then, each of up to three motion vectors is scaled based on the selected reference image, and any one of a median value, an average value, a minimum value, a maximum value, a weighted average value, or a mode value of the scaled maximum three motion vectors is used to determine sub-block A. A motion vector can be created.

A prediction block of the sub-block A may be generated using the motion vector of the sub-block A generated as above and the selected reference image.

2-2) Generation of multiple prediction blocks and final prediction blocks

After obtaining up to 3 pieces of motion information from the top sub-block, left sub-block and corresponding position block, and obtaining up to 3 prediction blocks through motion compensation from motion vectors of each of the obtained motion information and reference images corresponding thereto, , a weighted sum of these can generate a prediction block for sub-block A.

In this case, the motion vector of the block corresponding to the sub-block A may be scaled based on the reference image selected in 2-1) above. In this case, a prediction block may be obtained through motion compensation based on the selected reference image and the scaled motion vector.

19 is a flowchart illustrating an image decoding method according to an embodiment of the present invention.

Referring to FIG. 19 , a spatial merge candidate may be derived from at least one spatial neighboring block of the current block (S1901).

And, a temporal merge candidate can be derived from the block corresponding to the current block (S1902).

Here, the derivation of the spatial neighboring block and the temporal merge candidate has been specifically described in FIGS. 9 to 11, so it will be omitted.

In addition, a prediction block of the current block may be generated based on at least one of the derived spatial merge candidate and temporal merge candidate (S1903).

In this case, the reference picture for the temporal merge candidate may be selected based on the reference picture list of the current picture including the current block and the reference picture list of the corresponding position picture including the corresponding position block. Here, the reference image for the temporal merge candidate may be a reference image used for motion compensation using the temporal merge candidate or a reference image used for scaling the motion vector of the corresponding position block.

Meanwhile, a reference image for a temporal merge candidate may be selected based on whether the reference image list of the current image and the reference image list of the corresponding position image are the same.

For example, when the reference image list of the current image and the reference image list of the corresponding location image are the same, a reference image for a temporal merge candidate may be selected as a reference image used by a motion vector derived from a corresponding location block.

As another example, when at least one reference image is the same in the reference image list of the current image and the reference image list of the corresponding location image, one of the at least one identical reference image may be selected as the reference image for the temporal merge candidate.

Meanwhile, reference images for temporal merge candidates may be selected according to inter prediction directions. Here, the inter-screen prediction directions are L0, L1,... ,LX (X is a positive integer).

Meanwhile, the spatial merge candidate and the temporal merge candidate of the current block may be derived in units of sub-blocks of the current block.

In this case, a temporal merge candidate of a sub-block of the current block may be derived from a sub-block at the same location as a sub-block of the current block included in a corresponding block.

However, if the sub-block at the same position is unavailable, the temporal merge candidates of the sub-blocks of the current block are the sub-block at the center of the corresponding block, the left sub-block of the same-position sub-block, and the top sub-block of the same-position sub-block. It can be derived from any one of the sub-blocks.

Meanwhile, in the step of deriving a temporal merge candidate (S1902), the motion vector of the corresponding block is scaled based on each of the reference images of the reference image list of the current block, and the temporal merge candidate includes a plurality of scaled motion vectors. can induce

For example, a prediction block of the current block may be generated using a motion vector generated based on a weighted sum of a plurality of scaled motion vectors.

As another example, a plurality of temporary prediction blocks may be generated using each of a plurality of scaled motion vectors, and a prediction block of the current block may be generated based on a weighted sum of the plurality of generated temporary prediction blocks.

Meanwhile, in the step of deriving a temporal merge candidate ( S1902 ), a temporal merge candidate may be derived by scaling motion information of a corresponding location block based on a reference image for the temporal merge candidate.

In this case, deriving the temporal merge candidate by scaling the motion information of the corresponding position block is the image order count value between the current image including the current block and the reference image of the current block, and the corresponding position image including the corresponding position block and It may be selectively performed based on an image order count value between reference images of corresponding location blocks.

The intra-screen encoding/decoding process may be performed on each of the luminance and color difference signals. For example, in the intra-coding/decoding process, at least one method of deriving an intra-prediction mode, dividing a block, constructing a reference sample, and performing intra-prediction may be differently applied to a luminance signal and a chrominance signal.

The intra-screen encoding/decoding process for the luminance and color difference signals can be performed in the same way. For example, in the intra-prediction/decoding process applied to the luminance signal, at least one of intra-prediction mode derivation, block segmentation, reference sample configuration, and intra-prediction may be equally applied to the chrominance signal.

The above methods can be performed in the same way in the encoder and decoder. For example, in the process of intra-coding/decoding, at least one method of deriving an intra-prediction mode, dividing a block, constructing a reference sample, and performing intra-prediction may be equally applied to an encoder and a decoder. In addition, the order of applying the above methods may be different between the encoder and the decoder. For example, in performing intra-picture/decoding of the current block, the encoder may construct a reference sample and then encode the determined intra-prediction mode by performing one or more intra-prediction.

The above embodiments of the present invention may be applied according to the size of at least one of a coding block, a prediction block, a block, and a unit. The size herein may be defined as a minimum size and/or a maximum size for the above embodiments to be applied, or may be defined as a fixed size to which the above 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 above embodiments may be applied in a complex manner according to the size. In addition, the above embodiments of the present invention can be applied only when the size is greater than the minimum size and less than the maximum size. That is, the above embodiments can be applied only when the block size is included within a certain range.

For example, the above embodiments can be applied only when the size of an encoding/decoding target block is 8x8 or larger. For example, the above embodiments can be applied only when the size of an encoding/decoding object block is 16x16 or larger. For example, the above embodiments can be applied only when the size of an encoding/decoding target block is 32x32 or larger. For example, the above embodiments can be applied only when the size of an encoding/decoding object block is 64x64 or larger. For example, the above embodiments can be applied only when the size of an encoding/decoding object block is 128x128 or larger. For example, the above embodiments can be applied only when the size of an encoding/decoding object block is 4x4. For example, the above embodiments can be applied only when the size of an encoding/decoding target block is 8x8 or less. For example, the above embodiments may be applied only when the size of an encoding/decoding target block is 16x16 or less. For example, the above embodiments can be applied only when the size of an encoding/decoding object block is greater than or equal to 8x8 and less than or equal to 16x16. For example, the above embodiments can be applied only when the size of an encoding/decoding target block is greater than or equal to 16x16 and less than or equal to 64x64.

The above embodiments of the present invention may be applied according to a temporal layer. A separate identifier is signaled to identify a temporal layer to which the above embodiments are applicable, and the above embodiments can be applied to the temporal layer specified by the identifier. The identifier herein may be defined as a minimum layer and/or a maximum layer to which the embodiment is applicable, or may be defined as indicating a specific layer to which the embodiment is applied.

For example, the above embodiments may be applied only when the temporal layer of the current video is the lowest layer. For example, the above embodiments may be applied only when the temporal layer identifier of the current video is 0. For example, the above embodiments can be applied only when the temporal layer identifier of the current video is 1 or more. For example, the above embodiments may be applied only when the temporal layer of the current video is the highest layer.

As in the above embodiment of the present invention, the reference picture set used in the process of reference picture list construction and reference picture list modification is L0, L1, L2, L3. At least one or more reference image lists may be used.

According to the above embodiment of the present invention, one or more and up to N motion vectors of a block to be encoded/decoded may be used when calculating boundary strength in a deblocking filter. Here, N represents a positive integer greater than or equal to 1, and may be 2, 3, 4, and the like.

When predicting the motion vector, the motion vector is 16-pixel (16-pel) unit, 8-pixel (8-pel) unit, 4-pixel (4-pel) unit, integer-pixel unit, 1/2 -Pixel (1/2-pel) unit, 1/4-pixel (1/4-pel) unit, 1/8-pixel (1/8-pel) unit, 1/16-pixel (1/16-pel) ) unit, 1/32-pixel (1/32-pel) unit, and 1/64-pixel (1/64-pel) unit, the above embodiments of the present invention can also be applied. Also, when motion vector prediction is performed, a motion vector may be selectively used for each pixel unit.

A slice type to which the above embodiments of the present invention are applied is defined, and the above embodiments of the present invention can be applied according to the corresponding slice type.

For example, when the slice type is T (Tri-predictive)-slice, a prediction block is generated using at least three motion vectors, and a weighted sum of the at least three prediction blocks is calculated to obtain an encoding/decoding target block. can be used as the final prediction block of For example, when the slice type is Q (Quad-predictive)-slice, a prediction block is generated using at least 4 or more motion vectors, and a weighted sum of the at least 4 or more prediction blocks is calculated to obtain an encoding/decoding target block. can be used as the final prediction block of

The above embodiments of the present invention can be applied not only to inter-prediction and motion compensation methods using motion vector prediction, but also to inter-prediction and motion compensation methods using skip mode and merge mode.

The shape of a block to which the above embodiments of the present invention are applied may have a square shape or a non-square shape.

In the foregoing embodiments, the methods are described on the basis of a flow chart as a series of steps or units, but the present invention is not limited to the order of steps, and some steps may occur in a different order or concurrently with other steps as described above. can In addition, those skilled in the art will understand that the steps shown in the flow chart are not exclusive, that other steps may be included, or that one or more steps of the flow chart may be deleted without affecting the scope of the present invention. You will understand.

The foregoing embodiment includes examples of various aspects. It is not possible to describe all possible combinations to represent the various aspects, but those skilled in the art will recognize that other combinations are possible. Accordingly, it is intended that the present invention cover all other substitutions, modifications and variations falling within the scope of the following claims.

Embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded on a computer-readable recording medium. The computer readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the computer-readable recording medium may be specially designed and configured for the present invention, or may be known and usable to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floptical disks. media), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter or the like as well as machine language codes such as those produced by a compiler. The hardware device may be configured to act as one or more software modules to perform processing according to the present invention and vice versa.

In the above, the present invention has been described by specific details such as specific components and limited embodiments and drawings, but these are provided to help a more general understanding of the present invention, and the present invention is not limited to the above embodiments. , Those skilled in the art to which the present invention pertains may seek various modifications and variations from these descriptions.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments, and not only the claims described later, but also all modifications equivalent or equivalent to these claims belong to the scope of the spirit of the present invention. will do it

Claims (20)

In the video decoding method performed by the decoding device,
obtaining image information including information related to motion compensation through a bitstream;
Deriving a prediction mode for the current block;
deriving a motion information list including motion information candidates derived based on neighboring blocks of the current block;
deriving motion information of the current block based on the motion information list;
generating a predicted block of the current block based on the motion information; and
Generating a restoration block based on the predicted block,
The video decoding method, characterized in that the information related to the motion compensation includes motion vector precision information. According to claim 1,
Wherein the motion vector precision information indicates one of motion vector precision candidates including integer pixel precision, 1/4 pixel precision and 1/8 pixel precision. According to claim 1,
The video decoding method, characterized in that the motion vector precision information indicates one of the motion vector precision candidates as a picture level parameter. According to claim 1,
The video decoding method of claim 1 , wherein the motion vector precision information represents motion vector precision for motion vector differences. According to claim 1,
The motion vector precision information may be entropy-decoded based on at least one of a block size and a block depth of the current block. According to claim 1,
Characterized in that the motion vector precision information is shared with respect to all blocks in a maximum coding unit, video decoding method. According to claim 1,
Characterized in that the motion vector precision information is decoded from header syntax, video decoding method. In the video encoding method performed by the encoding device,
Deriving a prediction mode for the current block;
deriving a motion information list including motion information candidates derived based on neighboring blocks of the current block;
generating information related to motion compensation; and
Encoding video information including information related to the motion compensation;
The video encoding method, characterized in that the information related to the motion compensation includes motion vector precision information. According to claim 8,
The video encoding method of claim 1 , wherein the motion vector precision information indicates one of motion vector precision candidates including integer pixel precision, 1/4 pixel precision and 1/8 pixel precision. According to claim 8,
The video encoding method, characterized in that the motion vector precision information indicates one of the motion vector precision candidates as a picture level parameter. According to claim 8,
The video encoding method of claim 1 , wherein the motion vector precision information indicates motion vector precision for at least one of a motion vector and a motion vector difference. According to claim 8,
The motion vector precision information may be entropy-encoded based on at least one of a block size and a block depth of the current block. According to claim 8,
The video encoding method, characterized in that the motion vector precision information is shared for all blocks in a maximum coding unit. According to claim 8,
The video encoding method, characterized in that the motion vector precision information is encoded in header syntax. In the transmission method for image data,
Obtaining a bitstream of encoded image information, wherein the encoded image information derives a prediction mode for a current block; a motion information list including motion information candidates derived based on neighboring blocks of the current block; generating based on the steps of deriving, generating information related to motion compensation, and encoding image information including the information related to motion compensation; and
Transmitting the image data including the bitstream,
The transmission method, characterized in that the information related to the motion compensation includes motion vector precision information. According to claim 15,
characterized in that the motion vector precision information indicates one of motion vector precision candidates including integer pixel precision, 1/4 pixel precision and 1/8 pixel precision. According to claim 15,
The transmission method, characterized in that the motion vector precision information indicates one of motion vector precision candidates as a picture level parameter. According to claim 15,
The transmission method, characterized in that the motion vector precision information indicates motion vector precision for at least one of a motion vector and a motion vector difference. According to claim 15,
The transmission method, characterized in that the motion vector precision information is entropy-encoded based on at least one of a block size and a block depth of the current block. According to claim 15,
Characterized in that the motion vector precision information is encoded in header syntax.

Images