WO2019194653A1 - Image processing method for providing complex merge mode process of motion information, image decoding and encoding method using same, and apparatus thereof - Google Patents

Abstract

An image processing method according to an embodiment of the present invention comprises the steps of: dividing a picture of an image into a plurality of coding units that are basic units in which inter prediction or intra prediction is to be performed, and determining a current block for decoding a coding unit which divides the picture or the divided coding unit into quad tree and binary tree structures; variably constructing a merged mode motion information candidate list for predicting merged mode motion information of the current block, according to one or more merge mode candidates derived corresponding to the current block; and performing merge mode motion information prediction decoding of the current block by using the merge mode motion information candidate list.

Description

An image processing method for providing complex merge mode processing of motion information, an image decoding, encoding method, and apparatus therefor

The present invention relates to an image processing method, an image decoding and encoding method using the same, and an apparatus thereof, and more particularly, to an image processing method for providing complex merge mode processing of motion information, an image decoding and encoding method using the same, and an apparatus thereof. It is about.

Digital video technologies include digital television, digital direct broadcast systems, wireless broadcast systems, personal digital assistants (PDAs), laptop or desktop computers, digital cameras, digital recording devices, video gaming devices, video game consoles, cellular or satellite radio telephones, etc. It can be integratedly applied to a wide range of digital video devices, including. Digital video devices include video compression technologies such as MPEG-2, MPEG-4, or ITU-T H.264 / MPEG-4, Part 10, Advanced Video Coding (AVC), and High Efficiency Video Coding (HVC). By implementing the digital video information transmission and reception more efficiently. Video compression techniques perform spatial prediction and temporal prediction to remove or reduce redundancy inherent in video sequences.

Such image compression techniques include an inter prediction technique for predicting pixel values included in a current picture from a picture before or after the current picture, and an intra prediction for predicting pixel values included in a current picture using pixel information in the current picture. There are various technologies such as technology, entropy encoding technology that assigns a short code to a high frequency of appearance and long code to a low frequency of frequency, and it is possible to effectively compress and transmit or store image data using such image compression technology. .

In order to cost-effectively cope with various resolutions, frame rates, etc. according to such applications, a video decoding apparatus that can be easily processed according to the performance and function required for the application must be provided.

For example, in the image compression method, one picture is divided into a plurality of blocks having a predetermined size and encoding is performed. In addition, inter prediction and intra prediction techniques that remove redundancy between pictures are used to increase compression efficiency.

In this case, a residual signal is generated by using intra prediction and inter prediction, and the reason for obtaining the residual signal is that when coding with the residual signal, the amount of data is small and the data compression ratio is high, and the better the prediction, the residual signal. This is because the value of becomes small.

The intra prediction method predicts data of the current block by using pixels around the current block. The difference between the actual value and the predicted value is called the residual signal block. In the case of HEVC, the intra prediction method is increased from nine prediction modes used in H.264 / AVC to 35 prediction modes to further refine the prediction.

In the case of the inter prediction method, the most similar block is found by comparing the current block with blocks in neighboring pictures. At this time, the position information (Vx, Vy) of the found block is called a motion vector. The difference between pixel values in a block between the current block and the prediction block predicted by the motion vector is called a residual signal block (motion-compensated residual block).

As the intra prediction and the inter prediction are further subdivided, the amount of data of the residual signal is reduced, but the amount of computation for processing a video has greatly increased.

In particular, in the intra and inter prediction processes for image encoding and decoding, due to the complexity increase due to the division structure division and the difficulty of prediction, not only the encoding efficiency but also the decoder implementation and parallel processing exist. The prediction method may not be suitable for encoding a high resolution image.

The present invention is to solve the above problems, and to provide an image decoding, encoding method and apparatus for improving the efficiency of motion information prediction decoding and encoding of a high resolution image according to the diversification and segmentation of the partition structure. There is this.

In an image processing method according to an embodiment of the present invention for solving the above problems, a plurality of coding units in which a picture of an image is a basic unit in which inter prediction or intra prediction is performed Determining a current block for decoding the coding unit divided into the plurality of frames), and the coding unit obtained by dividing the picture or the divided coding unit into a quad tree and a binary tree structure; Variably constructing a merge mode motion information candidate list for predicting merge mode motion information of the current block according to one or more merge mode candidates derived corresponding to the current block; And performing merge mode motion information prediction decoding of the current block by using the merge mode motion information candidate list.

In addition, a method according to an embodiment of the present invention for solving the above problems, the image decoding method, comprising: receiving an encoded bitstream; Performing inverse quantization and inverse transformation on the input bitstream to obtain a residual block; Determining a current block for decoding a coding unit having a picture of an image divided into quad tree and binary tree structures corresponding to the residual block; Variably constructing a merge mode motion information candidate list for predicting merge mode motion information of the current block according to one or more merge mode candidates derived corresponding to the current block; Obtaining a prediction block by performing merge mode motion information prediction decoding of the current block by using the merge mode motion information candidate list; And reconstructing an image by using the prediction block and the residual block.

In addition, the method according to an embodiment of the present invention for solving the above problems, the image encoding method, comprising: dividing a picture of the image into a plurality of coding units divided into quad tree and binary tree structure; Determining a current block for encoding the coding unit; Variably constructing a merge mode motion information candidate list for predicting merge mode motion information of the current block according to one or more merge mode candidates derived corresponding to the current block; And performing prediction mode decoding of merge mode motion information of the current block by using the merge mode motion information candidate list to obtain a prediction block.

On the other hand, the method according to an embodiment of the present invention for solving the above problems can be implemented as a program for executing the method on a computer and a non-volatile recording medium that is stored in the computer can be read by the computer.

According to an embodiment of the present invention, an image decoding and encoding method for improving the efficiency of motion information prediction decoding and encoding of a high resolution image according to diversification and segmentation of a partition structure by providing efficient processing of motion information for a merge mode And the apparatus.

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

2 to 5 are diagrams for describing a first exemplary embodiment of a method of dividing and processing an image in block units.

6 is a block diagram illustrating an embodiment of a method of performing inter prediction in an image encoding apparatus.

7 is a block diagram illustrating a configuration of an image decoding apparatus according to an embodiment of the present invention.

8 is a block diagram illustrating an embodiment of a method of performing inter prediction in an image decoding apparatus.

FIG. 9 is a diagram for describing a second exemplary embodiment of a method of dividing and processing an image into blocks.

FIG. 10 is a diagram for describing a third embodiment of a method of dividing and processing an image into blocks.

FIG. 11 is a diagram for describing an example of a method of configuring a transform unit by dividing a coding unit into a binary tree structure.

FIG. 12 is a diagram for describing a fourth embodiment of a method of dividing and processing an image into blocks.

13 to 14 are diagrams for describing still another example of a method of dividing and processing an image into blocks.

15 and 16 are diagrams for describing embodiments of a method of determining a partition structure of a transform unit by performing rate distortion optimization (RDO).

17 is a block diagram illustrating in more detail the merge mode motion information decoder according to an embodiment of the present invention.

18 is a flowchart illustrating a merge mode candidate derivation and candidate list construction method according to an embodiment of the present invention.

19 to 20 are diagrams for explaining merge mode candidate derivation methods according to an embodiment of the present invention for each derivation process.

21 is a flowchart illustrating a method of constructing and updating a merge mode motion information candidate list according to an embodiment of the present invention.

22 to 24 are diagrams for explaining a configuration and update process of a merge mode motion information candidate list according to an embodiment of the present invention.

25 to 27 are diagrams for explaining high level syntax of header information according to an embodiment 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 the present specification, when it is determined that a detailed description of a related well-known configuration or function may obscure the gist of the present specification, the detailed description thereof will be omitted.

When a component is said to be "connected" or "connected" to another component, it may be directly connected to or connected to that other component, but it may be understood that another component may be present in between. Should be. In addition, the content described as "include" a specific configuration in the present invention does not exclude a configuration other than the configuration, it means that additional configuration may be included in the scope of the technical idea of the present invention or 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. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as the second component, and similarly, the second component may also be referred to as the first component.

In addition, the components shown in the embodiments of the present invention are shown independently to represent different characteristic functions, and do not mean that each component is made of separate hardware or one software component unit. In other words, each component is included in each component for convenience of description, and at least two of the components may be combined into one component, or one component may be divided into a plurality of components to perform a function. Integrated and separate embodiments of the components are also included within the scope of the present invention without departing from the spirit of the invention.

In addition, some of the components may not be essential components for performing essential functions in the present invention, but may be optional components for improving performance. The present invention can be implemented including only the components essential for implementing the essentials of the present invention except for the components used for improving performance, and the structure including only the essential components except for the optional components used for improving performance. Also included in the scope of the present invention.

1 is a block diagram illustrating a configuration of an image encoding apparatus according to an embodiment of the present invention. The image encoding apparatus 10 may include a picture divider 110, a transform unit 120, a quantization unit 130, and a scanning unit. The unit 131, the entropy encoder 140, the intra predictor 150, the inter predictor 160, the inverse quantizer 135, the inverse transformer 125, the post processor 170, and the picture storage 180 ), A subtractor 190 and an adder 195.

Referring to FIG. 1, the picture dividing unit 110 analyzes an input video signal, divides a picture into coding units, determines a prediction mode, and determines a size of a prediction unit for each coding unit.

In addition, the picture splitter 110 sends the prediction unit to be encoded to the intra predictor 150 or the inter predictor 160 according to a prediction mode (or a prediction method). In addition, the picture dividing unit 110 sends the prediction unit to be encoded to the subtracting unit 190.

Here, a picture of an image may be composed of a plurality of slices, and the slice may be divided into a plurality of coding tree units (CTUs) which are basic units for dividing a picture.

Also, a picture may be configured as a tile group, and when configured as a tile group, a tile group may be configured with one or more tiles. At this time, each tile may be divided into the CTUs. Here, the tile group related encoding information may be signaled through a tile group header. When a plurality of tile groups exist, they may be divided into tile group identification information, and encoding options may be selectively and adaptively applied between different tile groups.

The coding tree unit may be divided into one or two coding units (CUs), which are basic units on which inter prediction or intra prediction is performed.

The coding unit (CU) may be divided into one or more prediction units (PUs), which are basic units on which prediction is performed.

In this case, the encoding apparatus 10 determines one of inter prediction and intra prediction as a prediction method for each of the divided coding units (CUs), but differently predicts a prediction block for each prediction unit (PU). Can be generated.

Meanwhile, the coding unit CU may be divided into one or two transform units (TUs), which are basic units for transforming a residual block.

In this case, the picture dividing unit 110 may transmit the image data to the subtracting unit 190 in a block unit (for example, a prediction unit (PU) or a transformation unit (TU)) divided as described above.

Referring to FIG. 2, a coding tree unit (CTU) having a maximum size of 256 × 256 pixels may be divided into a quad tree structure and divided into four coding units (CUs) having a square shape.

The four coding units (CUs) having the square shape may be re-divided into quad tree structures, respectively, and the depth of the coding units CU divided into quad tree structures as described above may be any one of 0 to 3. It can have one integer value.

The coding unit CU may be divided into one or two or more prediction units (PUs) according to the prediction mode.

In the case of the intra prediction mode, when the size of the coding unit CU is 2Nx2N, the prediction unit PU may have a size of 2Nx2N shown in FIG. 3A or NxN shown in FIG. 3B. have.

Meanwhile, in the inter prediction mode, when the size of the coding unit CU is 2Nx2N, the prediction unit PU is 2Nx2N shown in FIG. 4A, 2NxN shown in FIG. 4B, and FIG. 4. Nx2N shown in (c) of FIG. 4, NxN shown in (d) of FIG. 4, 2NxnU shown in (e) of FIG. 4, 2NxnD shown in (f) of FIG. 4, shown in (g) of FIG. It may have a size of any one of nLx2N and nRx2N shown in (h) of FIG.

Referring to FIG. 5, the coding unit CU may be divided into a quad tree structure and divided into four transform units TUs having a square shape.

The four transform units (TUs) having a square shape may be re-divided into quad tree structures, and the depth of the transform units (TUs) divided into quad tree structures as described above may be any one of 0 to 3. It can have one integer value.

Here, when the coding unit CU is in the inter prediction mode, the prediction unit PU and the transform unit TU split from the coding unit CU may have a partition structure that is independent of each other.

When the coding unit CU is in the intra prediction mode, the transform unit TU split from the coding unit CU cannot be larger than the size of the prediction unit PU.

In addition, the transform unit TU divided as described above may have a maximum size of 64x64 pixels.

The transform unit 120 converts a residual block that is a residual signal between the original block of the input prediction unit PU and the prediction block generated by the intra predictor 150 or the inter predictor 160, and the transform is performed. It may be performed using the unit (TU) as a basic unit.

In the transformation process, different transform matrices may be determined according to a prediction mode (intra or inter), and since the residual signal of intra prediction has a direction according to the intra prediction mode, the transform matrix may be adaptively determined according to the intra prediction mode. have.

The transform unit may be transformed by two (horizontal and vertical) one-dimensional transform matrices. For example, in the case of inter prediction, one predetermined transform matrix may be determined.

On the other hand, in the case of intra prediction, when the intra prediction mode is horizontal, the probability of the residual block having the directionality in the vertical direction increases, so a DCT-based integer matrix is applied in the vertical direction, and DST-based or in the horizontal direction. Apply KLT-based integer matrix. When the intra prediction mode is vertical, an integer matrix based on DST or KLT may be applied in the vertical direction and a DCT based integer matrix in the horizontal direction.

In addition, in the DC mode, a DCT based integer matrix may be applied in both directions.

In the case of intra prediction, a transform matrix may be adaptively determined based on the size of a transform unit (TU).

The quantization unit 130 determines a quantization step size for quantizing the coefficients of the residual block transformed by the transform matrix, and the quantization step size may be determined for each quantization unit having a predetermined size or more.

The size of the quantization unit may be 8x8 or 16x16, and the quantization unit 130 quantizes coefficients of the transform block using a quantization matrix determined according to the quantization step size and the prediction mode.

In addition, the quantization unit 130 may use the quantization step size of the quantization unit adjacent to the current quantization unit as the quantization step size predictor of the current quantization unit.

The quantization unit 130 may search for the left quantization unit, the upper quantization unit, and the upper left quantization unit of the current quantization unit and generate a quantization step size predictor of the current quantization unit using one or two valid quantization step sizes. have.

For example, the quantization unit 130 may determine a valid first quantization step size found in the order as a quantization step size predictor, or determine an average value of two valid quantization step sizes found in the order as a quantization step size predictor, or If only one quantization step size is valid, this may be determined as a quantization step size predictor.

When the quantization step size predictor is determined, the quantization unit 130 transmits a difference value between the quantization step size and the quantization step size predictor of the current quantization unit to the entropy encoder 140.

On the other hand, the left coding unit, the upper coding unit, the upper left coding unit of the current coding unit does not all exist. Or there may be a coding unit previously present in the coding order within the largest coding unit.

Accordingly, candidates may be quantization step sizes of the quantization units adjacent to the current coding unit and the quantization unit immediately before the coding order within the maximum coding unit.

In this case, priority is set in the order of 1) the left quantization unit of the current coding unit, 2) the upper quantization unit of the current coding unit, 3) the upper left quantization unit of the current coding unit, and 4) the quantization unit immediately preceding the coding order. Can be. The order may be reversed and the upper left quantization unit may be omitted.

Meanwhile, the transform block quantized as described above is transferred to the inverse quantization unit 135 and the scanning unit 131.

The scanning unit 131 scans the coefficients of the quantized transform block and converts them into one-dimensional quantization coefficients. In this case, since the distribution of coefficients of the transform block after quantization may depend on the intra prediction mode, the scanning method is applied to the intra prediction mode. Can be determined accordingly.

Further, the coefficient scanning scheme may be determined differently according to the size of the transform unit, and the scan pattern may vary according to the directional intra prediction mode, in which case the scanning order of the quantization coefficients may be scanned in the reverse direction.

When the quantized coefficients are divided into a plurality of subsets, the same scan pattern may be applied to the quantization coefficients in each subset, and a zigzag scan or a diagonal scan may be applied to the scan patterns between the subsets.

Meanwhile, the scan pattern is preferably scanned in the forward direction from the main subset including DC to the remaining subsets, but the reverse direction is also possible.

In addition, a scan pattern between subsets may be set to be identical to a scan pattern of quantized coefficients in a subset, and the scan pattern between subsets may be determined according to an intra prediction mode.

Meanwhile, the encoding apparatus 10 may include information indicative of the position of the last non-zero quantization coefficient and the position of the last non-zero quantization coefficient in each subset in the transform unit PU to include the decoding apparatus ( 20).

The inverse quantization unit 135 inverse quantizes the quantized coefficients as described above, and the inverse transform unit 125 performs inverse transformation in units of transform units (TUs) to restore the inverse quantized transform coefficients into a residual block of a spatial domain. can do.

The adder 195 may generate a reconstructed block by adding the residual block reconstructed by the inverse transform unit 125 and the received prediction block from the intra predictor 150 or the inter predictor 160.

In addition, the post-processing unit 170 may perform a deblocking filtering process to remove the blocking effect occurring in the reconstructed picture, and a sample adaptive offset to compensate for the difference value from the original image in pixel units. A SAO) application process and a coding unit can perform an adaptive loop filtering (ALF) process to compensate for a difference value from an original image.

The deblocking filtering process may be applied to the boundary of the prediction unit (PU) or transform unit (TU) having a size of a predetermined size or more.

For example, the deblocking filtering process may include determining a boundary to filter, determining a boundary filtering strength to be applied to the boundary, determining whether to apply a deblocking filter, If it is determined to apply the deblocking filter, the method may include selecting a filter to be applied to the boundary.

On the other hand, whether the deblocking filter is applied depends on whether i) the boundary filtering intensity is greater than 0 and ii) the degree of change of pixel values at the boundary portions of two blocks (P block, Q block) adjacent to the boundary to be filtered. The value represented may be determined by whether the value is smaller than the first reference value determined by the quantization parameter.

It is preferable that the said filter is at least 2 or more. When the absolute value of the difference between two pixels positioned at the block boundary is greater than or equal to the second reference value, a filter that performs relatively weak filtering is selected.

The second reference value is determined by the quantization parameter and the boundary filtering intensity.

In addition, the sample adaptive offset (SAO) application process is to reduce the distortion (distortion) between the pixel and the original pixel in the image to which the deblocking filter is applied, the sample adaptive offset (SAO) application process in the unit of picture or slice. Whether to perform may be determined.

The picture or slice may be divided into a plurality of offset regions, and an offset type may be determined for each offset region, and the offset type may be a predetermined number of edge offset types (eg, four) and two band offsets. It can include a type.

For example, when the offset type is an edge offset type, an edge type to which each pixel belongs is determined and an offset corresponding thereto is applied, and the edge type may be determined based on a distribution of two pixel values adjacent to the current pixel. have.

The adaptive loop filtering (ALF) process may perform filtering based on a value obtained by comparing a reconstructed image and an original image that have undergone a deblocking filtering process or an adaptive offset application process.

The picture storage unit 180 receives the post-processed image data from the post-processing unit 170 and restores the image in a picture unit, and the picture may be an image in a frame unit or an image in a field unit.

The inter prediction unit 160 may perform motion estimation using at least one or more reference pictures stored in the picture storage unit 180, and may determine a reference picture index and a motion vector indicating the reference picture.

In this case, a prediction block corresponding to a prediction unit to be encoded may be extracted from a reference picture used for motion estimation among a plurality of reference pictures stored in the picture storage unit 180 according to the determined reference picture index and the motion vector. have.

The intra predictor 150 may perform intra prediction encoding by using the reconstructed pixel value inside the picture in which the current prediction unit is included.

The intra prediction unit 150 may receive the current prediction unit to be predictively encoded, and perform intra prediction by selecting one of a preset number of intra prediction modes according to the size of the current block.

The intra predictor 150 adaptively filters the reference pixel to generate the intra prediction block, and generates reference pixels using the available reference pixels when the reference pixel is not available.

The entropy encoder 140 may entropy encode quantized coefficients quantized by the quantizer 130, intra prediction information received from the intra predictor 150, motion information received from the inter predictor 160, and the like. Can be.

FIG. 6 is a block diagram illustrating an example of a configuration for performing inter prediction in the encoding apparatus 10. The inter prediction encoder illustrated in FIG. 6 includes a motion information determiner 161 and a motion information encoding mode determiner 162. FIG. , Motion information encoder 163, prediction block generator 164, residual block generator 165, residual block encoder 166, and multiplexer 167.

Referring to FIG. 6, the motion information determiner 161 determines motion information of the current block, the motion information includes a reference picture index and a motion vector, and the reference picture index is any one of a previously coded and reconstructed picture. Can be represented.

When the current block is unidirectional inter prediction coded, it represents one of the reference pictures belonging to list 0 (L0), and when the current block is bidirectional predictively coded, it is a reference picture indicating one of the reference pictures of list 0 (L0). It may include an index and a reference picture index indicating one of the reference pictures of the list 1 (L1).

In addition, when the current block is bidirectional predictively coded, the current block may include an index indicating one or two pictures of reference pictures of the composite list LC generated by combining the list 0 and the list 1.

The motion vector indicates a position of a prediction block in a picture indicated by each reference picture index, and the motion vector may be in pixel units (integer units) or sub pixel units.

For example, the motion vector may have a resolution of 1/2, 1/4, 1/8 or 1/16 pixels, and if the motion vector is not an integer unit, the prediction block may be generated from pixels of an integer unit. Can be.

The motion information encoding mode determiner 162 may determine an encoding mode for the motion information of the current block as one of a skip mode, a merge mode, and an AMVP mode.

The skip mode is applied when there are skip candidates having the same motion information as the motion information of the current block and the residual signal is 0. The skip mode is that the current block, which is the prediction unit PU, has a size equal to that of the coding unit CU. Can be applied when

The merge mode is applied when there is a merge candidate having the same motion information as the motion information of the current block, and the merge mode includes a residual signal when the current block has a different size or the same size as the coding unit CU. Applies in the case. Meanwhile, the merge candidate and the skip candidate may be the same.

The AMVP mode is applied when the skip mode and the merge mode are not applied, and an AMVP candidate having a motion vector most similar to the motion vector of the current block may be selected as an AMVP predictor.

The motion information encoder 163 may encode motion information according to a method determined by the motion information encoding mode determiner 162.

For example, the motion information encoder 163 may perform a merge motion vector encoding process when the motion information encoding mode is a skip mode or a merge mode, and may perform an AMVP encoding process when the motion information encoding mode is an AMVP mode.

The prediction block generator 164 generates a prediction block by using the motion information of the current block. When the motion vector is an integer unit, the prediction block generator 164 copies the block corresponding to the position indicated by the motion vector in the picture indicated by the reference picture index, and then copies the current block. Generate a predictive block of.

Meanwhile, when the motion vector is not an integer unit, the prediction block generator 164 may generate pixels of the prediction block from integer unit pixels in a picture indicated by the reference picture index.

In this case, the prediction pixel may be generated using an 8-tap interpolation filter for the luminance pixel, and the prediction pixel may be generated using a 4-tap interpolation filter for the chrominance pixel.

The residual block generator 165 generates a residual block using the current block and the prediction block of the current block. When the size of the current block is 2Nx2N, the residual block generator 165 uses the prediction block having a size of 2Nx2N corresponding to the current block and the current block. You can create a block.

On the other hand, when the size of the current block used for prediction is 2NxN or Nx2N, after obtaining the prediction block for each of the two 2NxN blocks constituting the 2Nx2N, the last prediction block having a size of 2Nx2N using the two 2NxN prediction block is Can be generated.

In addition, a 2Nx2N sized residual block may be generated using the 2Nx2N sized prediction block, and overlap smoothing is applied to the pixels of the boundary part to eliminate discontinuity of the boundary parts of two prediction blocks having 2NxN size. Can be.

The residual block encoder 166 may divide the residual block into one or more transform units (TUs) so that each transform unit TU may be transform encoded, quantized, and entropy encoded.

The residual block encoder 166 may transform the residual block generated by the inter prediction method using an integer-based transform matrix, and the transform matrix may be an integer-based DCT matrix.

Meanwhile, the residual block encoder 166 uses a quantization matrix to quantize coefficients of the residual block transformed by the transform matrix, and the quantization matrix may be determined by a quantization parameter.

The quantization parameter is determined for each coding unit CU having a predetermined size or more, and when the current coding unit CU is smaller than the predetermined size, the first coding unit in the coding order among the coding units CU within the predetermined size ( Since only the quantization parameter of the CU) is encoded and the quantization parameter of the remaining coding unit CU is the same as the above parameter, it may not be encoded.

In addition, coefficients of the transform block may be quantized using a quantization matrix determined according to the quantization parameter and the prediction mode.

The quantization parameter determined for each coding unit CU having a predetermined size or more may be predictively encoded using the quantization parameter of the coding unit CU adjacent to the current coding unit CU.

A quantization parameter predictor of the current coding unit CU may be generated by searching in the order of the left coding unit CU and the upper coding unit CU of the current coding unit CU using one or two valid quantization parameters. have.

For example, the first valid quantization parameter found in the above order may be determined as a quantization parameter predictor, and the left first coding unit (CU) is searched in order of the coding unit immediately before the coding order to quantize the first valid quantization parameter. Can be determined by the parameter predictor.

The coefficients of the quantized transform block are scanned and converted into one-dimensional quantization coefficients, and the scanning scheme may be set differently according to the entropy encoding mode.

For example, inter prediction coded quantization coefficients may be scanned in one predetermined manner (zigzag or diagonal raster scan) when encoded by CABAC, and may be scanned differently from the above method when encoded by CAVLC. Can be.

For example, the scanning method may be determined according to zigzag in case of inter, the intra prediction mode in case of intra, and the coefficient scanning method may be determined differently according to the size of a transform unit.

Meanwhile, the scan pattern may vary according to the directional intra prediction mode, and the scanning order of the quantization coefficients may be scanned in the reverse direction.

The multiplexer 167 multiplexes the motion information encoded by the motion information encoder 163 and the residual signals encoded by the residual block encoder 166.

The motion information may vary according to an encoding mode. For example, in case of skip or merge, the motion information may include only an index indicating a predictor, and in the case of AMVP, the motion information may include a reference picture index, a differential motion vector, and an AMVP index of the current block. .

Hereinafter, an embodiment of the operation of the intra predictor 150 shown in FIG. 1 will be described in detail.

First, the intra prediction unit 150 receives the prediction mode information and the size of the prediction unit PU from the picture division unit 110, and stores the reference pixel in the picture storage unit to determine the intra prediction mode of the prediction unit PU. Read from 180.

The intra predictor 150 determines whether a reference pixel is generated by examining whether there is a reference pixel that is not available, and the reference pixels may be used to determine an intra prediction mode of the current block.

If the current block is located at the upper boundary of the current picture, pixels adjacent to the upper side of the current block are not defined. If the current block is located at the left boundary of the current picture, pixels adjacent to the left of the current block are not defined. It may be determined that the pixels are not available pixels.

In addition, even when the current block is located at the slice boundary and pixels adjacent to the upper or left side of the slice are not pixels that are first encoded and reconstructed, it may be determined that the pixels are not usable pixels.

As described above, when there are no pixels adjacent to the left side or the upper side of the current block or there are no pre-coded and reconstructed pixels, the intra prediction mode of the current block may be determined using only the available pixels.

Meanwhile, reference pixels at positions that are not available may be generated using the available reference pixels of the current block. For example, when the pixels of the upper block are not available, the upper pixels may be used using some or all of the left pixels. Can be generated and vice versa.

That is, the reference pixel is generated by copying the available reference pixel at the position closest to the predetermined direction from the reference pixel at the position not available, or when the reference pixel is not available in the predetermined direction, the closest in the opposite direction. The reference pixel can be generated by copying the available reference pixel at the location.

Meanwhile, even when the upper or left pixels of the current block exist, it may be determined as a reference pixel that is not available according to the encoding mode of the block to which the pixels belong.

For example, when a block to which a reference pixel adjacent to an upper side of the current block belongs is a block that is inter-coded and reconstructed, the pixels may be determined as not available pixels.

In this case, reference pixels usable may be generated using pixels belonging to a block in which a block adjacent to the current block is intra-encoded, and the encoding apparatus 10 may determine that the reference pixels are available according to an encoding mode. It transmits to the decoding apparatus 20.

The intra predictor 150 determines the intra prediction mode of the current block by using the reference pixels, and the number of intra prediction modes allowable in the current block may vary depending on the size of the block.

For example, if the size of the current block is 8x8, 16x16, 32x32, there may be 34 intra prediction modes. If the size of the current block is 4x4, there may be 17 intra prediction modes.

The 34 or 17 intra prediction modes may be configured of at least one non-directional mode (non-directional mode) and a plurality of directional modes.

One or more non-directional modes may be DC mode and / or planar mode. When the DC mode and the planner mode are included in the non-directional mode, there may be 35 intra prediction modes regardless of the size of the current block.

In this case, two non-directional modes (DC mode and planner mode) and 33 directional modes may be included.

In the planner mode, the prediction block of the current block is formed by using at least one pixel value (or a prediction value of the pixel value, hereinafter referred to as a first reference value) and reference pixels positioned at the bottom-right side of the current block. Is generated.

The configuration of an image decoding apparatus according to an embodiment of the present invention may be derived from the configuration of the image encoding apparatus 10 described with reference to FIGS. 1 to 6. For example, as described with reference to FIGS. 1 to 6. By performing the same processes of the same image encoding method in reverse, the image can be decoded.

7 is a block diagram illustrating a configuration of a video decoding apparatus according to an embodiment of the present invention. The decoding apparatus 20 includes an entropy decoding unit 210, an inverse quantization / inverse transform unit 220, an adder 270, The deblocking filter 250, the picture storage unit 260, the intra predictor 230, the motion compensation predictor 240, and the intra / inter switch 280 are provided.

The entropy decoder 210 receives and decodes a bit stream encoded by the image encoding apparatus 10, divides the bit stream into intra prediction mode indexes, motion information, quantization coefficient sequences, and the like, and decodes the decoded motion information into a motion compensation predictor ( 240).

In addition, the entropy decoder 210 may transfer the intra prediction mode index to the intra predictor 230 and the inverse quantizer / inverse transformer 220, and may deliver the inverse quantization coefficient sequence to the inverse quantizer / inverse transformer 220. .

The inverse quantization / inverse transform unit 220 converts the quantization coefficient sequence into inverse quantization coefficients of a two-dimensional array, and selects one of a plurality of scanning patterns for the transformation, for example, the prediction mode of the current block (ie, , Intra prediction or inter prediction), and a scanning pattern may be selected based on the intra prediction mode.

The inverse quantization / inverse transform unit 220 restores the quantization coefficients by applying a quantization matrix selected from a plurality of quantization matrices to the inverse quantization coefficients of the two-dimensional array.

Meanwhile, different quantization matrices are applied according to the size of the current block to be reconstructed, and a quantization matrix may be selected based on at least one of the prediction mode and the intra prediction mode of the current block for the same size block.

The inverse quantization / inverse transform unit 220 inversely transforms the reconstructed quantization coefficients to reconstruct the residual block, and the inverse transform process may be performed using a transform unit (TU) as a basic unit.

The adder 270 reconstructs the image block by adding the residual block reconstructed by the inverse quantization / inverse transform unit 220 and the prediction block generated by the intra predictor 230 or the motion compensation predictor 240.

The deblocking filter 250 may perform deblocking filter processing on the reconstructed image generated by the adder 270 to reduce deblocking artifacts due to image loss due to the quantization process.

The picture storage unit 260 is a frame memory for storing a local decoded image on which the deblocking filter process is performed by the deblocking filter 250.

The intra predictor 230 restores the intra prediction mode of the current block based on the intra prediction mode index received from the entropy decoder 210, and generates a prediction block according to the restored intra prediction mode.

The motion compensation predictor 240 generates a prediction block for the current block from the picture stored in the picture storage unit 260 based on the motion vector information, and applies the selected interpolation filter when a motion compensation with a small precision is applied. Can be generated.

The intra / inter switch 280 may provide the adder 270 with the prediction block generated by either the intra predictor 230 or the motion compensation predictor 240 based on the encoding mode.

FIG. 8 is a block diagram illustrating an example of a configuration of performing inter prediction in the image decoding apparatus 20. The inter prediction decoder includes a demultiplexer 241, a motion information encoding mode determiner 242, and a merge mode motion. An information decoder 243, an AMVP mode motion information decoder 244, a prediction block generator 245, a residual block decoder 246, and a reconstruction block generator 247 are included. Here, the merge mode motion information decoder 243 and the AMVP mode motion information decoder 244 may be included in the motion information decoder 248.

Referring to FIG. 8, the de-multiplexer 241 demultiplexes the currently encoded motion information and the encoded residual signals from the received bitstream, and transmits the demultiplexed motion information to the motion information encoding mode determiner 242. The demultiplexed residual signal may be transmitted to the residual block decoder 246.

The motion information encoding mode determiner 242 determines the motion information encoding mode of the current block. If the skip_flag of the received bitstream has a value of 1, the motion information encoding mode determiner 242 determines that the motion information encoding mode of the current block is encoded as the skip encoding mode. can do.

If the skip_flag of the received bitstream has a value of 0 and the motion information received from the de-multiplexer 241 has only a merge index, the motion information encoding mode determiner 242 determines the motion information encoding mode of the current block. It may be determined that is encoded in the merge mode.

In addition, the motion information encoding mode determiner 242 has a skip_flag of the received bitstream having a value of 0, and the motion information received from the demultiplexer 241 has a reference picture index, a differential motion vector, and an AMVP index. In this case, it may be determined that the motion information encoding mode of the current block is encoded in the AMVP mode.

The merge mode motion information decoder 243 is activated when the motion information encoding mode determiner 242 determines that the motion information encoding mode of the current block is a skip or merge mode, and the AMVP mode motion information decoder 244 moves. The information encoding mode determiner 242 may be activated when the motion information encoding mode of the current block is determined to be an AMVP mode.

The prediction block generator 245 generates the prediction block of the current block by using the motion information reconstructed by the merge mode motion information decoder 243 or the AMVP mode motion information decoder 244.

When the motion vector is an integer unit, the prediction block of the current block may be generated by copying a block corresponding to the position indicated by the motion vector in the picture indicated by the reference picture index.

On the other hand, when the motion vector is not an integer unit, pixels of the prediction block are generated from integer unit pixels in the picture indicated by the reference picture index. In this case, an interpolation filter of 8 taps is used for a luminance pixel and a chrominance pixel is used. Predictive pixels may be generated using a 4-tap interpolation filter.

The residual block decoder 246 entropy decodes the residual signal and inversely scans the entropy decoded coefficients to generate a two-dimensional quantized coefficient block, and the inverse scanning scheme may vary according to an entropy decoding scheme.

For example, when the CABAC-based decoding is performed, the reverse scanning method may be applied in a diagonal raster inverse scan manner and in the case of the CAVLC-based decoding in a zigzag inverse scanning manner. In addition, the inverse scanning scheme may be determined differently according to the size of the prediction block.

The residual block decoder 246 dequantizes the coefficient block generated as described above using an inverse quantization matrix, and reconstructs a quantization parameter to derive the quantization matrix. Here, the quantization step size may be reconstructed for each coding unit of a predetermined size or more.

The residual block decoder 260 inversely transforms the inverse quantized coefficient block to restore the residual block.

The reconstruction block generation unit 270 generates a reconstruction block by adding the prediction block generated by the prediction block generation unit 250 and the residual block generated by the residual block decoding unit 260.

Hereinafter, an embodiment of a process of reconstructing the current block through intra prediction will be described with reference to FIG. 7 again.

First, the intra prediction mode of the current block is decoded from the received bitstream, and for this purpose, the entropy decoder 210 may reconstruct the first intra prediction mode index of the current block by referring to one of the plurality of intra prediction mode tables. Can be.

As the table shared by the plurality of intra prediction mode tables encoding apparatus 10 and the decoding apparatus 20, any one table selected according to the distribution of intra prediction modes for a plurality of blocks adjacent to the current block may be applied. .

For example, if the intra prediction mode of the left block of the current block and the intra prediction mode of the upper block of the current block are the same, the first intra prediction mode index of the current block is restored by applying the first intra prediction mode table, and not the same. Otherwise, the second intra prediction mode table may be applied to restore the first intra prediction mode index of the current block.

As another example, when the intra prediction modes of the upper block and the left block of the current block are both the directional intra prediction mode, the direction of the intra prediction mode of the upper block and the direction of the intra prediction mode of the left block. If within this predetermined angle, the first intra prediction mode index is restored by applying the first intra prediction mode table, and if outside the predetermined angle, the first intra prediction mode index is applied by applying the second intra prediction mode table. You can also restore.

The entropy decoder 210 transmits the first intra prediction mode index of the reconstructed current block to the intra predictor 230.

The intra prediction unit 230 that receives the index of the first intra prediction mode may determine the maximum possible mode of the current block as the intra prediction mode of the current block when the index has the minimum value (ie, 0). .

Meanwhile, when the index has a value other than 0, the intra prediction unit 230 compares the index indicated by the maximum possible mode of the current block with the first intra prediction mode index, and as a result of the comparison, the first intra prediction mode. If the index is not smaller than the index indicated by the maximum possible mode of the current block, the intra prediction mode corresponding to the second intra prediction mode index obtained by adding 1 to the first intra prediction mode index is determined as the intra prediction mode of the current block. Otherwise, the intra prediction mode corresponding to the first intra prediction mode index may be determined as the intra prediction mode of the current block.

The intra prediction mode allowable for the current block may consist of at least one non-directional mode (non-directional mode) and a plurality of directional modes.

One or more non-directional modes may be DC mode and / or planar mode. In addition, either DC mode or planner mode may be adaptively included in the allowable intra prediction mode set.

To this end, information specifying the non-directional mode included in the allowable intra prediction mode set may be included in the picture header or the slice header.

Next, the intra predictor 230 reads reference pixels from the picture storage unit 260 to generate an intra prediction block, and determines whether there is a reference pixel that is not available.

The determination may be performed according to the presence or absence of reference pixels used to generate the intra prediction block by applying the decoded intra prediction mode of the current block.

Next, when it is necessary to generate the reference pixel, the intra predictor 230 may generate reference pixels at positions that are not available using the available reference pixels reconstructed in advance.

Definition of a reference pixel that is not available and a method of generating the reference pixel may be the same as the operation of the intra prediction unit 150 of FIG. 1, but generate an intra prediction block according to the decoded intra prediction mode of the current block. The reference pixels used to selectively recover may be selectively restored.

In addition, the intra prediction unit 230 determines whether to apply a filter to the reference pixels to generate the prediction block, that is, whether to apply filtering to the reference pixels to generate the intra prediction block of the current block. It may be determined based on the decoded intra prediction mode and the size of the current prediction block.

The problem of blocking artifacts is that the larger the block size is, the larger the block size can increase the number of prediction modes for filtering the reference pixels, but if the block is larger than the predetermined size can be seen as a flat area, the complexity is reduced The reference pixel may not be filtered for.

If it is determined that the filter is required to be applied to the reference pixel, the intra predictor 230 filters the reference pixels by using a filter.

At least two or more filters may be adaptively applied according to the degree of difference between the steps between the reference pixels. The filter coefficient of the filter is preferably symmetrical.

In addition, the above two filters may be adaptively applied according to the size of the current block. When the filter is applied, a narrow bandwidth filter is used for a small block, and a wide bandwidth filter is used for a large block. May be applied.

In the DC mode, since the prediction block is generated with the average value of the reference pixels, there is no need to apply a filter. In the vertical mode where the correlation is in the vertical direction, the filter does not need to be applied to the reference pixel, and the image is horizontal. It may not be necessary to apply a filter to the reference pixel even in a horizontal mode that is correlated in the direction.

As described above, whether filtering is applied is also related to the intra prediction mode of the current block, and thus, the reference pixel may be adaptively filtered based on the intra prediction mode of the current block and the size of the prediction block.

Next, the intra prediction unit 230 generates a prediction block using reference pixels or filtered reference pixels according to the reconstructed intra prediction mode, and the generation of the prediction block is the same as the operation of the encoding apparatus 10. As such, detailed description thereof will be omitted.

The intra prediction unit 230 determines whether to filter the generated prediction block, and the filtering may be determined by using information included in a slice header or a coding unit header or according to an intra prediction mode of the current block.

When determining that the generated prediction block is to be filtered, the intra predictor 230 may generate a new pixel by filtering pixels at a specific position of the generated prediction block by using available reference pixels adjacent to the current block. .

For example, in the DC mode, a prediction pixel in contact with reference pixels among the prediction pixels may be filtered using a reference pixel in contact with the prediction pixel.

Therefore, the prediction pixels are filtered using one or two reference pixels according to the positions of the prediction pixels, and the filtering of the prediction pixels in the DC mode may be applied to the prediction blocks of all sizes.

Meanwhile, in the vertical mode, prediction pixels in contact with the left reference pixel among the prediction pixels of the prediction block may be changed by using reference pixels other than the upper pixel used to generate the prediction block.

Similarly, in the horizontal mode, the prediction pixels in contact with the upper reference pixel among the generated prediction pixels may be changed using reference pixels other than the left pixel used to generate the prediction block.

In this manner, the current block may be reconstructed using the prediction block of the current block reconstructed and the residual block of the decoded current block.

FIG. 9 illustrates a second exemplary embodiment of a method of dividing and processing an image into blocks.

Referring to FIG. 9, a coding tree unit (CTU) having a maximum size of 256 × 256 pixels may be first divided into a quad tree structure and divided into four coding units (CUs) having a square shape.

Here, at least one of the coding units divided into the quad tree structure may be divided into a binary tree structure and re-divided into two coding units (CUs) having a rectangular shape.

Meanwhile, at least one of the coding units divided into the quad tree structure may be divided into a quad tree structure and re-divided into four coding units (CUs) having a square shape.

At least one of the coding units re-divided into the binary tree structure may be divided into two binary tree structures and divided into two coding units (CUs) having a square or rectangular shape.

Meanwhile, at least one of the coding units re-divided into the quad tree structure may be divided into a quad tree structure or a binary tree structure and divided into coding units (CUs) having a square or rectangular shape.

As described above, coding blocks (CBs) formed by dividing into a binary tree structure are no longer divided and may be used for prediction and transformation. That is, the sizes of the prediction unit PU and the transform unit TU belonging to the coding block CB as shown in FIG. 9 may be equal to the size of the coding block CB.

The coding unit split into the quad tree structure as described above may be split into one or two prediction units (PUs) using the method described with reference to FIGS. 3 and 4.

In addition, the coding unit divided into the quad tree structure as described above may be divided into one or more transform units (TUs) by using the method as described with reference to FIG. 5, and the divided transform units (TU) May have a maximum size of 64x64 pixels.

The syntax structure used to divide and process an image in block units may indicate the partition information using a flag. For example, whether to split the coding unit CU may be indicated using split_cu_flag, and the depth of the coding unit CU split using the binary tree may be indicated using binary_depth. In addition, whether the coding unit (CU) is divided into a binary tree structure may be represented by a separate binary_split_flag.

Blocks divided by the method as described with reference to FIG. 9 (eg, a coding unit (CU), a prediction unit (PU), and a transform unit (TU)) have been described with reference to FIGS. 1 to 8. By applying the same methods, encoding and decoding on an image may be performed.

Hereinafter, other embodiments of a method of dividing a coding unit (CU) into one or more transform units (TUs) will be described with reference to FIGS. 10 to 15.

According to an embodiment of the present invention, the coding unit CU may be divided into a binary tree structure and divided into transform units (TUs) which are basic units for transforming a residual block.

Referring to FIG. 10, at least one of rectangular coding blocks CB0 and CB1 divided into a binary tree structure and having a size of Nx2N or 2NxN is divided into a binary tree structure again, and has a square transform unit having a size of NxN. Can be divided into TU0 and TU1.

As described above, the block-based image encoding method may perform prediction, transform, quantization, and entropy encoding steps.

In the prediction step, a prediction signal may be generated by referring to a block currently performing encoding and an existing coded image or a neighboring image, and thus a difference signal between the current block and the current block may be calculated.

On the other hand, in the transforming step, the difference signal is input, and the transform is performed using various transform functions. The transformed signal is classified into DC coefficients and AC coefficients and is energy compacted to improve encoding efficiency. Can be.

In addition, in the quantization step, quantization may be performed by inputting transform coefficients, and then an image may be encoded by performing entropy encoding on the quantized signal.

On the other hand, the image decoding method is performed in the reverse order of the above encoding process, the image quality distortion may occur in the quantization step.

As a method for reducing image distortion while improving encoding efficiency, the size or shape of a transform unit (TU) and the type of transform function to be applied may be varied according to the distribution of the differential signal input to the input and the characteristics of the image in the conversion step. have.

For example, when a block-based motion estimation process is used to find a block similar to the current block, a difference is measured using a cost measurement method such as a sum of absolute difference (SAD) or mean square error (MSE). The signal distribution may occur in various forms according to the characteristics of the image.

Accordingly, effective encoding can be performed by selectively determining the size or shape of the transform unit CU based on the distribution of various differential signals to perform the transform.

For example, when a differential signal is generated in any coding block CBx, an effective transform may be performed by dividing the coding block CBx into a binary tree structure and dividing it into two transform units (TUs). . The DC value can generally be said to represent the average value of the input signal, so that when the differential signal is received at the input of the conversion process, the DC value can be effectively represented by dividing the coding block (CBx) into two transform units (TUs). have.

Referring to FIG. 11, a square coding unit CU0 having a size of 2N × 2N may be divided into a binary tree structure and divided into rectangular transform units TU0 and TU1 having a size of N × 2N or 2N × N.

According to another embodiment of the present invention, as described above, the step of dividing the coding unit (CU) into a binary tree structure may be repeated two or more times to divide the coding unit (CU) into a plurality of transform units (TUs).

Referring to FIG. 12, a rectangular coding block CB1 having a size of Nx2N is divided into a binary tree structure, and a block having a size of the divided NxN is further divided into a binary tree structure to N / 2xN or NxN / 2. After configuring a rectangular block having a size of N, the block having a size of N / 2xN or NxN / 2 is divided into a binary tree structure again to square transform units having a size of N / 2 × N / 2 (TU1, TU2) , TU4, TU5).

Referring to FIG. 13, a square coding block CB0 having a size of 2Nx2N is divided into a binary tree structure, and a block having a size of Nx2N is further divided into a binary tree structure to have a square having a size of NxN. After constructing the block, the block having the size of NxN may be further divided into a binary tree structure and divided into rectangular transform units TU1 and TU2 having the size of N / 2xN.

Referring to FIG. 14, a rectangular coding block CB0 having a size of 2N × N is divided into a binary tree structure, and a block having the size of divided NxN is further divided into a quad tree structure to have a size of N / 2 × N / 2. Square units may be divided into TU1, TU2, TU3, and TU4.

See FIGS. 1 through 8 for blocks divided by the method as described with reference to FIGS. 10 through 14 (eg, coding unit (CU), prediction unit (PU), and transform unit (TU)). By applying the methods as described above, encoding and decoding on an image may be performed.

Hereinafter, embodiments of a method of determining, by the encoding apparatus 10 according to the present invention, a block division structure will be described.

The picture division unit 110 included in the image encoding apparatus 10 performs rate distortion optimization (RDO) according to a preset order, and thus is capable of splitting a coding unit (CU), a prediction unit (PU), and a transform as described above. The partition structure of the unit TU may be determined.

For example, to determine the block division structure, the picture division unit 110 performs a rate distortion optimization-quantization (RDO-Q) while selecting an optimal block division structure in terms of bitrate and distortion. You can decide.

Referring to FIG. 15, when the coding unit CU has a form of 2Nx2N pixel size, the 2Nx2N pixel size shown in (a), the NxN pixel size shown in (b), and the Nx2N pixel size shown in (c) , RD may be performed in the order of transform unit (PU) partition structure of 2N × N pixel size shown in (d) to determine an optimal partition structure of the transform unit PU.

Referring to FIG. 16, when the coding unit CU has a form of Nx2N or 2NxN pixel size, the pixel size of Nx2N (or 2NxN) shown in (a), the pixel size of NxN shown in (b), N / 2xN (or NxN / 2) and NxN pixel sizes shown in (c), N / 2xN / 2, N / 2xN and NxN pixel sizes shown in (d), N shown in (e) An RDO may be performed in a transform unit (PU) partition structure order of a pixel size of 2 × N to determine an optimal partition structure of the transform unit PU.

In the above, the block division method of the present invention has been described with an example in which a block division structure is determined by performing RDO (Rate distortion Optimization). However, the picture division unit 110 may have a sum of absolute difference (SAD) or mean square error (MSE). By deciding the block division structure using), it is possible to reduce the complexity and maintain proper efficiency.

Hereinafter, an image processing method for providing merge mode processing and an encoding and decoding method according to the embodiment of the present invention will be described in detail.

The merge mode process merges the motion information of the current block and the adjacent block during inter prediction using the motion information, so that the coding unit according to the embodiment of the present invention or the coding unit according to an embodiment of the present invention performs quadtree and binary processing. As it is complexly divided into a tree structure, it may be processed inefficiently.

Accordingly, when the merge mode motion information decoder 243 constructs a merge mode motion information candidate list for selecting a merge candidate in merge mode motion information decoding, the merge mode motion information decoder 243 is derived in response to the current block. According to one or more merge mode candidates, the merge mode motion information candidate list for the prediction of the merge mode motion information of the current block is variably configured to provide an improvement in encoding efficiency according to adaptive processing on a complex partitioned block structure. Can be.

Variable and adaptive variables for constructing the merge mode motion information candidate list include temporal and spatial adjacency, each block size, partition type, motion information signaled by the encoder, candidate list configuration information signaled by the encoder, or context of a block. At least one of the information may be exemplified, and the merge mode motion information decoder 243 further provides a composite merge mode motion information candidate list construction process, a probability based update process, and a cumulative motion predictor candidate calculation process using the same. By doing so, the encoding and decoding efficiency can be improved.

17 is a block diagram illustrating in more detail the merge mode motion information decoder according to an embodiment of the present invention.

Referring to FIG. 17, the merge mode motion information decoder 243 according to an embodiment of the present invention may include a merge type classifier 2431, a temporal adjacent co-location merge mode candidate derivator 2432, and a spatial merge mode candidate derivator ( 2433, the corresponding block merge mode candidate derivator 2434, the composite merge mode candidate derivator 2435, the sub-block merge mode candidate derivator 2439, the merge candidate list constructer 2436, the merge candidate selector 2437, and the probability An information update unit 2438 is included.

However, the merge candidate is not limited to the candidates shown in FIG. 17, and the type of the merge candidate is calculated through equations using motion information of the adjacent blocks in addition to the block unit or sub-block unit candidates of the temporally and spatially adjacent blocks. Motion information is also referred to generically.

The merge mode motion information decoder 243 according to an embodiment of the present invention may adaptively derive the merge mode motion information for each of the subdivided types according to the complex partition structure, and the derived merge mode motion information may be a merge candidate. It may be inserted into the merge mode motion information candidate list configured in the list constructing unit 2436. In addition, the priority in the merge mode motion information candidate list according to each type may be determined.

To this end, the merge type classifier 2431 may determine a merge type corresponding to the current block and determine a merge mode derivation process according to the merge type.

The merge type information may be determined according to the merge mode encoding efficiency for the current block, and may be obtained from header information signaled corresponding to the current block during decoding.

To this end, the decoding apparatus 20 may parse the separately signaled index or flag corresponding to the merge mode and the type information. In addition, motion vector predictor (MVP) information and corresponding motion information (Motion information) corresponding thereto may be signaled to derive the merge mode motion information.

For example, the information signaled corresponding to the merge mode may indicate the position of one merge target block among spatial, temporal or complex adjacent merge target block candidates of the current block, or may indicate index information.

In addition, the motion information encoded in response to the merge mode may be coding block unit-based motion information of one merge target block among spatial, temporal or complex neighboring merge target block candidates of the current block, and the signaling information may be such a motion. It may also indicate index information of the information.

In more detail, the merge mode information signaled corresponding to the merge mode may be included in the header information. The header information may include a Sequence Parameter Set (SPS) header or a tile group header, and the header information may include size information of a merge mode list or a merge mode list according to a tile group type. It may indicate whether to perform intra prediction.

In more detail, the merge type classifier 2431 may determine, from the SPS header or the tile group header, merge mode information available for the current block, and for the determination, the merge mode signaled from the encoding apparatus 10. Information can be parsed.

Accordingly, the merge mode motion information decoder 243 may selectively perform at least one of a plurality of merge mode candidate configuration processes according to the type information I / B / P of the tile group or the slice.

For example, the merge mode motion information decoder 243 may selectively construct a merge mode candidate list for inter prediction prediction decoding corresponding to a coding block in a tile group of a B or P type. When the tile group to which the coding block to be decoded belongs is a B type, inter prediction may be performed through bidirectional prediction, and thus the merge mode motion information decoder 243 adds separate reference picture list selection information from the merge mode signaling information. By parsing, additional motion prediction candidate (MVP) index information to be added to the merge candidate list may be derived.

In addition, when the type information of the tile group to which the coding block to be decoded belongs is I type, the merge mode motion information decoder 243 may form a merge candidate list by decoding motion information encoded in the merge mode. The candidate list may be used for decoding using the intra prediction mode.

Here, the size of the merge mode candidate list may be specified for each merge mode type or tile group and may have different sizes. For example, the merge mode candidate list size may be derived through merge mode signaling information included in the tile group header. The merge mode signaling information may include maximum number of MVP candidate information, and the merge mode motion information decoder 243 may subtract a predetermined number corresponding to the tile group to which the current block belongs, The list size according to the merge mode of the tile group to which the block belongs may be determined in advance.

The encoding mode signaling information signaled in correspondence to the coding block unit (CU) may include at least one of skip mode flag information (CU_Skip_Flag), prediction mode information (Pred_mode_flag, Pred_mode_ibc_flag), and merge mode flag information (Merge_flag) of the block. Each flag information may indicate the type and configuration of the prediction mode.

For example, when the current block is encoded in the merge mode, the value of the merge mode flag (Merge_Flag) may be true, and the merge candidate list constructer 2436 may be configured according to the process of the induced derivative selected from the plurality of merge candidate derivators. Merge mode candidate list can be constructed.

The merge mode motion information decoder 243 decodes the merge mode candidate motion information selected from the finally configured merge mode candidate list and the additional motion information of the current block corresponding thereto to decode the merge mode motion information of the current block. You can decide.

In addition, the additional motion information may indicate differential motion information generated according to the prediction of the current block. For example, the additional motion information may include MVD (Motion Vector Difference) or Reference Index information, or these. It may include a value calculated by a combination of.

According to the merge mode candidate derivation process determined by the merge type classifier 2431, a derivation unit for deriving each merge mode candidate may selectively operate.

The merge type classifier 2431 determines a merge type for the current block by using signaling information for classifying the merge type of the current block. In this case, as an embodiment of the merge type of the current block, it may be classified into a spatial merge type and a temporal merge type. Alternatively, as another embodiment, the method may be classified into a block unit merge type and a sub-block unit merge type.

In this case, the sub-block refers to a block unit for dividing the current block into a plurality of blocks, and the sub-block unit merge type refers to a case in which a motion vector is merged and used in each sub-block unit. .

A temporal adjacent co-location merge mode candidate derivation unit 2432 derives the merge mode candidate from motion information of a block at the same position as that of the current block in a pre-decoded picture temporally adjacent to the current block.

The spatial merge mode candidate derivator 2433 derives the merge mode candidate from motion information of at least one spatially adjacent block based on a boundary line divided by the current block, the quad tree, and the binary tree structure.

The corresponding block merge mode candidate derivator 2434 derives the merge mode candidate from motion information of a block of a corresponding position determined corresponding to the position of the current block in a pre-decoded picture that is temporally adjacent to the current block.

The composite merge mode candidate derivator 2435 is based on a boundary line divided by the quad tree and the binary tree structure for a block having the same position as the current block in a pre-decoded picture temporally adjacent to the current block. The merge mode candidate is derived from spatially adjacent motion information of at least one block.

The sub-block merge mode candidate derivator 2439 derives the merge mode candidate from one or more sub-blocks constituting a block at the same position as the current block in a pre-decoded picture that is temporally adjacent to the current block.

The merge candidate list forming unit 2436 constructs a merge mode motion information candidate list by using one or more merge mode candidates derived according to the merge type classification.

Accordingly, the merge candidate list constructer 2436 uses the at least one candidate of a temporally adjacent co-location merge mode candidate, a spatial merge mode candidate, a corresponding block merge mode candidate, a composite merge mode candidate, and a sub block merge mode candidate. A merge mode motion information candidate list may be constructed.

Here, the merge mode motion information candidate list may include one or more arrays of motion information indexed by the merge mode motion information candidates according to a preset priority.

For example, the priority of the merge mode motion information candidate list construction may be determined according to the type information classification of the merge mode, and the merge candidate list construction unit 2436 may include one or more merge mode motion information classified according to the type information. The candidate list can be constructed. The type information classification may include spatial merge type information and temporal merge type information.

Accordingly, the merge mode motion information candidate list may be generated as each candidate list corresponding to each type information or as one merged mode motion information candidate list.

In addition, the merge candidate selector 2437 uses the current block using temporal and spatial adjacency, variable and adaptive variables such as each block size and partition type, and merge mode signaling information signaled from the encoding apparatus 10. May select a candidate for merge.

For example, the merge candidate selector 2437 may select a merge mode candidate using block size information of a block to be decoded.

Also, for example, the merge candidate selector 2437 may determine the merge type of the merge mode candidate to be selected by using the block size information of the block to be decoded. The merge candidate selector 2437 parses the sub-block merge flag information when the horizontal and vertical minimum sizes of the block to be decoded are equal to or greater than a predetermined size (for example, 8), thereby selecting the sub-block merge mode candidates from the current block. It may be decided whether to select as a merge candidate.

In addition, the merge mode motion information decoder 243 may combine the inter prediction mode and the intra prediction mode (CIIP) when the size of a block to be decoded is 64x64 or more and smaller than 128 x 128. ) May be further parsed as encoding information, select a merge candidate of the current block based on the parsed encoding information, and perform merge mode decoding.

For example, combined mode encoding may be performed to support inter-picture prediction and intra-picture prediction combining mode in response to a coding block to be currently decoded. In this case, the merge mode motion information decoder 243 determines whether the block division form of the block to be decoded is a vertically or horizontally divided coding block, and if the vertically or horizontally divided is, the merge mode information is determined from the intra prediction mode information. Acquire and perform merge mode decoding. A separate flag indicating merge mode decoding in this manner may be signaled, and the merge mode motion information decoder 243 may display a signal value of the signaled separate flag (CIIP mode flag, Ciip_flag), block division type, and block size information. Accordingly, merge mode decoding as described above may be performed. If the block size and the split form are different, motion information prediction decoding according to inter prediction may be separately performed.

The merge candidate selector 2437 may select a merge candidate corresponding to the current block from the merge mode motion information candidate list.

Accordingly, the prediction block generator 245 may generate the prediction block by performing motion information prediction decoding on the current block by using the merge candidate.

The probability information updater 2438 may update probability information for each candidate list of the merge mode motion information candidate list by using the selected merge candidate.

Here, the updated merge mode motion information candidate list may be used as a merge mode motion information candidate list of the next block merged by the current block and the merge mode.

In addition, the probability information updater 2438 accumulates and updates previously decoded motion information corresponding to slices in a picture, a tile group, or a tile in a tile group, as shown in FIG. 24, to be used for deriving motion information of the current block. can do.

More specifically, the previously decoded motion information may include a motion vector and the like, and may be used as one of motion vector candidates of the current block.

To this end, the probability information updater 2438 may construct a separate virtual motion candidate list (VMVCL). The VMVCL may be initialized to one of a row unit or a column unit of CTUs in the tile and the tile group according to the tile group, tile unit, or scan order in which the corresponding coding block is located.

In addition, the virtual motion candidate list has a characteristic of being updated as decoding is performed, and the number of MV Candidates included in the list may be variably changed. The virtual motion candidate list may be constructed according to a FIFO algorithm or the like, and the motion information of the virtual motion candidate list may be added to a candidate of the merge mode candidate list of the current block and used for motion prediction of the current block. .

18 is a flowchart illustrating a merge mode candidate derivation and candidate list construction method according to an embodiment of the present invention.

Referring to FIG. 18, the merge mode motion information decoder 243 according to an embodiment of the present invention first identifies merge type information in the merge type classifier 2431 (S101) and merges the merge type information. The mode candidate derivation process is determined (S103).

Accordingly, by the determined merge mode candidate derivation process, the derived unit of each of the induction units 2432, 2433, 2434, 2345, and 2436 obtains the merge mode motion information candidate (S105).

Here, the determination of the derivation process may be predetermined according to the merge type or sequentially determined according to the priority set in advance for the merge type. For example, the temporal adjacent co-merge mode candidate derivator 2432 may be prioritized to derive a merge mode candidate prior to the composite merge mode candidate derivator 2435.

The merge candidate list constructing unit 2436 constructs the one or more merge mode motion information candidate lists according to the type of the merge mode candidate and the priority in the list (S107).

19 to 20 are diagrams for explaining merge mode candidate derivation methods according to an embodiment of the present invention for each derivation process.

19 is a diagram for describing a method of determining a spatial merge mode candidate of the spatial merge mode candidate derivator 2433 corresponding to the current block.

Referring to FIG. 19, a block having the same position as the current block in a pre-decoded picture temporally adjacent to the current block corresponding to a current block is divided by the quad tree and the binary tree structure. The merge mode candidate may be derived from the motion information of at least one block that is spatially adjacent to the defined boundary line.

As such, motion information of blocks predetermined to be spatially adjacent may be derived as candidates for spatial merge mode, and as shown in FIG. 19A, basically, a left block 310, an upper block 311, The upper right block 312, the lower left block 313, and the upper left block 314 may be included.

In addition, the derivation range of the spatial merge mode candidate according to an embodiment of the present invention can be extended, and signaling information such as a separate index or a flag can be transmitted in the encoding apparatus 10.

That is, FIG. 19 (B) shows the extended merge mode motion information candidate of the current target block. The block of the AR position, together with the motion information of the block 311 at the A position among the blocks located at the top of the current target block The motion information of 316 may be extended and used as a spatial merge mode motion information candidate.

Such an extended case may be applied when the upper block of the current target block is further divided into a plurality of blocks, and thus the A block 311 and the AR block 316 have different motion information. For example, the A block or the AR block may be a block finally divided into binary, so that the motion information may be different by a predetermined reference or more.

However, when the upper block of the current target block is not divided into a plurality of blocks, the motion information of the expanded AR block is spatial when the motion information of the A block 311 and the AR block 316 is compared and the same. It may be excluded from the merge mode motion information candidate or may not be added to the merge mode motion information candidate list.

Meanwhile, as shown in FIG. 19B, in addition to the motion information of the block 310 at the L position among the blocks located on the left of the current target block, the motion information of the block 318 at the BL location is also expanded. It may be used as a merge mode motion information candidate.

This extended case may be applied to a case in which the left block of the current target block is further divided into a plurality of blocks, and the L block 310 and the BL block 318 have different motion information. For example, the L block or the BL block may be a block finally divided into binaries, and motion information may be different from each other by a predetermined reference or more.

However, when the left block of the current target block is not divided into a plurality of blocks, the motion information of the extended BL block is excluded from the spatial merge mode motion information candidate when the motion information of the L and BL positions is the same. It may not be added to the merge mode motion information candidate list.

As such, when the merge candidate list constructer 2436 constructs a list by deriving a motion candidate from the neighboring block, the order may be set in advance.

For example, when the merge candidate list generator 2436 may obtain motion information from the neighboring block, the merge mode candidate list of the current block in the order of the left block, the upper block, the upper right block, the lower left block, and the upper left block. Can be configured.

In addition, after the merge candidate list construction unit 2436 performs motion information derivation on a spatially neighboring block, the merge candidate list constructer 2436 additionally derives a temporal co-location motion candidate according to the presence or absence of duplicate motion information in the list to merge the current mode. It is also possible to construct a candidate list.

20 is a diagram illustrating a temporal and complex merge mode candidate derivation process according to an exemplary embodiment of the present invention.

20 (A) illustrates a method of deriving a merge mode candidate of the temporal adjacent co-merge mode candidate derivation unit 2432, and according to an embodiment of the present invention, the temporal adjacent co-merge mode candidate derivation unit 2432 is a current target block. One temporal motion vector derived from temporal collocated merge candidates may be obtained as the merge mode motion information candidate.

For example, the temporal adjacent co-location merge mode candidate inducing unit 2432 may identify a block at the same position as the current target block in a reference picture corresponding to the current picture, and is derived from the co-located block. One temporal motion vector may be derived as the merge mode motion information candidate.

As an example according to an embodiment of the present disclosure, when the characteristic of the tile group is B (two-way), the temporal adjacent co-merge mode candidate derivation unit 2432 corresponds to a block for performing current decoding, and thus includes two reference picture lists. One reference picture list of 0 and reference picture list 1 may be selected to derive motion information of a collocated block, and the selection information of the reference picture list may be arbitrary flag information (Collcated_from_ref_list_L0_flag). It may be signaled separately.

The temporal adjacent co-merge mode candidate derivator 2432 derives a temporal neighbor candidate from the reference picture list 0 when the signaled value is true. On the contrary, the temporal adjacent co-merge mode candidate derivator 2432 may derive a temporal prediction candidate (Temporal MVP) from reference picture list 1 when the signaled value is false.

For example, in FIG. 20A, a merge mode motion information candidate may be derived from a motion vector of a lower right block 421 of a colocated block of a temporal adjacent picture.

Meanwhile, if the motion vector is not derived in the lower right block 421, the temporal adjacent co-merge mode candidate derivator 2432 may derive the merge mode motion information candidate from the motion vector of the center position 422. have.

Meanwhile, FIG. 20B illustrates a merge mode candidate derivation of the corresponding block merge mode candidate derivation unit 2434 according to an embodiment of the present invention, and the corresponding block merge mode candidate derivation unit 2434 is based on the current target block. The merge mode motion information candidate may be derived from the motion information of the block of the corresponding position determined corresponding to the position of the current block of the adjacent reference picture, which may be referred to as an enhanced temporal merge mode motion candidate predictor (ATMVP). In this case, the corresponding location may be derived using motion information derived from spatially adjacent blocks of the current block according to a predefined rule.

For example, as shown in FIG. 20B, a plurality of motion vectors in units of sub-blocks derived from corresponding position blocks may exist, and according to an embodiment of the present invention, corresponding blocks The merge mode candidate derivator 2434 may determine a temporal neighboring motion vector TEMPORAL VECTOR derived from the target block as the corresponding block merge mode candidate.

Here, in the merge mode candidate list structure of the current target block, the merge mode candidate list constructing unit 2436 may firstly insert candidates obtained from the corresponding block merge mode candidate deriving unit 2434 into the list.

20 (C) shows at least one of spatially adjacent motion information candidates corresponding to the same position block in a time-adjacent picture of the current target block, based on a boundary line divided by the quad tree and the binary tree structure. The composite merge mode candidate derivation process of deriving the merge mode candidate from the motion information of the block is shown.

As shown in FIG. 20C, the composite merge mode candidate derivator 2435 may select the merge mode candidate from a motion vector or a sub-block unit motion vector derived from a spatially adjacent block based on a position of a current block among blocks of a temporally adjacent picture. Can be induced. This may be referred to as a spatiotemporal merge mode motion information candidate (STMVP).

In addition, the composite merge mode candidate derivator 2435 is configured to derive motion information of the first sub-block 440 of the current block, in which the left sub-block 442 and the upper sub-block 441 are spatially adjacent to each other. The merge candidates may be derived according to a complex operation using the motion vector average value or the median value of the sub-block 443 of the co-located block.

However, when there is a non-referenced subblock among the spatially and temporally adjacent subblocks, an average value or a median value of motion vectors of one or more remaining subblocks except for the corresponding subblock may be derived as the composite merge candidate. It may be.

Meanwhile, the merge mode motion information decoder 243 may generate new motion information by combining the merge mode differential motion information encoded by the encoding apparatus 10 with the merge candidate, and the generated motion information may be added to the merge mode list. Can be added. Here, the merge mode differential motion information may include additional distance information and direction information corresponding to the merge candidate.

For example, the merge mode motion information decoder 243 may generate new motion information using motion information of one of the merge mode candidate lists and prediction parameter information such as decoded MVD information.

In more detail, the prediction parameter information may be offset information including distance information and direction information. The merge mode motion information decoder 243 may generate new motion information by combining it with the previously selected merge candidate and add it to the merge candidate list.

The distance information may be indicated by a preset table, and index information corresponding thereto may indicate the distance information.

For example, the newly generated motion information is called merge motion information as Merge_MV, and the merge mode list is MergeCandidateList [i] (i = 0, 쪋, Maximum number of mergecandidate), and the distance information of the offset information is Merge_MVD_Dis, When direction information is represented by Merge_MVD_Dir,

TempMV = MergeCandidateList [0]; or

TempMV = MergeCandidateList [1];

Merge_MV_Cand = TempMV + (Merge_MVD_Dis * Merge_MVD_Dir);

Next, new motion information (Merge_MVCand) using one of the motion information derived from the merge candidate list may be generated.

In addition, signaling information indicating that additional new motion information should be generated using motion information of the previously configured merge candidate list may be signaled from the encoding apparatus 10, and may be indicated by a flag such as Merge_MVD_Flag. . When Merge_MVD_Flag is true, Offset parameters Merge_MVD_Distance_Index and Merge_MVD_Dir_Index may be indicated, and their values may be indicated by separate table indexes.

In addition, the merge mode motion information decoder 243 may derive the actual difference signal value using the Merge_MVD_Distance_Index and the scaling information derived from the header information. Merge_MVD_Dir_index may be direction information and may include four direction information.

Meanwhile, the merge candidate list construction unit 2436 may calculate average merge candidate motion information to determine one of candidates added to the merge candidate list.

For example, in constructing a merge candidate list of a coding unit to be currently decoded, the merge candidate list constructing unit 2436 performs a merge mode candidate derivation process, but the current merge candidate list size is the maximum size of the merge candidate list ( If less than MaxNumMergeCand), the average merge candidate motion information may be generated and inserted into the list.

To this end, the merge candidate list constructer 2436 may calculate the average motion vector value AVG_MV using one or more motion information obtained from the previously configured merge candidate list.

MV_Cand0 = MergeCandidateList [0];

MV_Cand1 = MergeCandidateList [1];

Merge_AVG_MV = (MV_Cand0 + MV_Cand1) / 2;

Accordingly, the merge candidate list construction unit 2436 may conditionally insert the Merge_AVG_MV generated by the above operation into the merge candidate list as follows.

MergeCandidateList [i] = Merge_AVG_MV; (1 <i <Maximum Number of Merge Candidate)

Meanwhile, as shown in FIG. 20D, the sub-block merge mode candidate inducing unit 2435 may include one or more sub-blocks constituting a block at the same position as the current block in the pre-decoded picture that is temporally adjacent to the existing current block. The merge mode candidate may be derived from.

Referring to FIG. 20 (D), there may be a plurality of motion vectors in units of sub-blocks derived from co-located blocks of temporally adjacent reference pictures of the current target block, and the sub-block merge mode candidate derivation unit 2435 ) May obtain a merge mode motion information candidate with reference to the subblock.

21 is a flowchart illustrating a method of constructing and updating a merge mode motion information candidate list according to an embodiment of the present invention, and FIGS. 22 to 24 are views illustrating a configuration and configuration of a merge mode motion information candidate list according to an embodiment of the present invention. Figures are schematic diagrams for explaining the update process.

Referring to FIG. 21, when a merge candidate is selected from the merge candidate list in the merge candidate selector 2437 (S201), the motion compensation predictor 240 may perform merge mode decoding of motion information using the selected merge candidate. It may be (S203).

Thereafter, the probability information updater 2438 updates the probability information of the merge candidates corresponding to the selected merge candidate index (S205), and the merge candidate list construction unit 2436 orders the merge candidate list according to the updated probability information. Change processing (S207).

22 illustrates an example of a merge mode motion information candidate list constructed according to an embodiment of the present invention.

In constructing the merge mode motion information candidate list of the current block, the merge candidate list construction unit 2436 may add the merge candidates derived to the candidate list according to the predefined configuration order as described above. Each added merge candidate may correspond to derived type information or location information. For example, derived spatial information (L, AL, AR, BL, BR, etc.) may be illustrated in the case of a spatial merge candidate, and type information such as ATMVP, STMVP, etc. may be illustrated in the case of other merge candidates as described above. Such location and type information may be considered in determining the priority when constructing each candidate list.

In addition, the candidate list may be configured as a single list integrated as shown in FIG. 22A, and as shown in FIG. 22B, a plurality of lists separated by each type or position (list0, list1, ...) It may be configured as.

In the encoding apparatus 10, when referring to motion information in a block spatially adjacent to the current block, one of a plurality of reference picture lists may be selected according to the characteristic information of the tile group, and the reference picture list selection information may be selected from a decoding apparatus ( 20).

The decoding apparatus 20 may derive temporally adjacent ATMVP (motion information induced by Temporal Motion Vector prediction) from the same block using the received reference picture list selection information. ATMVP may be added to the merge candidate list.

Accordingly, when the merge candidate list forming unit 2436 constructs a plurality of merge lists, the ATMVP can be added to the 0th index of the second merge mode list list 1 as shown in FIG. 22 (B). The reason for the addition to the 0 th index is that the temporally adjacent ATMVPs in the collocated block are most similar and thus one of the candidates having the highest coding efficiency when constructing the second merge mode list.

Meanwhile, as an example of priority, the motion information of the block 310 adjacent to the spatial left side of the current block 300 may be inserted as a candidate first, and then the block adjacent to the spatial top of the current block 300 ( The motion information of 311) may be added as the second candidate. Subsequently, after adding candidates of the two spatially adjacent spatial merge mode candidate derivators 2433 to the list, ATMVP candidates of the temporal adjacent co-merge mode candidate derivator 2432 may be added, and then the composite merge mode candidates are added. After the STMVP candidates of the induction unit 2435 are added, the subblock unit candidates of the extended subblock merge mode candidate induction unit 2435 are added, or the spatially adjacent candidates or the corresponding blocks of the spatial merge mode candidate derivation unit 2433 are added. Temporal adjacent candidates of the merge mode candidate derivator 2434 may be added.

Meanwhile, in adding the candidates to the list, the candidate information may be added only when the motion information of each reference block can be referred to. The case where the motion information of each reference block is referenceable may mean a case in which the reference block is coded in the inter prediction mode.

23 illustrates a probability information table and list update according to an embodiment of the present invention.

In constructing the merge mode motion information candidate list of the current block, the encoding apparatus 10 or the decoding apparatus 20 according to an embodiment of the present invention initializes the candidate list as shown in FIG. By performing probability update as is performed, the order of list construction can be adaptively changed.

For example, as shown in FIG. 23B, the candidate list constructing unit 2436 includes respective initial probability values a, b, c, d, e, f,... Corresponding to merge candidates added to the list. ..., n) may be assigned, and the candidate list constructing unit 2436 may form the list sequentially by sorting in descending order according to the probability table.

23 (C), according to the probability information update (a ', b', c ', d', e ', f', ..., n ') of the probability information updater 2438, The candidate list constructer 2436 may rearrange the order of the list.

The probability value may be a value of a specific integer form or may be determined according to a selection rate or a generation rate of the target merge candidate.

In addition, the initial probability value may use a predefined value or may be transmitted from the encoding apparatus 10 to the decoding apparatus 20. The probability information table may be implemented in various implementations.

When the merge candidate candidate block of the current block is selected, the probability information updater 2438 may update the probability value and the table by using the result value.

Based on the probability value changed according to the result of performing the update, the merge mode candidate list constructer 2436 may be again sorted in descending order.

According to another embodiment of the present invention, the probability information table may be configured as a plurality of individual tables according to the encoding information of the current target block, and the encoding information may include a block size, a partition type, a block depth, and the like. It may include at least one of the information.

Meanwhile, FIG. 24 illustrates a merge mode candidate list initial configuration and a sequential update method according to an embodiment of the present invention. Referring to FIG. 24, the size of a coding tree block (CTB) may be extended to 256 × 256, a merge mode candidate list for the extended coding tree block is configured, and probability updates for each coding block in the coding tree block are updated. And update processing can be performed.

To this end, the merge mode motion candidate list may be derived from a predetermined neighboring block corresponding to the initial block.

Then, whenever inter-prediction decoding is performed on sub-coded blocks divided from coding tree blocks, a list is constructed according to probability information for merge candidates determined by the above-described derivation processes from temporal and spatial neighboring blocks. The order can be updated.

Accordingly, after the initial block, the merge candidate selector 2437, which processes the block N updated in stages, uses the merge mode motion candidate list in which the N-order probability is accumulated from the temporal spatially adjacent or complex derived neighboring blocks, thereby encoding efficiency. This optimized merge motion information can be selected.

The cumulative decoding probability information according to the merge mode selection information may be sequentially accumulated according to a raster scan order, and then, from the current block N, the next N + 1 block, N + 2 block, .. And further sequentially update to the last block (Block End) in the current coding tree block, so that merge mode motion information coding efficiency and accuracy of subsequent processed blocks can be further increased.

25 to 27 are diagrams for explaining high level syntax of header information according to an embodiment of the present invention.

Collocated_from_l0_flag indicates whether the temporal merge mode motion candidate predictor (ATMVP) is derived. If the variable value is 1, ATMVP can be derived from reference picture list 0, and if the value is 0, ATMVP can be derived from reference picture list 1. have.

In addition, Merge_MVD_Flag described above may indicate that additional new motion information should be generated using motion information of a previously configured merge candidate list, which may be indicated as a tile_group_mmvd_enabled_flag corresponding to a tile group in FIG. 25. In addition, distance and direction corresponding to each mmvd_flag may be indicated to merge data, respectively.

The merge mode signaling information may include maximum number of MVP candidate information and may be indicated by a flag such as Max_num_merge_cand, which may be allocated to each subblock. For example, the merge mode motion information decoder 243 may predetermine the list size according to the merge mode of the tile group to which the current block belongs by subtracting a predetermined number corresponding to the tile group to which the current block belongs.

In addition, when a merge mode list is constructed by deriving motion information from a subblock, the merge mode motion information decoder 243 may determine whether to encode merge mode motion information of each subblock by parsing separate flag information (merge_subblock_flag). have.

Alternatively, when the merge mode motion information decoder 243 is encoded in the CIIP mode as described above, the merge mode motion information decoder 243 may determine whether the merge mode is encoded in the CIIP mode by using a separate CIIP mode flag (Ciip_flag). In the case of the CIIP mode, intra or inter prediction may be determined according to the presence or absence of a Most provable mode (MPM) flag, and merge mode may be applied to inter prediction.

Meanwhile, inter prediction may be performed by a triangle. In this case, the merge mode list may be generated, and merge mode information according to the triangular prediction mode may be signaled through merge_triangle_flag, merge_triangle_split_dir (division direction information), and index information.

The method according to the present invention described above may be stored in a computer-readable recording medium that is produced as a program for execution on a computer, and examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape , Floppy disks, optical data storage devices, and the like, and also include those implemented in the form of carrier waves (eg, transmission over the Internet).

The computer readable recording medium can be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion. In addition, functional programs, codes, and code segments for implementing the method can be easily inferred by programmers in the art to which the present invention belongs.

In addition, although the preferred embodiment of the present invention has been shown and described above, the present invention is not limited to the specific embodiments described above, but the technical field to which the invention belongs without departing from the spirit of the invention claimed in the claims. Of course, various modifications can be made by those skilled in the art, and these modifications should not be individually understood from the technical spirit or prospect of the present invention.

Claims (13)

In the image processing method, A picture of an image is divided into a plurality of coding units, which are basic units on which inter prediction or intra prediction is performed, and a quad tree of the picture or the split coding unit. Determining a current block for decoding the coding unit divided into a binary tree structure; Variably constructing a merge mode motion information candidate list for predicting merge mode motion information of the current block according to one or more merge mode candidates derived corresponding to the current block; And Performing predictive decoding of merge mode motion information of the current block by using the merge mode motion information candidate list. Image processing method. The method of claim 1, The configuring step, Deriving the merge mode candidate from motion information of a block at a position identical to a position of the current block in a pre-decoded picture temporally adjacent to the current block; Image processing method. The method of claim 1, The configuring step, Deriving the merge mode candidate from motion information of a block of a corresponding position determined corresponding to the position of the current block in a pre-decoded picture temporally adjacent to the current block; Image processing method. The method of claim 1, The configuring step, Deriving the merge mode candidate from motion information of at least one spatially adjacent block based on the boundary line divided by the current block, the quad tree, and the binary tree structure; Image processing method. The method of claim 1, The configuring step, Movement of at least one block spatially adjacent to a block divided by the quad tree and the binary tree structure with respect to a block at the same position as the current block in a pre-decoded picture temporally adjacent to the current block Deriving the merge mode candidate from information. Image processing method. The method of claim 1, The configuring step, Deriving the merge mode candidate from at least one sub-block constituting a block at the same position as the position of the current block in a pre-decoded picture temporally adjacent to the current block; Image processing method. The method of claim 1, The configuring step, Constructing the merge mode motion information candidate list using at least one candidate of a temporally adjacent co-location merge mode candidate, a spatial merge mode candidate, a corresponding block merge mode candidate, a compound merge mode candidate, and a sub-block merge mode candidate. doing Image processing method. The method of claim 1, The performing of the motion information prediction decoding may include: Selecting a merge candidate corresponding to the current block from the merge mode motion information candidate list; And Performing motion information prediction decoding of the current block by using the merge candidate; Updating probability information for each candidate list of the merge mode motion information candidate list by using the selected merge candidate; Image processing method. The method of claim 1, The updated merge mode motion information candidate list is Used as a merge mode motion information candidate list of the next block merged by the current block and the merge mode Image processing method. The method of claim 1, Variably constructing the merge mode motion information candidate list, Identifying the type information of the merge mode; And Constructing one or more merge mode motion information candidate lists classified according to the type information. Image processing method. The method of claim 10, The type information includes spatial merge type information and temporal merge type information. Image processing method. In the video decoding method, Receiving an encoded bitstream; Performing inverse quantization and inverse transformation on the input bitstream to obtain a residual block; Determining a current block for decoding a coding unit having a picture of an image divided into quad tree and binary tree structures corresponding to the residual block; Variably constructing a merge mode motion information candidate list for predicting merge mode motion information of the current block according to one or more merge mode candidates derived corresponding to the current block; Obtaining a prediction block by performing merge mode motion information prediction decoding of the current block by using the merge mode motion information candidate list; And Reconstructing an image using the prediction block and the residual block. Image Decoding Method. In the video encoding method, Dividing a picture of an image into a plurality of coding units divided into quad tree and binary tree structures; Determining a current block for encoding the coding unit; Variably constructing a merge mode motion information candidate list for predicting merge mode motion information of the current block according to one or more merge mode candidates derived corresponding to the current block; And Performing prediction mode decoding of merge mode motion information of the current block using the merge mode motion information candidate list to obtain a prediction block; Image coding method.

Images