KR102449210B1 - Method for intra prediction and apparatus thereof - Google Patents

Abstract

In the intra prediction method according to the present invention, based on a pixel in a current luma block and a neighboring luma pixel located around the current luma block, luma reference information including a pixel in a luma reference block and a neighboring luma reference pixel located around the luma reference block generating, based on a first neighboring chroma reference pixel corresponding to a neighboring luma reference pixel among neighboring luma reference pixels and neighboring chroma pixels located around the current chroma block, deriving a prediction parameter for the current chroma block and luma reference and deriving a prediction value of the pixel in the current chroma block based on the pixel in the block and the prediction parameter.

Description

Intra prediction method and device therefor

The present invention relates to image processing, and more particularly, to an intra prediction method and apparatus.

Recently, as broadcasting services with high definition (HD) resolution are expanding not only domestically but also worldwide, many users are getting used to high-resolution and high-definition images, and accordingly, many institutions are spurring the development of next-generation imaging devices. In addition, as interest in UHD (Ultra High Definition) having a resolution four times higher than that of HDTV increases along with HDTV, a compression technology for higher resolution and high-definition images is required.

For image compression, inter prediction technology that predicts pixel values included in the current picture from temporally previous and/or later pictures, predicting pixel values included in the current picture using pixel information in the current picture An intra prediction technique, an entropy encoding technique in which a short code is assigned to a symbol having a high frequency of occurrence and a long code is assigned to a symbol having a low frequency of occurrence, may be used.

An object of the present invention is to provide an image encoding method and apparatus capable of improving image encoding/decoding efficiency.

Another technical object of the present invention is to provide an image decoding method and apparatus capable of improving image encoding/decoding efficiency.

Another technical object of the present invention is to provide an intra prediction method and apparatus capable of improving image encoding/decoding efficiency.

One embodiment of the present invention is an intra prediction method. The method includes generating luma reference information including a pixel in a luma reference block and a neighboring luma reference pixel located in the vicinity of the luma reference block, based on a pixel in the current luma block and a neighboring luma pixel located around the current luma block , deriving a prediction parameter for the current chroma block based on a first neighboring chroma pixel corresponding to the neighboring luma reference pixel among the neighboring luma reference pixel and neighboring chroma pixels located around the current chroma block, and the luma The method may include deriving a prediction value of a pixel in the current chroma block based on the pixel in the reference block and the prediction parameter. Here, in the generating of the luma reference information, it is determined whether the neighboring luma pixel is available according to whether a second neighboring chroma pixel corresponding to the neighboring luma pixel is available from among the neighboring chroma pixels located around the current chroma block. and a pixel value of the neighboring luma pixel may be determined according to whether the neighboring luma pixel is available.

In the generating of the luma reference information, when the second neighboring chroma pixels are available, it may be determined that the neighboring luma pixels are available.

When the position of the second neighboring chroma pixel is [x, y], the x is -1 and the y has one of 0 to N _U *2-1, based on the second neighboring chroma pixel A location of the neighboring luma pixel, at which availability is determined, may be [x, 2y] or [x, 2y+1], and N _U may indicate a horizontal length and a vertical length of the current chroma block.

When the position of the second neighboring chroma pixel is [x, y], the y is -1, and the x has one of 0 to N _U *2-1, based on the second neighboring chroma pixel A location of the neighboring luma pixel, at which availability is determined, may be [2x, y] or [2x+1, y], and N _U may indicate a horizontal length and a vertical length of the current chroma block.

When the position of the second neighboring chroma pixel is [-1, -1], the position of the neighboring luma pixel at which availability is determined based on the second neighboring chroma pixel may be [-1, -1]. have.

In the generating of the luma reference information, when it is determined that the neighboring luma pixel is not available, a pixel value of an available pixel among pixels located around the current luma block may be allocated to the neighboring luma pixel.

In the generating of the luma reference information, when it is determined that the neighboring luma pixel is a pixel on a left pixel line located adjacent to the left side of the current luma block and it is determined that the neighboring luma pixel is available, it is located adjacent to the left side of the neighboring luma pixel A pixel value of the luma pixel may be determined as a pixel value of the neighboring luma pixel.

In the generating of the luma reference information, when it is determined that the neighboring luma pixel is a pixel on an upper pixel line located adjacent to the upper end of the current luma block and it is determined that the neighboring luma pixel is available, a luma pixel located at the same position as the neighboring luma pixel A pixel value of may be determined as a pixel value of the neighboring luma pixel.

Another embodiment of the present invention is a video decoding method. The method includes generating luma reference information including a pixel in a luma reference block and a neighboring luma reference pixel located in the vicinity of the luma reference block, based on a pixel in the current luma block and a neighboring luma pixel located around the current luma block , deriving a prediction parameter for the current chroma block based on a first neighboring chroma pixel corresponding to the neighboring luma reference pixel among the neighboring luma reference pixel and neighboring chroma pixels located around the current chroma block; generating a prediction block for the current chroma block by deriving a prediction value of a pixel in the current chroma block based on the pixel in the reference block and the prediction parameter, and for the current chroma block based on the prediction block It may include generating a reconstruction block. Here, in the generating of the luma reference information, it is determined whether the neighboring luma pixel is available according to whether a second neighboring chroma pixel corresponding to the neighboring luma pixel is available from among the neighboring chroma pixels located around the current chroma block. and a pixel value of the neighboring luma pixel may be determined according to whether the neighboring luma pixel is available.

In the generating of the luma reference information, when the second neighboring chroma pixels are available, it may be determined that the neighboring luma pixels are available.

When the position of the second neighboring chroma pixel is [x, y], the x is -1 and the y has one of 0 to N _U *2-1, based on the second neighboring chroma pixel A location of the neighboring luma pixel, at which availability is determined, may be [x, 2y] or [x, 2y+1], and N _U may indicate a horizontal length and a vertical length of the current chroma block.

When the position of the second neighboring chroma pixel is [x, y], the y is -1, and the x has one of 0 to N _U *2-1, based on the second neighboring chroma pixel A location of the neighboring luma pixel, at which availability is determined, may be [2x, y] or [2x+1, y], and N _U may indicate a horizontal length and a vertical length of the current chroma block.

When the position of the second neighboring chroma pixel is [-1, -1], the position of the neighboring luma pixel at which availability is determined based on the second neighboring chroma pixel may be [-1, -1]. have.

In the generating of the luma reference information, when it is determined that the neighboring luma pixel is not available, a pixel value of an available pixel among pixels located around the current luma block may be allocated to the neighboring luma pixel.

In the generating of the luma reference information, when it is determined that the neighboring luma pixel is a pixel on a left pixel line located adjacent to the left side of the current luma block and it is determined that the neighboring luma pixel is available, it is located adjacent to the left side of the neighboring luma pixel A pixel value of the luma pixel may be determined as a pixel value of the neighboring luma pixel.

In the generating of the luma reference information, when it is determined that the neighboring luma pixel is a pixel on an upper pixel line located adjacent to the upper end of the current luma block and it is determined that the neighboring luma pixel is available, a luma pixel located at the same position as the neighboring luma pixel A pixel value of may be determined as a pixel value of the neighboring luma pixel.

According to the image encoding method according to the present invention, image encoding/decoding efficiency can be improved.

According to the image decoding method according to the present invention, image encoding/decoding efficiency can be improved.

According to the intra prediction method according to the present invention, image encoding/decoding efficiency can be improved.

1 is a block diagram illustrating a configuration of an image encoding apparatus to which the present invention is applied according to an embodiment.
2 is a block diagram illustrating a configuration of an image decoding apparatus to which the present invention is applied according to an embodiment.
3 is a conceptual diagram schematically illustrating an embodiment in which one unit is divided into a plurality of sub-units.
4 shows an example of a prediction direction of an intra prediction mode and a mode value assigned to each prediction direction.
5 is a flowchart schematically illustrating an embodiment of an inter-color component prediction method for performing prediction on a chroma component based on a luma component.
6 illustrates an embodiment of a method of deriving luma reference information when the size of the current luma block and the size of the current chroma block are different from each other.
7 is a diagram schematically illustrating an embodiment of a process for generating luma reference information based on encoding parameters.
8 is a diagram schematically illustrating an embodiment of a process for generating luma reference information when an unavailable pixel exists among neighboring luma pixels.
9 is a diagram for explaining an embodiment of a process of generating luma reference information after performing a padding process.
10 is a diagram for explaining another embodiment of a process of generating luma reference information after performing a padding process.
11 is a diagram for explaining another embodiment of a process of generating luma reference information after performing a padding process.
12 is a diagram for explaining another embodiment of a process of generating luma reference information after performing a padding process.
13 is a diagram schematically illustrating an embodiment of a process of deriving first color component reference information to be used for prediction of a second color component block when the size of the first color component block is smaller than the size of the second color component block .
14 is a diagram schematically illustrating an embodiment of a process of deriving first color component reference information to be used for prediction of a second color component block when the size of the first color component block is the same as the size of the second color component block .
15 is a diagram schematically illustrating an embodiment of neighboring luma pixels used to generate luma reference information.
16 is a diagram for explaining an embodiment of a process of deriving a prediction parameter and a process of deriving prediction values of pixels in a current chroma block.
17 is a flowchart schematically illustrating an embodiment of a decoding process according to the present invention.
FIG. 18 is a diagram for explaining an embodiment of a process of deriving luma reference information and prediction parameters for a current chroma block when a decoding process is performed as in the embodiment of FIG. 17 .
19A and 19B are diagrams for explaining another embodiment of a decoding process according to 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, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present specification, the detailed description thereof will be omitted.

When it is said that a component is “connected” or “connected” to another component, it may be directly connected or connected to the other component, but it is understood that other components may exist in between. it should be In addition, the description of "including" a specific configuration in the present invention does not exclude configurations other than the corresponding configuration, and means that additional configurations may be included in the practice of the present invention or the scope of the technical spirit of the present invention.

Terms such as first, second, etc. may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, a second component may also be referred to as a first component.

In addition, the components shown in the embodiment of the present invention are shown independently to represent different characteristic functions, and it does not mean that each component is composed of separate hardware or a single software component. That is, each component is listed as each component for convenience of description, and at least two components of each component are combined to form one component, or one component can be divided into a plurality of components to perform a function, and each Integrated embodiments and separate embodiments of the components are also included in the scope of the present invention without departing from the essence of the present invention.

In addition, some of the components are not essential components for performing essential functions in the present invention, but may be optional components for merely improving performance. The present invention can be implemented by including only essential components to implement the essence of the present invention except for components used for improving performance, and only having a structure including essential components excluding 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 to which the present invention is applied according to an embodiment.

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

The image encoding apparatus 100 may encode an input image in an intra mode or an inter mode and output a bitstream. Intra prediction refers to intra prediction, and inter prediction refers to inter prediction. In the intra mode, the switch 115 may be switched to the intra mode, and in the inter mode, the switch 115 may be switched to the inter mode. The image encoding apparatus 100 may generate a prediction block for an input block of an input image, and then encode a residual between the input block and the prediction block.

In the intra mode, the intra prediction unit 120 may generate a prediction block by performing spatial prediction using pixel values of an already encoded block around the current block.

In the inter mode, the motion prediction unit 111 may obtain a motion vector by finding a region that best matches the input block in the reference image stored in the reference picture buffer 190 during the motion prediction process. The motion compensator 112 may generate a prediction block by performing motion compensation using a motion vector. Here, the motion vector is a two-dimensional vector used for inter prediction, and may indicate an offset between the current encoding/decoding target image and the reference image.

In the prediction process of the intra prediction unit 120 , the motion prediction unit 111 , and/or the motion compensation unit 112 , a processing unit in which a prediction method or a prediction mode is determined and a processing unit in which prediction is performed may be the same, but may be different from each other. may be As an example, when a processing unit for which a prediction method is determined is a PU, a processing unit on which prediction is performed may be a TU.

The subtractor 125 may generate a residual block by the difference between the input block and the generated prediction block. The transform unit 130 may perform transform on the residual block to output a transform coefficient. In addition, the quantization unit 140 may quantize the input transform coefficient according to the quantization parameter to output a quantized coefficient.

The entropy encoding unit 150 may output a bit stream by performing entropy encoding based on the values calculated by the quantization unit 140 or encoding parameter values calculated during the encoding process.

When entropy encoding is applied, a small number of bits are allocated to a symbol having a high probability of occurrence and a large number of bits are allocated to a symbol having a low probability of occurrence to express the symbol, so that bits for symbols to be encoded The size of the column may be reduced. Accordingly, compression performance of image encoding may be improved through entropy encoding. The entropy encoder 150 may use an encoding method such as exponential golomb, Context-Adaptive Variable Length Coding (CAVLC), or Context-Adaptive Binary Arithmetic Coding (CABAC) for entropy encoding.

Since the image encoding apparatus according to the embodiment of FIG. 1 performs inter prediction encoding, that is, inter prediction encoding, the currently encoded image needs to be decoded and stored to be used as a reference image. Accordingly, the quantized coefficient is inverse quantized by the inverse quantization unit 160 and inversely transformed by the inverse transformation unit 170 . The inverse-quantized and inverse-transformed coefficients are added to the prediction block through an adder 175 to generate a reconstructed block.

The reconstructed block passes through the filter unit 180, and the filter unit 180 applies at least one of a deblocking filter, a sample adaptive offset (SAO), and an adaptive loop filter (ALF) to the reconstructed block or the reconstructed picture. can do. The filter unit 180 may be referred to as an adaptive in-loop filter. The deblocking filter can remove block distortion at the boundary between blocks. The SAO may add an appropriate offset value to a pixel value to compensate for a coding error. The ALF may perform filtering based on a value obtained by comparing the reconstructed image and the original image. The reconstructed block passing through the filter unit 180 may be stored in the reference picture buffer 190 .

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

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

The image decoding apparatus 200 may receive a bitstream output from the encoder, perform decoding in an intra mode or an inter mode, and output a reconstructed image, that is, a reconstructed image. In the case of the intra mode, the switch may be switched to the intra mode, and in the case of the inter mode, the switch may be switched to the inter mode. The image decoding apparatus 200 may generate a reconstructed block, that is, a reconstructed block by obtaining a residual block from the received bitstream, generating a prediction block, and adding the residual block and the prediction block.

The entropy decoding unit 210 may entropy-decode the input bitstream according to a probability distribution to generate symbols including symbols in the form of quantized coefficients. The entropy decoding method is similar to the entropy encoding method described above.

When the entropy decoding method is applied, a small number of bits are allocated to a symbol having a high occurrence probability and a large number of bits are allocated to a symbol having a low occurrence probability to represent the symbol, so that the size of the bit stream for each symbol is can be reduced. Accordingly, the compression performance of image decoding may be improved through the entropy decoding method.

The quantized coefficient is inverse quantized by the inverse quantizer 220 and inverse transformed by the inverse transform unit 230 , and as a result of inverse quantization/inverse transformation of the quantized coefficient, a residual block may be generated.

In the intra mode, the intra prediction unit 240 may generate a prediction block by performing spatial prediction using pixel values of an already encoded block around the current block. In the inter mode, the motion compensator 250 may generate a prediction block by performing motion compensation using a motion vector and a reference image stored in the reference picture buffer 270 .

In the prediction process of the intra prediction unit 240 and/or the motion compensator 250 , a processing unit in which a prediction method or a prediction mode is determined and a processing unit in which prediction is performed may be the same or different from each other. As an example, when a processing unit for which a prediction method is determined is a PU, a processing unit on which prediction is performed may be a TU.

The residual block and the prediction block may be added through the adder 255 , and the added block may pass through the filter unit 260 . The filter unit 260 may apply at least one of a deblocking filter, SAO, and ALF to a reconstructed block or a reconstructed picture. The filter unit 260 may output a reconstructed image, that is, a reconstructed image. The reconstructed image may be stored in the reference picture buffer 270 and used for inter prediction.

Hereinafter, a unit may mean a unit of image encoding and decoding. When encoding and decoding an image, an encoding or decoding unit means a divided unit when encoding or decoding an image by dividing it, so a coding unit (CU), a prediction unit (PU: Prediction Unit), and a transformation unit (TU). : Transform Unit), etc. In addition, in the embodiments to be described later, a unit may also be referred to as a block. One unit may be further divided into smaller sub-units.

3 is a conceptual diagram schematically illustrating an embodiment in which one unit is divided into a plurality of sub-units.

One unit may be hierarchically divided with depth information based on a tree structure. Each divided sub-unit may have depth information. Since the depth information indicates the number and/or degree of division of the unit, it may include information on the size of the sub-unit.

310 of FIG. 3 shows an embodiment of a tree structure in which one unit is divided into a plurality of sub-units. In addition, reference numeral 320 of FIG. 3 shows an embodiment of a unit divided into a plurality of sub-units. The unit shown at 320 of FIG. 3 may be, for example, a CU or a TU.

Referring to 310 of FIG. 3 , the highest node may be referred to as a root node and may have the smallest depth value. In this case, the highest node may have a depth of level 0 and may represent an undivided first unit.

A lower node having a depth of level 1 may indicate a unit in which the initial unit is divided once, and a lower node having a depth of level 2 may indicate a unit in which the initial unit is divided twice. For example, in 320 of FIG. 3 , a unit a corresponding to a node a is a unit divided once from the initial unit, and may have a depth of level 1.

A leaf node of level 3 may indicate a unit in which the initial unit is divided three times. For example, unit d corresponding to node d in 320 of FIG. 3 is a unit divided three times in the initial unit, and may have a depth of level 3. Accordingly, the leaf node of level 3, which is the lowest node, may have the deepest depth.

Hereinafter, in embodiments to be described below, the encoding/decoding target block may be referred to as a current block in some cases. Also, when intra prediction is performed on an encoding/decoding object block, the encoding/decoding object block may be referred to as a prediction object block.

On the other hand, the image signal may be composed of color components (color components) constituting the color space (color space). For example, the image signal may include three color components representing three primary color components of light. The three color components representing the three primary color components of light may be represented by R (Red), G (Green), and B (Blue). The R, G, and B signals may be converted into one luma component and two chroma components to reduce a frequency band used for image processing. In this case, one image signal may include one luma component and two chroma components. Here, the luma component is a component representing the brightness of the screen and may be represented by Y. In addition, the two chroma components are components representing the color of the screen, and may be expressed as U, V, Cb, or Cr. Accordingly, the color space of the image may be expressed as YUV or YCbCr as an example. Hereinafter, in embodiments to be described below, a block having a luma component may be referred to as a luma block, and a block having a chroma component may be referred to as a chroma block.

The human eye has a characteristic of being sensitive to a luma component signal and insensitive to a chroma component signal. When the R, G, and B signals are converted into a luma signal and a chroma signal using these characteristics, a frequency band used for image processing may be reduced. Also, due to this characteristic, the number of pixels of the chroma component in one image or block may be smaller than the number of pixels of the luma component.

For example, in a 4:2:0 image format, the number of pixels in the horizontal direction of the chroma block corresponding to the luma block is 1/2 of the number of pixels in the horizontal direction of the luma block, and the number of pixels in the vertical direction of the chroma block corresponding to the luma block The number may correspond to 1/2 of the number of pixels in the vertical direction of the luma block. In the 4:2:2 image format, the number of pixels in the horizontal direction of the chroma block corresponding to the luma block is 1/2 of the number of pixels in the horizontal direction of the luma block, and the number of pixels in the vertical direction of the chroma block corresponding to the luma block is the luma block may be equal to the number of pixels in the vertical direction of In the 4:4:4 image format, the number of pixels in the horizontal direction and the vertical direction of the chroma block corresponding to the luma block may be the same as the number of pixels in the horizontal direction and the vertical direction of the luma block, respectively.

Meanwhile, intra prediction may be performed according to the intra prediction mode of the current block. Since the intra prediction mode of the current block has a high correlation with the prediction mode of a neighboring block, the prediction mode of the current block may be encoded based on the prediction mode of the neighboring block adjacent to the current block block. In this case, MPM (Most Probable Mode) for the current block may be used.

4 shows an example of a prediction direction of an intra prediction mode and a mode value assigned to each prediction direction. Hereinafter, in the present specification, a pixel may have the same meaning as a sample, and in embodiments to be described later, a pixel may be referred to as a sample in some cases.

As described above in the embodiments of FIGS. 1 and 2 , the encoder and the decoder may generate a prediction block by performing intra prediction based on pixel information in the current picture. That is, when performing intra prediction, the encoder and the decoder may perform directional prediction and/or non-directional prediction based on at least one reconstructed reference pixel. Here, the prediction block may mean a block generated as a result of performing intra prediction. The prediction block may correspond to at least one of a coding unit (CU), a prediction unit (PU), and a transform unit (TU).

As described above, intra prediction may be performed according to the intra prediction mode of the current block. The number of intra prediction modes that the current block can have may be a predetermined fixed value, or a value determined differently according to the size of the prediction block. For example, the number of intra prediction modes that the current block may have may be 35 or 36.

Referring to 410 of FIG. 4 , for example, in the case of a vertical mode in which the mode value is 1, prediction may be performed in the vertical direction based on the pixel value of the reference pixel, and in the case of a horizontal mode in which the mode value is 2, the reference pixel Prediction may be performed in the horizontal direction based on the pixel value of . Even in a directional mode other than the above-described mode, the encoder and the decoder may perform intra prediction using the reference pixel according to the corresponding angle.

In 410 of FIG. 4 , a prediction mode having a mode value of 0 may be referred to as a planar mode, and a prediction mode having a mode value of 3 may be referred to as a DC mode. The planar mode (eg, Intra_Planar) and the DC mode (eg, Intra_DC) may correspond to the non-directional mode. For example, in the DC mode, a prediction block may be generated by averaging pixel values of a plurality of reference pixels. In the planner mode, the position of the reference pixel to be used for prediction of the prediction pixel may be determined based on the position of the prediction object pixel, and the prediction value of the prediction object pixel may be derived based on the reference pixel existing at the determined position.

In addition, a prediction mode having a mode value of 35 in 410 of FIG. 4 may be referred to as an LM mode (eg, Intra_FromLuma). The LM mode may refer to a prediction mode in which a prediction value of a chroma component (eg, U or V) pixel is determined according to a pixel value of a luma component (eg, Y). A specific embodiment of intra prediction in the LM mode will be described later.

The number of intra prediction modes and/or a mode value assigned to each intra prediction mode are not limited to the above-described embodiment, and may be differently determined according to implementation and/or necessity. For example, the prediction direction of the intra prediction mode and the mode value assigned to each prediction mode may be determined differently from 410 of FIG. 4 as in 420 of FIG. 4 . Also, as an example in the embodiment 420 of FIG. 4 , the LM mode may be applied as in 410 of FIG. 4 . In this case, a mode value of 35 may be assigned to the LM mode. Hereinafter, in the embodiments to be described below, it is assumed that intra prediction is performed based on the intra prediction mode as in 410 of FIG. 4 , unless otherwise stated for convenience of description.

Meanwhile, since the luma component and the chroma component in the image have a correlation, the prediction mode of the chroma component may be encoded based on the prediction mode of the corresponding luma component, and the decoder predicts the chroma component based on the prediction mode of the luma component. mode can be derived. Accordingly, the prediction mode information of the chroma component transmitted from the encoder to the decoder may be a separate value used to derive the prediction mode of the chroma component in relation to the prediction mode of the luma component, not the prediction mode of the chroma component itself. Table 1 below shows an embodiment of a prediction mode of a luma component and a chroma prediction mode determined according to a transmitted value.

[Table 1]

Referring to Table 1, a value transmitted from the encoder to the decoder may be a value assigned to intra_chroma_pred_mode. IntraPredMode indicates an intra prediction mode of a luma component. For example, when the intra_chroma_pred_mode value is 2 and the IntraPredMode value is 26, the intra prediction mode value of the chroma component may be 10. The intra_chroma_pred_mode and IntraPredMode are not limited to their names. In addition, values assigned to each prediction mode may be differently determined according to implementation and/or necessity as an embodiment.

In the embodiment of Table 1, when the value of intra_chroma_pred_mode is 4, the prediction mode of the chroma component may be called an LM mode, and when the value of the intra_chroma_pred_mode is 5, the prediction mode of the chroma component may be called a DM mode. The LM mode is a prediction mode in which the prediction value of the chroma component pixel is determined according to the pixel value of the luma component, and the DM mode is a prediction mode in which the prediction mode of the luma component is used as the prediction mode of the chroma component as it is.

In addition, the encoder encodes LM flag information indicating whether the LM mode is applied according to the prediction mode of the chroma component in the current sequence, that is, whether the LM mode can be applied to the intra prediction of the chroma component in the current sequence, to the decoder. can be transmitted The LM flag information may be transmitted to the decoder through, for example, a syntax element called chroma_pred_from_luma_enabled_flag. For example, when 1 is assigned to chroma_pred_from_luma_enabled_flag, the flag may indicate that the LM mode may be used for intra prediction of a chroma component. In addition, when 0 is assigned to chroma_pred_from_luma_enabled_flag, the flag may indicate that the LM mode is not used for intra prediction of a chroma component.

When the LM flag is used, when 1 is assigned to chroma_pred_from_luma_enabled_flag, the chroma prediction mode may be determined according to the embodiment of Table 1. However, when 0 is assigned to chroma_pred_from_luma_enabled_flag, since the LM mode is not used for intra prediction of the chroma component, the chroma prediction mode determined according to the prediction mode of the luma component and the transmitted value may be determined differently from the embodiment of Table 1. For example, when the LM mode is not used for intra prediction of a chroma component, the chroma prediction mode may be represented as in the example of Table 2 below.

[Table 2]

Meanwhile, since a plurality of color components constituting the color space have correlations with each other, prediction of another color component (second color component) is performed based on one color component (first color component), and thus the second color Encoding efficiency for components may be improved. This prediction process may also be referred to as 'prediction between color components'. As described above, the color components constituting the color space may include R, G, B, a luma component (eg, Y) and/or a chroma component (eg, U, V, Cb, Cr).

For example, in the above-described LM mode, prediction of the chroma component may be performed based on the luma component. In this case, the first color component may correspond to the luma component, and the second color component may correspond to the chroma component. In the LM mode, the correlation between the luma component and the chroma component may be expressed as a linear relationship using a prediction parameter. A specific example of a process of performing prediction on the chroma component based on the luma component will be described later.

The first color component and the second color component may have various combinations, but in the present specification, the first color component is referred to as a luma component and the second color component is referred to as a chroma component for convenience of description. However, the embodiments described below are not limited thereto and may be applied in the same or similar manner to all possible combinations of the first color component and the second color component. In this case, in embodiments to be described later, the luma component may be viewed as a first color component, and the chroma component may be viewed as a second color component.

5 is a flowchart schematically illustrating an embodiment of an inter-color component prediction method for performing prediction on a chroma component based on a luma component. A prediction process, which will be described later, may be performed by the intra prediction unit of the encoding apparatus and the decoding apparatus.

Hereinafter, in the present specification, a chroma block to be encoded/decoded is referred to as a current chroma block. In addition, the 'restored' luma block corresponding to the current chroma block is referred to as a current luma block. For example, the current luma block may be a block existing in the same position as the current chroma block. As another example, the current luma block may be a block existing at a predetermined relative position with respect to the current chroma block or a predetermined fixed position.

Referring to FIG. 5 , the encoder and the decoder may derive luma reference information to be used for prediction of the current chroma block based on pixels in the current luma block and pixels around the current luma block ( S510 ).

The luma reference information may mean pixel values of a luma component to be used for prediction of a current chroma block. Accordingly, in this specification, the luma reference information may be referred to as a 'luma reference pixel'. In this case, luma reference pixels corresponding to pixels in the current chroma block may constitute a 'luma reference block'. In addition, in the present specification, a pixel positioned around the current luma block is referred to as a 'peripheral luma pixel', a pixel positioned near the current chroma block is referred to as a 'peripheral chroma pixel', and a luma reference pixel positioned around the luma reference block is referred to as a 'peripheral chroma pixel'. This is called the 'peripheral luma reference pixel'. In this case, the luma reference block may correspond to the current chroma block, and the neighboring luma reference pixel may correspond to the neighboring chroma pixel. In this specification, the luma reference information may be used as a concept including a pixel in a luma reference block and a neighboring luma reference pixel.

Meanwhile, as described above, the current luma block and the current chroma block may have different sizes depending on the image format. In this case, the encoder and the decoder perform down-sampling and/or up-sampling on pixels in the current luma block and pixels around the current luma block, thereby performing luma having the same size as the current chroma block. A reference block and its corresponding neighboring luma reference pixels may be derived.

In this case, the encoder and the decoder may differently generate luma reference information based on the encoding parameter. Here, the encoding parameter is a parameter necessary for encoding and decoding, and may include information that can be inferred during encoding or decoding as well as information that is encoded in the encoder and transmitted to the decoder, such as syntax elements, and encodes or decodes an image. It can mean the information you need when you do it. The encoding parameter may include, for example, an encoding mode, an intra prediction mode, an inter prediction mode, and a quantization parameter (QP).

Also, there may be unavailable pixels among pixels located in the vicinity of the current luma block, the luma reference block, and/or the current chroma block. In this case, the encoder and the decoder may perform a padding process on pixels that are not available. Here, the padding may refer to a process of filling in a new pixel value based on an available pixel in a position of an unavailable pixel.

Specific embodiments of the luma reference information derivation process will be described later.

Referring back to FIG. 5 , the encoder and the decoder may derive prediction parameters for the current chroma block based on the luma reference information and the chroma component information ( S520 ).

As an example, the encoder and the decoder may derive a prediction parameter based on pixel values of neighboring luma reference pixels and pixel values of neighboring chroma pixels. Here, the prediction parameters may be alpha (α) and beta (β) based on the least squares method.

Specific embodiments of the prediction parameter derivation process will be described later.

Referring back to FIG. 5 , the encoder and the decoder may derive prediction values of pixels in the current chroma block based on the prediction parameter and the luma reference pixel value in the luma reference block ( S530 ). That is, the encoder and the decoder may perform intra prediction on the current chroma block based on the prediction parameter and the luma reference pixel value in the luma reference block. A specific embodiment of the process of deriving prediction values of pixels in the current chroma block will be described later.

6 illustrates an embodiment of a method of deriving luma reference information when the size of the current luma block and the size of the current chroma block are different from each other.

As described above, the current luma block and the current chroma block may have different sizes according to an image format. For example, when the color space of the image is YUV 4:2:0 or YCbCr 4:2:0, the size of the current luma block may be 4 times larger than the size of the current chroma block. In this case, the encoder and the decoder may derive pixels in the luma reference block having the same size as the current chroma block based on the current luma block, and may derive neighboring luma reference pixels located around the luma reference block. That is, in this case, the encoder and the decoder may generate the luma reference information to correspond to the size of the current chroma block.

610 of FIG. 6 illustrates an embodiment of the current luma block 612 , the neighboring luma pixels 614 , the current chroma block 616 , and the neighboring chroma pixels 618 . Referring to 610 of FIG. 6 , when the size of the current chroma block 616 is 4x4, the size of the corresponding current luma block 612 may be 8x8. In this case, the encoder and the decoder may generate luma reference pixels corresponding to the 4x4 luma reference block.

620 of FIG. 6 illustrates an embodiment of a method for deriving luma reference information.

In the embodiment of 620 of FIG. 6 , as in 610 of FIG. 6 , when the size of the current luma block 612 is 8x8 and the size of the current chroma block 616 is 4x4, the luma reference information 626 and 628 derivation method This is shown.

Referring to 620 of FIG. 6 , the encoder and the decoder, based on the pixel in the current luma block 612 and the neighboring luma pixels 624 , the luma reference block 626 and the neighboring luma reference block 626 to be used for prediction of the current chroma block 616 . Luma reference pixels 628 may be generated. Here, the neighboring luma pixels 624 used for deriving the luma reference information are pixels included in the two pixel lines closest to the left of the current luma block 612 , and at the top of the current luma block 612 . Pixels included in one pixel line located closest to each other may be included.

Meanwhile, in 620 of FIG. 6 , since the sizes of the current luma block 612 and the current chroma block 616 are different from each other, the encoder and the decoder have the same size (4x4) as the luma reference block 626 of the current chroma block 616 . and corresponding neighboring luma reference pixels 628 may be generated. At this time, since the size of the current chroma block 616 is smaller than the size of the current luma block 612 , the encoder and the decoder down-sample the pixels in the current luma block 612 and neighboring luma pixels 624 ( luma reference information can be generated by performing down sampling).

Hereinafter, the coordinates (eg, (x, y)) of the present specification may be coordinates based on the position of the pixel present at the upper leftmost corner in the block. For example, in the present specification, the coordinates of the pixel present at the upper left of the block may be (0,0).

In one embodiment, the encoder and decoder are a 2-tap filter, 3-tap filter and/or 4-tap filter for pixels in the current luma block 612 and surrounding luma pixels 624 . Downsampling may be performed by applying a (4-tap) filter or the like. In this case, the down-sampling process may be expressed by Equation 1 below as an example.

[Equation 1]

p'[x,y] = (p[2x,2y] + p[2x,2y+1]) >> 1

p'[x,y] = (p[2x+1,2y] + p[2x+1,2y+1]) >> 1

p'[x,y] = (p[2x-1,2y+1] + 2*p[2x,2y+1] + p[2x+1,2y+1] + 2) >> 2

Here, p[x,y] may represent the pixel values of the restored luma component pixels, that is, pixels in the current luma block 612 and neighboring luma pixels 624 . Also, p'[x,y] may represent pixel values of the luma reference pixels 626 and 628 generated according to the size of the current chroma block 616 .

In another embodiment, the encoder and the decoder may perform down-sampling by selecting pixels to be used as luma reference pixels from among the pixels in the current luma block 612 and the neighboring luma pixels 624 according to a predetermined operation. In this case, for example, the encoder and the decoder may select one pixel from a pixel group having a predetermined size and use it as a luma reference pixel. This down-sampling process may be performed by, for example, at least one of the down-sampling methods represented by Equations 2, 3, and 4 below.

[Equation 2]

p'[x,y] = p[2x,2y]

[Equation 3]

p'[x,y] = p[2x,2y+1]

[Equation 4]

p'[x,y] = p[2x+1,2y]

Also, different downsampling schemes may be applied to pixels in the current luma block 612 , neighboring luma pixels located at the top of the current luma block 612 , and neighboring luma pixels located to the left of the current luma block 612 . have. For example, when the down-sampling method according to Equation 1 is applied to pixels in the current luma block 612 , the down-sampling method according to Equation 2 is applied to neighboring luma pixels located at the top of the current luma block 612 . Alternatively, the down-sampling method according to Equation 3 may be applied to neighboring luma pixels located on the left side of the current luma block 612 . Here, although not all combinations of possible embodiments to which the down-sampling method is applied are described, those skilled in the art will recognize that other combinations of the above-described down-sampling methods may be used.

630 of FIG. 6 illustrates another embodiment of a method for deriving luma reference information.

In the embodiment of 630 of FIG. 6 , as in 610 of FIG. 6 , when the size of the current luma block 612 is 8x8 and the size of the current chroma block 616 is 4x4, the luma reference information 636 and 638 derivation method This is shown.

Referring to 630 of FIG. 6 , the encoder and the decoder, based on the pixel in the current luma block 612 and the neighboring luma pixels 634 , the luma reference block 636 and the neighboring luma reference block 636 to be used for prediction of the current chroma block 616 . Luma reference pixels 638 may be generated.

In 630 of FIG. 6 , unlike in 620 of FIG. 6 , pixels included in “one” pixel line located closest to the left of the current luma block 612 and closest to the top of the current luma block 612 . By performing down-sampling based on pixels included in one positioned pixel line, the luma reference information 636 and 638 may be derived. That is, in the embodiment 630 of FIG. 6 , only pixels 634 included in one pixel line at the left and upper sides of the current luma block 612 may be used to derive the luma reference information 636 and 638 .

Since the downsampling performing method in 630 of FIG. 6 is the same as in 620 of FIG. 6 , a detailed description of the downsampling performing method will be omitted.

Meanwhile, the encoder and the decoder may differently generate luma reference information based on a block including neighboring luma pixels and/or an encoding parameter of the current luma block. As described above, the encoding parameter may include an encoding mode, an intra prediction mode, an inter prediction mode, and a quantization parameter (QP). For example, the encoder and the decoder may generate luma reference information by using only pixel values of pixels having low distortion.

As an embodiment, the encoder and the decoder may differently generate luma reference information based on the encoding mode. For example, the encoder and the decoder may use only pixels in an intra-mode coded block among pixels in the current luma block and neighboring luma pixels to derive luma reference information.

A further embodiment of a method for generating luma reference information based on an encoding parameter will be described later with reference to FIG. 7 .

7 is a diagram schematically illustrating an embodiment of a process for generating luma reference information based on encoding parameters.

710 of FIG. 7 illustrates an embodiment of a method of generating luma reference information 716 and 718 based on the intra prediction mode of the current luma block 712 . In 710 of FIG. 7 , it is assumed that the size of the current luma block 712 is 8x8 and the size of the current chroma block is 4x4.

In 710 of FIG. 7 , a luma reference block 716 and neighboring luma reference pixels 718 to be used for prediction of the current chroma block are generated based on the pixels in the current luma block 712 and the neighboring luma pixels 714 . can be Since the size of the current chroma block is 4x4, the size of the generated luma reference block 716 may be 4x4.

At this time, since the intra prediction mode of the current luma block 712 is a vertical mode 722 , the encoder and the decoder use only pixels located adjacent to the top of the current luma block 712 among the neighboring luma pixels 714 . may be used to generate luma reference information 716 and 718 . Accordingly, the peripheral luma reference pixels 718 derived from 710 of FIG. 7 may be pixels located adjacent to the top of the luma reference block 716 .

Unlike in 710 of FIG. 7 , if the intra prediction mode of the current luma block is a horizontal mode, the encoder and the decoder may use only pixels located adjacent to the left side of the current luma block among neighboring luma pixels to generate luma reference information. have. In this case, the derived neighboring luma reference pixels may be pixels located adjacent to the left side of the luma reference block. When the intra prediction mode of the current luma block is the DC mode or the planar mode, the encoder and the decoder may use both pixels located at the left side of the current luma block and pixels located at the top of the current luma block to generate luma reference information. In this case, the derived neighboring luma reference pixels may include both pixels located on the left side of the luma reference block and pixels located on the upper side of the luma reference block.

730 of FIG. 7 illustrates an embodiment of a method of generating luma reference information 752 and 754 based on quantization parameters of neighboring luma blocks 734 , 736 , and 738 located in the vicinity of the current luma block 732 . In 730 of FIG. 7 , it is assumed that the size of the current luma block 732 is 8x8 and the size of the current chroma block is 4x4.

Referring to 730 of FIG. 7 , the quantization parameter value of the upper peripheral luma block 734 located adjacent to the upper end of the current luma block 732 is 22, and the upper left corner located at the upper left corner outside the current luma block 732 . The quantization parameter value of the peripheral luma block 736 is 26. Also, the quantization parameter value of the left peripheral luma block 738 located adjacent to the left side of the current luma block 732 is 28. In this case, the encoder and the decoder may use the neighboring luma pixels belonging to the neighboring luma block having the smallest quantization parameter value among the neighboring luma pixels 744 , 746 , and 748 to generate the luma reference information.

In 730 of FIG. 7 , the upper peripheral luma block 734 among the upper peripheral luma block 734, the upper left peripheral luma block 736, and the left peripheral luma block 738 has the smallest quantization parameter value (eg, 22). Therefore, the peripheral luma pixels 744 in the upper peripheral luma block 734 may be used to generate luma reference information. That is, in 730 of FIG. 7 , the encoder and the decoder convert the pixels in the current luma block 732 and the neighboring luma pixels 744 in the upper peripheral luma block 734 to the luma reference block 752 and the neighboring luma reference pixels ( 754) can be used.

Since the size of the current chroma block in 730 of FIG. 7 is 4x4, the size of the generated luma reference block 752 may be 4x4. Also, the peripheral luma reference pixels 754 derived from 730 of FIG. 7 may be pixels located adjacent to the upper end of the luma reference block 752 .

In the embodiment of FIG. 7 , a downsampling process performed to generate the luma reference information is similar to that in the embodiment of FIG. 6 , so a detailed description thereof will be omitted herein.

8 is a diagram schematically illustrating an embodiment of a process for generating luma reference information when an unavailable pixel exists among neighboring luma pixels.

As described above, among the neighboring luma pixels located in the vicinity of the current luma block, there may be unavailable pixels. Cases in which the neighboring luma pixels are not available include when the current luma block is located adjacent to the boundary of the current picture, when the current luma block is located adjacent to the boundary of the current slice, and constrained intra (CIP) for the current luma block Prediction) may be applied.

810 of FIG. 8 illustrates an embodiment of a method of deriving luma reference information when a current luma block is located adjacent to a boundary of a current picture and/or a current slice. Referring to 810 of FIG. 8 , the current luma block 812 may be a block located adjacent to the left boundary 818 of the current picture and/or the current slice. In this case, the neighboring luma pixels 816 located on the left side of the current luma block 812 are located outside the current picture and/or the current slice, and thus may correspond to pixels that are not available.

In this case, according to an embodiment, the encoder and the decoder may perform a padding process on pixels that are not available before performing the downsampling process. As described above, padding may refer to a process of filling new pixel values in positions of pixels that are not available.

As in 810 of FIG. 8 , when the neighboring luma pixels 816 located on the left side of the current luma block 812 are not available, the encoder and the decoder are available neighboring luma pixels located on the top of the current luma block 812 . A pixel value of at least one pixel among the pixels 814 may be filled in positions of neighboring luma pixels located on the left side of the current luma block 812 . For example, the process of filling the pixel value in the first pixel line closest to the left of the current luma block 812 may be expressed by Equation 5 below.

[Equation 5]

p[-1,y] = p[0,-1], (y=0-7)

Here, p[x,y] may represent a reconstructed pixel value of the luma component.

The process of filling a pixel value into a pixel line located second closest to the left of the current luma block 812 , that is, a second pixel line located adjacent to the left side of the first pixel line, may be expressed by Equation 6 below as an example. can

[Equation 6]

p[-2,y] = p[0,-1], (y=0-7)

When the padding process is performed, the encoder and the decoder may use the newly filled pixel value to generate the luma reference information. That is, the encoder and the decoder based on the pixel values of the available neighboring luma pixels 814, the pixel values of the pixels 816 filled by the padding process, and the pixel values of the pixels in the current luma block 812 luma reference information can create

In another embodiment, the encoder and decoder may not use neighboring luma pixels that are not available to generate luma reference information. That is, the encoder and the decoder may generate luma reference information based on pixels in the current luma block and available neighboring luma pixels. For example, in 810 of FIG. 8 , pixels in the current luma block 812 and available neighboring luma pixels 814 positioned above the current luma block 812 may be used to generate luma reference information.

In the process of generating the luma reference information in 810 of FIG. 8 , at least one of the luma reference information generating methods described above in the embodiments of FIGS. 6 and 7 may be applied together with the above-described embodiments.

820 of FIG. 8 illustrates an embodiment of a method of deriving luma reference information when constrained intra prediction (CIP) is applied.

The CIP may refer to an intra prediction method in which only pixels within a block encoded/decoded by intra prediction among blocks located in the vicinity of the prediction object block are used for prediction. When CIP is applied, pixels in a block encoded/decoded by inter prediction may not be used for intra prediction.

For example, the encoder may transmit flag information on whether or not CIP is applied to the decoder. In this case, the decoder may determine whether to apply the CIP based on the flag information. Flag information on whether or not CIP is applied may be expressed as 'constrained_intra_pred_flag', for example.

Referring to 820 of FIG. 8 , the upper peripheral luma block 824 located at the upper end of the current luma block 822 is a block encoded/decoded in the intra mode, and the left side located on the left side of the current luma block 822 . The peripheral luma block 826 may be a block encoded/decoded in an inter mode. In this case, the neighboring luma pixels 836 located in the left peripheral luma block 826 may correspond to pixels that are not available.

In this case, according to an embodiment, the encoder and the decoder may perform a padding process on pixels that are not available before performing the downsampling process. As described above, padding may refer to a process of filling new pixel values in positions of pixels that are not available. In this case, the encoder and the decoder may fill the unused neighboring luma pixels with pixel values of the available neighboring luma pixels.

For example, when the neighboring luma pixels 836 located in the left peripheral luma block 826 are not available as in 820 of FIG. A pixel value of at least one pixel among the luma pixels 834 may be filled in positions of adjacent luma pixels 836 that are not available. As an example, a process of performing padding based on pixels in the upper peripheral luma block 824 may be represented by Equation 7 below.

[Equation 7]

p[-2,y] = p[-2,-1] (y=0-7) or

p[-2,y] = p[-1,-1] (y=0-7) or

p[-1,y] = p[-1,-1] (y=0-7)

As another example, the pixels 838 positioned adjacent to the lower end of the left peripheral luma block 826 may be pixels in the intra-mode coded/decoded block. In this case, the encoder and the decoder include a pixel located at the top of the left peripheral luma block (eg, a pixel present at a position of (-2,-1) or a pixel present at a position of (-1, -1)) and a left periphery of the luma block. Neighboring luma pixels 836 in which the pixel average value of a pixel located at the bottom of the luma block (eg, a pixel at a position of (-2,8) or a pixel at a position of (-1,8)) is not available ) can be filled in. This padding process may be expressed by Equation (8) below as an example.

[Equation 8]

p[-2,y] = (p[-2,-1] + p[-2,8] + 1)>>1, (y=0-7) or

p[-2,y] = (p[-1,-1] + p[-1,8] + 1)>>1, (y=0~7)

*163 When the padding process is performed, the encoder and the decoder may use the newly filled pixel value to generate luma reference information. That is, the encoder and the decoder based on the pixel values of the available neighboring luma pixels 834, the pixel values of the pixels 836 filled by the padding process, and the pixel values of the pixels in the current luma block 822, luma reference information can create

As another embodiment, in 820 of FIG. 8 , neighboring luma pixels that are not available to the encoder and decoder may not be used to generate luma reference information. That is, the encoder and the decoder may generate luma reference information based on pixels in the current luma block and available neighboring luma pixels. For example, in 820 of FIG. 8 , pixels in the current luma block 822 and available neighboring luma pixels 834 located in the upper peripheral luma block 824 may be used to generate luma reference information.

In the luma reference information generation process in 820 of FIG. 8 , at least one of the luma reference information generation methods described above in the embodiments of FIGS. 6 and 7 may be applied together with the above-described embodiments.

9 is a diagram for explaining an embodiment of a process of generating luma reference information after performing a padding process.

In the embodiment of FIG. 9 , the first left pixel line closest to the left of the current luma block 910 and the neighboring luma pixels in the first upper pixel line closest to the top of the current luma block 910 generate luma reference information. Examples of when used in In the embodiment of FIG. 9 , a pixel line located second closest to the left of the current luma block 910 , that is, a pixel line located adjacent to the left side of the first left pixel line will be referred to as a second left pixel line. Also, a pixel line that is second closest to the upper end of the current luma block 910 , that is, a pixel line positioned adjacent to the upper end of the first upper pixel line will be referred to as a second upper pixel line.

In the embodiment of FIG. 9 , it is assumed that the intra prediction mode of the current chroma block to be predicted is the LM mode. As an example, the LM mode may be represented by Intra_FromLuma. Also, in FIG. 9 , it is assumed that the size of the current luma block 910 is 8x8 and the size of the current chroma block is 4x4. As described above, the size of the luma reference block to be used for prediction of the current chroma block may be the same as the size of the current chroma block. Accordingly, the size of the luma reference block 930 derived in the embodiment of FIG. 9 may be 4x4.

In addition, among the neighboring luma pixels 922, 924, 926, 928 shown in the embodiment of FIG. 9, the neighboring luma pixels corresponding to 922 and the neighboring luma pixels corresponding to 926 are available pixels, It is assumed that the neighboring luma pixels corresponding to 924 and the neighboring luma pixels corresponding to 928 are unavailable pixels. In this case, the encoder and the decoder may generate luma reference information after performing a padding process on the neighboring luma pixels that are not available.

9 , the encoder and the decoder may perform a padding process after copying neighboring luma pixel values on the second left pixel line to the first left pixel line. That is, the encoder and the decoder replace the peripheral luma pixel values of the first left pixel line with the peripheral luma pixel values of the second left pixel line, and then, after A padding process may be performed. A process of copying the peripheral luma pixel values on the second left pixel line to the first left pixel line may be represented by Equation 9 below, for example.

[Equation 9]

P _Y [x,y] = P _Y [x-1,y], (x=-1, y=0~2*nS-1)

*173 Here, P _Y may represent a pixel value of a neighboring luma pixel. Also, nS may indicate the horizontal length and vertical length of the current luma block 910 .

After the neighboring luma pixel values on the second pixel line are copied to the first left pixel line, the encoder and the decoder perform a padding process on the neighboring luma pixels, thereby deriving pixel values of neighboring luma pixels to be used for generating luma reference information. can do. Pixel values of neighboring luma pixels to be used for generating luma reference information may be expressed by Equation 10 below.

[Equation 10]

P _Y [x,y], (x=-1, y=-1~nS*2-1), (x=0~nS*2-1, y=-1)

Pixels that are not available among the surrounding luma pixels may constitute a pixel group as shown in 924 and 928 of FIG. 9 . In the embodiment of FIG. 9 , for convenience of description, a group of unavailable pixels on the first left pixel line (eg, four pixels located on the left in 928 of FIG. 9 ) is referred to as a left pixel group, and the first upper pixel line The group of unavailable pixels 924 on the top is referred to as the top pixel group. 9 , pixels belonging to the upper pixel group may be represented by P _Y [x,y], (x=nS~nS*2-1, y=-1), and pixels belonging to the left pixel group can be expressed as P _Y [x,y], (x=-1, y=nS~nS+(nS>>1)-1).

The encoder and the decoder may determine whether available pixels located adjacent to the pixel group exist at both ends of each pixel group (eg, a left pixel group and an upper pixel group) to perform padding.

For example, an available pixel located adjacent to a pixel group may exist only at one end of the pixel group. Referring to FIG. 9 , an available pixel may not exist on the right side of the upper pixel group 924 , and an available pixel c may exist on the left side of the upper pixel group 924 . Here, pixel c may be represented by P _Y [nS-1,-1].

In this case, there may be only one available pixel positioned adjacent to the upper pixel group 924 within the first upper pixel line. Accordingly, the encoder and the decoder may fill the pixel values of the available pixels c located adjacent to the upper pixel group into positions of the unavailable pixels in the upper pixel group 924 . This padding process may be expressed by Equation 11 below as an example.

[Equation 11]

P _Y [x,y] = P _Y [nS-1,-1], (x=nS~nS*2-1, y=-1)

As another example, available pixels located adjacent to a pixel group may exist at both ends of the pixel group. Referring to FIG. 9 , an available pixel b may be adjacent to an upper end of the left pixel group, and an available pixel a may be adjacent to a lower end of the left pixel group. Here, the pixel a may be represented by P _Y [-1, nS+(nS>>1)], and the pixel b may be represented by P _Y [-1, nS-1].

In this case, two available pixels located adjacent to the left pixel group in the first left pixel line may exist. Accordingly, the encoder and the decoder may fill the pixel average values of the available pixels (a, b) located adjacent to the left pixel group into the positions of the unavailable pixels in the left pixel group. This padding process may be expressed by Equation 12 below as an example.

[Equation 12]

P _Y [x,y] = (P _Y [-1,nS-1] + P _Y [-1,nS+(nS>>1)] + 1) >> 1,

*187(x=-1, y=nS~nS+(nS>>1)-1)

By the above-described padding process, pixel values of the neighboring luma pixels on the first left pixel line and the neighboring luma pixels on the first upper pixel line may be determined. At this time, the encoder and the decoder based on the pixel values of the neighboring luma pixels on the first left pixel line, the neighboring luma pixels on the first upper pixel line, and the pixels in the current luma block 910, luma reference information (see luma). A block 930 and peripheral luma reference pixels 940 may be generated. However, the neighboring luma pixels existing at the position of [-1,-1] may not be used to generate the luma reference information. In the embodiment of FIG. 9 , the luma reference information generation process may be expressed by Equation 13 below as an example.

[Equation 13]

P _Y '[x,y] = (P _Y [2x-1,y] + 2*P _Y [2x,y] + P _Y [2x+1,y] + 2) >> 2, (x=0 ~nS-1, y=-1)

P _Y '[x,y] = (P _Y [x,2y] + P _Y [x,2y+1]) >> 1, (x=-1, y=0~nS-1)

P _Y '[x,y] =(P _RecY [2x,2y] + P _RecY [2x,2y+1]) >> 1, (x=0~nS-1, y=0~nS-1)

Here, P _RecY may represent a pixel value of a pixel in the current luma block. In addition, P _Y ′ may indicate a pixel value corresponding to luma reference information (eg, a pixel value of a pixel in a luma reference block and/or a pixel value of a neighboring luma reference pixel).

10 is a diagram for explaining another embodiment of a process of generating luma reference information after performing a padding process.

In the embodiment of FIG. 10 , the first left pixel line closest to the left of the current luma block 1010 and the neighboring luma pixels in the first upper pixel line closest to the top of the current luma block 1010 generate luma reference information. Examples of when used in In the embodiment of FIG. 10 , a pixel line located second to the left of the current luma block 1010 , that is, a pixel line located adjacent to the left of the first left pixel line, will be referred to as a second left pixel line. Also, a pixel line that is second closest to the top of the current luma block 1010 , that is, a pixel line positioned adjacent to the top of the first top pixel line will be referred to as a second top pixel line.

In the embodiment of FIG. 10 , it is assumed that the intra prediction mode of the current chroma block to be predicted is the LM mode. As an example, the LM mode may be represented by Intra_FromLuma. Also, in FIG. 10 , it is assumed that the size of the current luma block 1010 is 8x8 and the size of the current chroma block is 4x4. As described above, the size of the luma reference block to be used for prediction of the current chroma block may be the same as the size of the current chroma block. Accordingly, the size of the luma reference block 1030 derived in the embodiment of FIG. 10 may be 4x4.

In addition, among the neighboring luma pixels 1022, 1024, 1026, and 1028 shown in the embodiment of FIG. 10, the neighboring luma pixels corresponding to 1022 and the neighboring luma pixels corresponding to 1026 are available pixels, It is assumed that the neighboring luma pixels corresponding to 1024 and the neighboring luma pixels corresponding to 1028 are unavailable pixels. In this case, the encoder and the decoder may generate luma reference information after performing a padding process on the neighboring luma pixels that are not available.

In the embodiment of FIG. 10 , the encoder and the decoder perform a padding process on the neighboring luma pixels that are not available on the first left pixel line and the first upper pixel line, so that the pixel values of the neighboring luma pixels P _Y [x,y] , (x=-1, y=-1 to nS*2-1), (x=0 to nS*2-1, y=-1) can be derived. Here, P _Y may represent a pixel value of a neighboring luma pixel. Also, nS may indicate the horizontal length and vertical length of the current luma block 1010 .

Pixels that are not available among the surrounding luma pixels may constitute a pixel group as shown in 1024 and 1028 of FIG. 10 . In the embodiment of FIG. 10 , for convenience of description, a group of unavailable pixels on the first left pixel line (eg, four pixels located on the left in 1028 of FIG. 10 ) is referred to as a left pixel group, and the first upper pixel line The group of pixels 1024 that are not available on the top is referred to as the top pixel group. In the embodiment of FIG. 10 , pixels belonging to the upper pixel group may be represented by P _Y [x,y], (x=nS~nS*2-1, y=-1), and pixels belonging to the left pixel group can be expressed as P _Y [x,y], (x=-1, y=nS~nS+(nS>>1)-1).

The encoder and the decoder may determine whether available pixels located adjacent to the pixel group exist at both ends of each pixel group (eg, a left pixel group and an upper pixel group) to perform padding.

For example, an available pixel located adjacent to a pixel group may exist only at one end of the pixel group. Referring to FIG. 10 , an available pixel may not exist on the right side of the upper pixel group 1024 , and an available pixel c may exist on the left side of the upper pixel group 1024 . Here, pixel c may be represented by P _Y [nS-1,-1].

In this case, there may be only one available pixel positioned adjacent to the upper pixel group 1024 within the first upper pixel line. Accordingly, the encoder and the decoder may fill the pixel values of the available pixels c located adjacent to the upper pixel group into positions of the unavailable pixels in the upper pixel group 1024 . This padding process may be expressed by Equation 14 below as an example.

[Equation 14]

P _Y [x,y] = P _Y [nS-1,-1], (x=nS~nS*2-1, y=-1)

As another example, available pixels located adjacent to a pixel group may exist at both ends of the pixel group. Referring to FIG. 10 , an available pixel b may be adjacent to an upper end of the left pixel group, and an available pixel a may be adjacent to a lower end of the left pixel group. Here, the pixel a may be represented by P _Y [-1, nS+(nS>>1)], and the pixel b may be represented by P _Y [-1, nS-1].

In this case, two available pixels located adjacent to the left pixel group in the first left pixel line may exist. Accordingly, the encoder and the decoder may fill the pixel average values of the available pixels (a, b) located adjacent to the left pixel group into the positions of the unavailable pixels in the left pixel group. This padding process may be expressed by Equation 15 below as an example.

[Equation 15]

P _Y [x,y] = (P _Y [-1,nS-1] + P _Y [-1,nS+(nS>>1)] + 1) >> 1,

*209(x=-1, y=nS~nS+(nS>>1)-1)

By the above-described padding process, pixel values of the neighboring luma pixels on the first left pixel line and the neighboring luma pixels on the first upper pixel line may be determined. Based on the pixel values of the neighboring luma pixels on the first left pixel line, the neighboring luma pixels on the first upper pixel line, and the pixels in the current luma block 1010 , the encoder and the decoder determine the luma reference information (luma reference block 1030 ). ) and peripheral luma reference pixels 1040 ).

In this case, available pixels may exist among the pixels on the second left pixel line positioned adjacent to the left of the first left pixel line. If a pixel located adjacent to the left side of a pixel on the first left pixel line (a pixel on the second left pixel line) is available, the encoder and decoder instead of the pixel on the first left pixel line located adjacent thereto a second left pixel line A pixel on the image may be used to generate luma reference information. However, if a pixel located adjacent to the left of a pixel on the first left pixel line (a pixel on the second left pixel line) is not available, the encoder and decoder refer to the pixel on the first left pixel line on which the padding process has been performed as luma. It can be used to generate information.

In the embodiment of FIG. 10 , a process of generating the luma reference information may be expressed by Equation 16 below as an example.

[Equation 16]

P _Y '[x,y] = (P _Y [2x-1,y] + 2*P _Y [2x,y] + P _Y [2x+1,y] + 2) >> 2, (x=0 ~nS-1, y=-1)

P _Y '[x,y] = (P _RecY [2x,2y] + P _RecY [2x,2y+1]) >> 1, (x=0~nS-1, y=0~nS-1)

If P _Y [2x,2y], (x=-1, y=0~nS-1) is available,

P _Y '[x,y] = (P _Y [2x,2y] + P _Y [2x,2y+1]) >> 1, (x=-1, y=0~nS-1)

If P _Y [2x,2y], (x=-1, y=0~nS-1) is not available,

P _Y '[x,y] = (P _Y [x,2y] + P _Y [x,2y+1]) >> 1, (x=-1, y=0~nS-1)

Here, P _RecY may represent a pixel value of a pixel in the current luma block. In addition, P _Y ′ may indicate a pixel value corresponding to luma reference information (eg, a pixel value of a pixel in a luma reference block and/or a pixel value of a neighboring luma reference pixel).

11 is a diagram for explaining another embodiment of a process of generating luma reference information after performing a padding process.

In the embodiment of FIG. 11 , peripheral luma pixels in the first left pixel line closest to the left of the current luma block 1110 and the first upper pixel line closest to the top of the current luma block 1110 generate luma reference information. Examples of when used in In the embodiment of FIG. 11 , a pixel line located second closest to the left of the current luma block 1110, that is, a pixel line located adjacent to the left of the first left pixel line, will be referred to as a second left pixel line. Also, a pixel line that is second closest to the upper end of the current luma block 1110 , that is, a pixel line positioned adjacent to the upper end of the first upper pixel line will be referred to as a second upper pixel line.

In the embodiment of FIG. 11 , it is assumed that the intra prediction mode of the current chroma block to be predicted is the LM mode. As an example, the LM mode may be represented by Intra_FromLuma. Also, in FIG. 11 , it is assumed that the size of the current luma block 1110 is 8x8 and the size of the current chroma block is 4x4. As described above, the size of the luma reference block to be used for prediction of the current chroma block may be the same as the size of the current chroma block. Accordingly, the size of the luma reference block 1130 derived in the embodiment of FIG. 11 may be 4x4.

In addition, among the neighboring luma pixels 1122, 1124, 1126, and 1128 shown in the embodiment of FIG. 11, the neighboring luma pixels corresponding to 1122 and the neighboring luma pixels corresponding to 1126 are available pixels, It is assumed that the neighboring luma pixels corresponding to 1124 and the neighboring luma pixels corresponding to 1128 are unavailable pixels. In this case, the encoder and the decoder may generate luma reference information after performing a padding process on the neighboring luma pixels that are not available.

In the embodiment of FIG. 11 , the encoder and the decoder may perform a padding process on neighboring luma pixels that are not available on the first left pixel line, the second left pixel line, and the first upper pixel line. In addition, the encoder and the decoder may copy neighboring luma pixel values on the second left pixel line to the first left pixel line. Pixel values of neighboring luma pixels to be used for generating reference information, which are generated by the above-described process, may be expressed by Equation 17 below.

[Equation 17]

P _Y [x,y], (x=-1, y=-1~nS*2-1), (x=0~nS*2-1, y=-1)

Here, P _Y may represent a pixel value of a neighboring luma pixel belonging to the first left pixel line and/or the first upper pixel line. Also, nS may indicate the horizontal length and vertical length of the current luma block 1110 .

Pixels that are not available among the surrounding luma pixels may constitute a pixel group as shown in 1124 and 1128 of FIG. 11 . In the embodiment of FIG. 11 , for convenience of explanation, a group of unavailable pixels on the first left pixel line (eg, four pixels located on the left in 1128 of FIG. 11 ) is referred to as a first left pixel group, and the second left pixel group is referred to as a first left pixel group. The group of unavailable pixels on the pixel line (eg, the four pixels located on the right in 1128 in FIG. 11 ) is referred to as the second left pixel group, and the group of unavailable pixels 1124 on the first top pixel line The group is called the top pixel group. In the embodiment of FIG. 11 , pixels belonging to the upper pixel group may be represented by P _Y [x,y], (x=nS~nS*2-1, y=-1), and are located in the first left pixel group. Pixels belonging to P _Y [x,y], (x=-1, y=nS~nS+(nS>>1)-1) may be expressed, and pixels belonging to the second left pixel group are P _Y2 [x , y], (x=-2, y=nS~nS+(nS>>1)-1). Here, P _Y2 may represent a pixel value of a neighboring luma pixel belonging to the second left pixel line.

The encoder and the decoder determine whether there are available pixels located adjacent to the pixel group at both ends of each pixel group (eg, the first left pixel group, the second left pixel group, and the top pixel group) to perform padding. can be judged

In one embodiment, an available pixel located adjacent to a pixel group may exist only at one end of the pixel group. Referring to FIG. 11 , an available pixel may not exist on the right side of the upper pixel group 1124 , and an available pixel c may exist on the left side of the upper pixel group 1124 . Here, pixel c may be represented by P _Y [nS-1,-1].

In this case, there may be only one available pixel positioned adjacent to the upper pixel group 1124 within the first upper pixel line. Accordingly, the encoder and the decoder may fill the pixel values of the available pixels c located adjacent to the upper pixel group into positions of the unavailable pixels in the upper pixel group 1124 . This padding process may be expressed by Equation 18 below as an example.

[Equation 18]

P _Y [x,y] = P _Y [nS-1,-1], (x=nS~nS*2-1, y=-1)

In another embodiment, available pixels located adjacent to a group of pixels may exist at both ends of the group of pixels.

For example, in the embodiment of FIG. 11 , an available pixel b may be adjacent to an upper end of the first left pixel group, and an available pixel a may be adjacent to a lower end of the first left pixel group. Here, the pixel a may be represented by P _Y [-1, nS+(nS>>1)], and the pixel b may be represented by P _Y [-1, nS-1].

In this case, two available pixels located adjacent to the left pixel group in the first left pixel line may exist. Accordingly, the encoder and the decoder may fill the pixel average values of the available pixels a and b located adjacent to the first left pixel group into positions of the unavailable pixels in the first left pixel group. This padding process may be expressed by Equation 19 below as an example.

[Equation 19]

P _Y [x,y] = (P _Y [-1,nS-1] + P _Y [-1,nS+(nS>>1)] + 1) >> 1,

(x=-1, y=nS~nS+(nS>>1)-1)

As another example, in the embodiment of FIG. 11 , an available pixel e may be adjacent to an upper end of the second left pixel group, and an available pixel d may be adjacent to a lower end of the second left pixel group. Here, the pixel d may be represented by P _Y2 [-2, nS+(nS>>1)], and the pixel e may be represented by P _Y2 [-2, nS-1].

In this case, there may be two available pixels positioned adjacent to the left pixel group in the second left pixel line. Accordingly, the encoder and the decoder may fill the pixel average values of the available pixels (e, d) located adjacent to the second left pixel group into positions of the unavailable pixels in the second left pixel group. This padding process may be expressed by Equation 20 below as an example.

[Equation 20]

P _Y2 [x,y] = (P _Y2 [-2,nS-1] + P _Y2 [-2,nS+(nS>>1)] + 1) >> 1,

(x=-2, y=nS~nS+(nS>>1)-1)

The pixel values on the second left pixel line determined after the padding process is performed may be expressed as, for example, P _Y2 [x,y], (x=-2, y=-1 to 2*nS-1).

By the above-described padding process, pixel values of neighboring luma pixels on the first left pixel line, the second left pixel line, and the first upper pixel line may be determined. In this case, the encoder and the decoder may derive pixel values on the first left pixel line to be used for generating luma reference information by copying the neighboring luma pixel values on the second left pixel line to the first left pixel line. That is, the encoder and the decoder may replace the peripheral luma pixel values of the first left pixel line with the peripheral luma pixel values of the second left pixel line. A process of copying the peripheral luma pixel values on the second left pixel line to the first left pixel line may be represented by Equation 21 below as an example.

[Equation 21]

P _Y [x,y] = P _Y2 [x-1,y], (x=-1, y=0~2*nS-1)

After the peripheral luma pixel values on the second primary pixel line are copied to the first left pixel line, the encoder and decoder perform the first left pixel line peripheral luma pixels, the first upper pixel line peripheral luma pixels on the first top pixel line and the current luma block. Based on the pixel values of the pixels in 1110 , luma reference information (the luma reference block 1130 and the neighboring luma reference pixels 1140 ) may be generated. In the embodiment of FIG. 11 , the luma reference information generation process may be expressed by Equation 22 below as an example.

[Equation 22]

P _Y '[x,y] = (P _Y [2x-1,y] + 2*P _Y [2x,y] + P _Y [2x+1,y] + 2) >> 2, (x=0 ~nS-1, y=-1)

P _Y '[x,y] = (P _Y [x,2y] + P _Y [x,2y+1]) >> 1, (x=-1, y=0~nS-1)

P _Y '[x,y] = (P _RecY [2x,2y] + P _RecY [2x,2y+1]) >> 1, (x=0~nS-1, y=0~nS-1)

Here, P _RecY may represent a pixel value of a pixel in the current luma block. In addition, P _Y ′ may indicate a pixel value corresponding to luma reference information (eg, a pixel value of a pixel in a luma reference block and/or a pixel value of a neighboring luma reference pixel).

12 is a diagram for explaining another embodiment of a process of generating luma reference information after performing a padding process.

12 , the first left pixel line closest to the left of the current luma block 1210 and the neighboring luma pixels within the first upper pixel line closest to the top of the current luma block 1210 are used to generate luma reference information. Examples of when used are shown. In the embodiment of FIG. 12 , the pixel line located second closest to the left side of the current luma block 1210, that is, the pixel line located adjacent to the left side of the first left pixel line, will be referred to as a second left pixel line. Also, a pixel line that is second closest to the upper end of the current luma block 1210 , that is, a pixel line positioned adjacent to the upper end of the first upper pixel line, will be referred to as a second upper pixel line.

In the embodiment of FIG. 12 , it is assumed that the intra prediction mode of the current chroma block to be predicted is the LM mode. As an example, the LM mode may be represented by Intra_FromLuma. Also, in FIG. 12 , it is assumed that the size of the current luma block 1210 is 8x8 and the size of the current chroma block is 4x4. As described above, the size of the luma reference block to be used for prediction of the current chroma block may be the same as the size of the current chroma block. Accordingly, the size of the luma reference block 1230 derived in the embodiment of FIG. 12 may be 4x4.

In addition, among the neighboring luma pixels 1222, 1224, 1226, and 1228 shown in the embodiment of FIG. 12, the neighboring luma pixels corresponding to 1222 and the neighboring luma pixels corresponding to 1226 are available pixels, It is assumed that the neighboring luma pixels corresponding to 1224 and the neighboring luma pixels corresponding to 1228 are unavailable pixels. In this case, the encoder and the decoder may generate luma reference information after performing a padding process on the neighboring luma pixels that are not available.

In the embodiment of FIG. 12 , the encoder and the decoder perform a padding process on the neighboring luma pixels that are not available on the first left pixel line and the first upper pixel line, so that the pixel values of the neighboring luma pixels P _Y [x,y] , (x=-1, y=-1 to nS*2-1), (x=0 to nS*2-1, y=-1) can be derived. Here, P _Y may represent a pixel value of a neighboring luma pixel. Also, nS may indicate the horizontal length and vertical length of the current luma block 1210 .

Among the neighboring luma pixels, pixels that are not available may constitute a pixel group as shown in 1224 and 1228 of FIG. 12 . In the embodiment of FIG. 12 , for convenience of description, a group of unavailable pixels on the first left pixel line (eg, four pixels located on the left in 1228 of FIG. 12 ) is referred to as a left pixel group, and the first upper pixel line The group of pixels 1224 that are not available on the top is referred to as the top pixel group. 12 , pixels belonging to the upper pixel group may be represented by P _Y [x,y], (x=nS~nS*2-1, y=-1), and pixels belonging to the left pixel group can be expressed as P _Y [x,y], (x=-1, y=nS~nS+(nS>>1)-1).

The encoder and the decoder may determine whether available pixels located adjacent to the pixel group exist at both ends of each pixel group (eg, a left pixel group and an upper pixel group) to perform padding.

For example, an available pixel located adjacent to a pixel group may exist only at one end of the pixel group. Referring to FIG. 12 , an available pixel may not exist on the right side of the upper pixel group 1224 , and an available pixel c may exist on the left side of the upper pixel group 1224 . Here, pixel c may be represented by P _Y [nS-1,-1].

In this case, there may be only one available pixel positioned adjacent to the upper pixel group 1224 within the first upper pixel line. Accordingly, the encoder and the decoder may fill the pixel values of the available pixels c located adjacent to the upper pixel group into positions of the unavailable pixels in the upper pixel group 1224 . This padding process may be expressed by Equation 23 below as an example.

[Equation 23]

P _Y [x,y] = P _Y [nS-1,-1], (x=nS~nS*2-1, y=-1)

As another example, available pixels located adjacent to a pixel group may exist at both ends of the pixel group. Referring to FIG. 12 , an available pixel b may be adjacent to an upper end of the left pixel group, and an available pixel a may be adjacent to a lower end of the left pixel group. Here, the pixel a may be represented by P _Y [-1, nS+(nS>>1)], and the pixel b may be represented by P _Y [-1, nS-1].

In this case, two available pixels located adjacent to the left pixel group in the first left pixel line may exist. Accordingly, the encoder and the decoder may fill the pixel average values of the available pixels (a, b) located adjacent to the left pixel group into the positions of the unavailable pixels in the left pixel group. This padding process may be expressed by Equation 24 below as an example.

[Equation 24]

P _Y [x,y] = (P _Y [-1,nS-1] + P _Y [-1,nS+(nS>>1)] + 1) >> 1,

*271(x=-1, y=nS~nS+(nS>>1)-1)

By the above-described padding process, pixel values of the neighboring luma pixels on the first left pixel line and the neighboring luma pixels on the first upper pixel line may be determined. Based on the pixel values of the neighboring luma pixels on the first left pixel line, the neighboring luma pixels on the first upper pixel line, and the pixels in the current luma block 1210 , the encoder and the decoder determine the luma reference information (luma reference block 1230 ). ) and peripheral luma reference pixels 1240 ). In the embodiment of FIG. 12 , the luma reference information generation process may be expressed by Equation 25 below as an example.

[Equation 25]

P _Y '[x,y] = (P _Y [2x-1,y] + 2*P _Y [2x,y] + P _Y [2x+1,y] + 2) >> 2, (x=0 ~nS-1, y=-1)

P _Y '[x,y] =(P _RecY [2x,2y] + P _RecY [2x,2y+1]) >> 1, (x=0~nS-1, y=0~nS-1)

P _Y '[x,y] = (P _Y [x,2y] + P _Y [x,2y+1]) >> 1, (x=-1, y=0~nS-1)

Here, P _RecY may represent a pixel value of a pixel in the current luma block. In addition, P _Y ′ may indicate a pixel value corresponding to luma reference information (eg, a pixel value of a pixel in a luma reference block and/or a pixel value of a neighboring luma reference pixel).

Referring to the above description, in the embodiment of FIG. 12 , pixels on the second left pixel line are not used to generate luma reference information, but peripheral luma pixels on the first left pixel line and peripheral luma pixels on the first upper pixel line Only pixels and pixels in the current luma block 1210 may be used to generate luma reference information (luma reference block 1230 and surrounding luma reference pixels 1240 ).

Meanwhile, in FIGS. 5 to 12 , embodiments in which the first color component is the luma component and the second color component is the chroma component are described, but the present invention is not limited thereto. As mentioned above, the present invention can be applied in the same or similar manner to all possible combinations of first and second color components.

6 to 12 described above, embodiments in which the size of the first color component block (eg, the luma block) is greater than the size of the second color component block (eg, the chroma block) have been described, but the first color The component block may have a smaller size than the second color component block or may have the same size.

Even in this case, the encoder and the decoder may derive the first color component reference information to be used for prediction of the second color component block based on the pixels in the first color component block and pixels around the first color component block.

Here, the first color component reference information may mean pixel values of the first color component to be used for prediction of the second color component block. Accordingly, in this specification, the first color component information may be referred to as a 'first color component pixel'. In this case, the first color component pixels corresponding to the pixels in the second color component block may constitute a 'first color component reference block'. In addition, in the present specification, a pixel located around the first color component block is referred to as a 'peripheral first color component pixel', and a pixel located near the second color component block is referred to as a 'peripheral second color component pixel', A pixel positioned around the one color component reference block is referred to as a 'peripheral first color component reference pixel'. In this case, the first color component reference block may correspond to the second color component block, and the neighboring first color component reference pixel may correspond to the neighboring second color component pixel. In this specification, the first color component reference information may be used as a concept including a pixel in the first color component reference block and a neighboring first color component reference pixel.

A process of deriving first color component reference information to be used for prediction of the second color component block based on the first color component block when the first color component block has a size smaller than or equal to that of the second color component block may be different from the above-described embodiment of FIGS. 6 to 12 .

13 is a diagram schematically illustrating an embodiment of a process of deriving first color component reference information to be used for prediction of a second color component block when the size of the first color component block is smaller than the size of the second color component block .

1310 of FIG. 13 shows an embodiment of a first color component block 1312 , peripheral first color component pixels 1314 , a second color component block 1316 , and peripheral second color component pixels 1318 . . Referring to 1310 of FIG. 13 , when the size of the second color component block 1316 is 8x8, the size of the corresponding first color component block 1312 may be 4x4. In this case, the encoder and the decoder may generate first color component reference pixels corresponding to the 8x8 first color component reference block.

1320 of FIG. 13 illustrates an embodiment of a method for deriving first color component reference information.

In the embodiment of 1320 of FIG. 13 , as in 1310 of FIG. 13 , first color component reference information when the size of the first color component block 1312 is 4x4 and the size of the second color component block 1316 is 8x8 The derivation method is shown.

Referring to 1320 of FIG. 13 , the encoder and the decoder are configured to use a first color component block 1316 to be predicted based on a pixel in the first color component block 1312 and surrounding first color component pixels 1314 . A one color component reference block 1326 and peripheral first color component reference pixels 1328 may be generated. Here, the peripheral first color component pixels 1314 used for deriving the first color component reference information are, for example, pixels included in one pixel line located closest to the left side of the first color component block 1312; and pixels included in one pixel line located closest to the upper end of the first color component block 1312 .

Meanwhile, in 1320 of FIG. 13 , since the sizes of the first color component block 1312 and the second color component block 1316 are different from each other, the encoder and the decoder have the same size (8x8) as the second color component block 1316 . A first color component reference block 1326 and corresponding neighboring first color component reference pixels 1328 may be generated. In this case, since the size of the first color component block 1312 is smaller than the size of the second color component block 1316 , the encoder and the decoder are configured to include pixels in the first color component block 1312 and surrounding first color component pixels. The first color component reference information may be generated by performing up-sampling on the elements 1314 .

When unavailable pixels exist among the neighboring first color component pixels 1314 , the encoder and the decoder may perform upsampling after performing a padding process on the unavailable pixels. Since the padding process is similar to that in the above-described embodiments, a detailed description thereof will be omitted herein.

A process of performing up-sampling on the pixels in the first color component block 1312 and the surrounding first color component pixels 1314 may be expressed by Equation 26 below, for example.

[Equation 26]

*293p'[2x,2y] = p[x,y]

p'[2x+1,2y] = (p[x,y] + p[x+1,y])>>1

p'[2x,2y+1] = (p[x,y] + p[x,y+1])>>1

p'[2x+1,2y+1] = (p[x+1,y] + p[x,y+1])>>1

Here, p[x,y] may represent pixel values of the reconstructed first color component pixels, that is, pixels in the first color component block 1312 and surrounding first color component pixels 1314 . Also, p'[x,y] may represent pixel values of the first color component reference pixels 1326 and 1328 generated according to the size of the second color component block 1316 .

14 is a diagram schematically illustrating an embodiment of a process of deriving first color component reference information to be used for prediction of a second color component block when the size of the first color component block is the same as the size of the second color component block .

As described above, when the 4:4:4 image format is used, that is, when the color space of the image is RGB 4:4:4 or YUV 4:4:4, the size of the luma block and the chroma block may be the same. have. Accordingly, in this case, if the first color component block is a luma block and the second color component block is a chroma block, the size of the first color component block and the size of the second color component block may be the same. Also, when the first color component block is a chroma block corresponding to U and the second color component block is a chroma block corresponding to V, the size of the first color component block and the size of the second color component block may be the same. . In the embodiment of FIG. 14 , it is assumed that the size of the first color component block 1412 is 8x8 and the size of the second color component block 1422 is 8x8.

Referring to FIG. 14 , the encoder and the decoder use a first color component reference block to be used for prediction of a second color component block based on a pixel in the first color component block 1412 and surrounding first color component pixels 1414 . 1422 and surrounding first color component reference pixels 1424 may be generated. The generated first color component reference block 1422 may exist at a position corresponding to the second color component block. Here, the peripheral first color component pixels 1414 used for deriving the first color component reference information are, for example, pixels included in one pixel line located closest to the left side of the first color component block 1412; and pixels included in one pixel line located closest to the upper end of the first color component block 1412 .

In this case, since the size of the first color component block 1412 is the same as the size of the second color component block, the encoder and the decoder perform down sampling and sub sampling when generating the first color component information. and up-sampling may not be performed. Accordingly, the pixel values of the pixels in the first color component block 1412 and the pixel values of the surrounding first color component pixels 1414 may be used as the first color component reference information as they are.

Also, in the intra mode, a lossless encoding/decoding method may be applied instead of a lossy encoding/decoding method using transform/quantization. In the lossless encoding/decoding method, a process such as transform quantization may not be applied to the encoding object block. When the size of the first color component block 1412 is the same as the size of the second color component block, even when a lossless encoding/decoding method is applied, the first color component reference information may be derived as in the above-described embodiment. can In this case, the encoder and the decoder may not perform down-sampling, sub-sampling, and up-sampling when generating the first color component information. The generated first color component reference block 1422 may exist at a position corresponding to the second color component block.

Meanwhile, in the above-described embodiments, both upper peripheral luma pixels positioned at the top of the current luma block and left peripheral luma pixels positioned to the left of the current luma block may be used to generate luma reference information. However, the present invention is not limited thereto, and neighboring luma pixels used for generating luma reference information may be determined in various ways. For example, only the upper peripheral luma pixels may be used to generate the luma reference information, and only the left peripheral luma pixels may be used to generate the luma reference information. Also, neighboring luma pixels used to generate luma reference information may be determined according to the intra prediction mode of the current luma block.

15 is a diagram schematically illustrating an embodiment of neighboring luma pixels used to generate luma reference information.

*305 In the embodiment of FIG. 15 , it is assumed that the size of the current luma block 1512 is 8x8 and the size of the corresponding current chroma block is 4x4. As an example, FIG. 15 may correspond to an embodiment in which an image format of YUV 4:2:0 is used. Also, it is assumed that pixel values are already filled in the peripheral luma pixels A, B, C, D, E, and F shown in FIG. 15 through the above-described padding process. And, in FIG. 15 , it is assumed that the mode value of the intra prediction mode as in 410 of FIG. 4 is used.

In the embodiment of FIG. 15 , based on at least one of the neighboring luma pixels A, B, C, D, E, and F and pixels within the current luma block 1512, a luma reference block ( 1522 , 1532 or 1542 ) and peripheral luma reference pixels 1524 , 1534 or 1544 may be generated. In this case, since the size of the current chroma block is 4x4, the size of the generated luma reference block 1522, 1532, or 1542 may be 4x4.

In the embodiment of FIG. 15 , the neighboring luma pixels to be used for generating the luma reference information may be obtained by, for example, at least one of the combinations shown in Table 3 below.

[Table 3]

In the embodiment of Table 3, P may mean that left peripheral luma pixels located on the left side of the current luma block 1512 and upper peripheral luma pixels located on the upper side of the current luma block 1512 are used to generate luma reference information. . Also, Q may mean that upper peripheral luma pixels located at the upper end of the current luma block 1512 are used to generate luma reference information. And, R may mean that the left peripheral luma pixels located on the left side of the current luma block 1512 are used to generate luma reference information.

In addition, in the embodiment of Table 3, L indicates that when the left peripheral luma pixels are used to generate the luma reference information, pixels on the first left pixel line closest to the left of the current luma block 1512 are used to generate the luma reference information can mean And, M may mean that when the left neighboring luma pixels are used to generate the luma reference information, pixels on the second left pixel line located second to the left of the current luma block 1512 are used to generate the luma reference information. have. In this case, pixels on the first left pixel line may be used together for sub-sampling.

Also, N may mean that neighboring luma pixels to be used for generating luma reference information are determined based on the intra prediction mode of the current luma block 1512 . In this case, the luma reference information may be adaptively generated based on the intra prediction mode of the current luma block 1512 . When the luma reference information is generated based on the intra prediction mode of the current luma block 1512 , the left peripheral luma pixels used may be, for example, pixels on the first left pixel line, or pixels on the second left pixel line as another example. may be

For example, in an embodiment to which a combination of P and L is applied, upper peripheral luma pixels located at the upper end of the current luma block 1512 and left peripheral luma pixels on the first left pixel line may be used to generate luma reference information. . Accordingly, in this case, peripheral luma pixels of [A, D] or [A, B, D, F] shown in 1510 of FIG. 15 may be used to generate luma reference information.

When the luma reference information is generated based on [A, D], the generated luma reference information may be shown, for example, as shown in 1520 of FIG. 15 . In 1520 of FIG. 15 , the neighboring luma reference pixels 1524 may include pixels adjacent to an upper end of the luma reference block 1522 and pixels adjacent to the left side of the luma reference block 1522 .

As another example, in an embodiment to which a combination of Q and L is applied, upper peripheral luma pixels located at the upper end of the current luma block 1512 may be used to generate luma reference information. Accordingly, in this case, peripheral luma pixels of [A, B] shown in 1510 of FIG. 15 may be used to generate luma reference information.

When the luma reference information is generated based on [A, B], the generated luma reference information may be shown, for example, as shown in 1530 of FIG. 15 . In 1530 of FIG. 15 , the peripheral luma reference pixels 1534 may be pixels located at the upper end of the luma reference block 1532 .

As another example, in an embodiment to which a combination of R and L is applied, left peripheral luma pixels on the first left pixel line may be used to generate luma reference information. Accordingly, in this case, the neighboring luma pixels of [D, F] shown in 1510 of FIG. 15 may be used to generate the luma reference information.

When the luma reference information is generated based on [D, F], the generated luma reference information may be illustrated, for example, as in 1540 of FIG. 15 . In 1540 of FIG. 15 , the peripheral luma reference pixels 1544 may be pixels located on the left side of the luma reference block 1542 .

As another example, in an embodiment to which a combination of P and M is applied, upper peripheral luma pixels located at the upper end of the current luma block 1512 and left peripheral luma pixels on the second left pixel line may be used to generate luma reference information. have. Accordingly, in this case, the surrounding luma pixels of [A, C] or [A, B, C, E] shown in 1510 of FIG. 15 may be used to generate the luma reference information. The luma reference information generated when the combination of P and M is applied may be illustrated, for example, as in 1520 of FIG. 15 .

As another example, in an embodiment to which a combination of Q and M is applied, upper peripheral luma pixels located at the upper end of the current luma block 1512 may be used to generate luma reference information. Accordingly, in this case, peripheral luma pixels of [A, B] shown in 1510 of FIG. 15 may be used to generate luma reference information. Luma reference information generated when a combination of Q and M is applied may be illustrated, for example, as in 1530 of FIG. 15 .

As another example, in an embodiment to which a combination of R and M is applied, left peripheral luma pixels on the second left pixel line may be used to generate luma reference information. Accordingly, in this case, peripheral luma pixels of [C, E] shown in 1510 of FIG. 15 may be used to generate luma reference information. The luma reference information generated when the combination of R and M is applied may be illustrated, for example, as in 1540 of FIG. 15 .

As another example, in an embodiment to which a combination of P and N is applied, based on the intra prediction mode of the current luma block 1512 , upper peripheral luma pixels located at the upper end of the current luma block 1512 and the current luma block 1512 ), left neighboring luma pixels located on the left side may be determined as neighboring luma pixels to be used for generating luma reference information.

In this case, the intra prediction mode of the current luma block 1512 may be a first intra prediction mode that uses both pixels located at the top and the left of the current luma block 1512 . As an example, the first intra prediction mode is a DC mode, a planar mode, or a mode value of 4, 19, 11, 20, 5, 21, 12, 22, 27, 15, 28, 8, 29, 16 or 30. can be a mode.

The luma reference information generated when a combination of P and N is applied may be illustrated, for example, as in 1520 of FIG. 15 .

As another example, in an embodiment to which a combination of Q and N is applied, based on the intra prediction mode of the current luma block 1512 , upper peripheral luma pixels located at the upper end of the current luma block 1512 are used to generate luma reference information It may be determined by the surrounding luma pixels to be

In this case, the intra-prediction mode of the current luma block 1512 may be a second intra-prediction mode that mainly uses a pixel located at an upper end of the current luma block 1512 . For example, the second intra prediction mode may be a mode having a mode value of 1, 23, 13, 24, 6, 25, 14, 26, or 7.

Luma reference information generated when a combination of Q and N is applied may be illustrated, for example, as in 1530 of FIG. 15 .

As another example, in an embodiment to which a combination of R and N is applied, based on the intra prediction mode of the current luma block 1512 , left neighboring luma pixels located on the left side of the current luma block 1512 are used to generate luma reference information It may be determined by the surrounding luma pixels to be

In this case, the intra prediction mode of the current luma block 1512 may be a third intra prediction mode mainly using a pixel located to the left of the current luma block 1512 . For example, the third intra prediction mode may be a mode having a mode value of 2, 31, 17, 32, 9, 33, 18, 34 or 10.

The luma reference information generated when a combination of R and N is applied may be illustrated, for example, as in 1540 of FIG. 15 .

When the pixel values of the neighboring luma reference pixels are derived based on the determined neighboring luma pixels according to the above-described embodiment, the prediction parameter may be derived using the neighboring chroma pixels located at the same position as each of the neighboring luma reference pixels. . A specific embodiment of the prediction parameter derivation process will be described later.

16 is a diagram for explaining an embodiment of a process of deriving a prediction parameter and a process of deriving prediction values of pixels in a current chroma block.

16 illustrates a luma reference block 1610 , neighboring luma reference pixels 1620 , a current chroma block 1630 , and reconstructed neighboring chroma pixels 1640 . In the embodiment of FIG. 16 , the size of the luma reference block 1610 and the size of the current chroma block 1630 may be 8x8, respectively.

As an embodiment, the encoder and the decoder may derive prediction parameters for the current chroma block 1630 based on the luma reference information and the chroma component information. As an example, the encoder and the decoder may derive a prediction parameter based on pixel values of the neighboring luma reference pixels 1620 and pixel values of the neighboring chroma pixels 1640 . Here, the prediction parameters may be alpha (α) and beta (β) based on the least squares method.

A process of deriving a prediction parameter based on the pixel values of the neighboring luma reference pixels 1620 and the neighboring chroma pixels 1640 may be represented by Equation 27 below, for example.

[Equation 27]

In Equation 27, p1 may indicate a pixel value of the neighboring luma reference pixels 1620 , and p2 may indicate a pixel value of the neighboring chroma pixels 1640 . In addition, nS may indicate the horizontal and vertical lengths of the luma reference block 1610 and the current chroma block 1630 , and 'Bitdepth' may indicate the bit depth of a chroma component pixel. And, abs(x) may represent the absolute value of x. 'lmDiv' is a variable determined according to the value of 'a2s', and may be determined, for example, according to Table 4 below.

[Table 4]

In Equation 27, a may represent alpha (α), and b may represent beta (β).

In the above-described prediction parameter deriving process, an unavailable pixel may exist among the neighboring luma reference pixels 1620 and/or the neighboring chroma pixels 1640 . In this case, for example, the encoder and the decoder may derive the prediction parameter based on only available pixels among the neighboring luma reference pixels 1620 and/or the neighboring chroma pixels 1640 .

In another embodiment, the encoder and the decoder do not derive a prediction parameter based on the neighboring luma reference pixels 1620 and the neighboring chroma pixels 1640 , but use a default prediction parameter as the current chroma block 1630 . It can also be used for prediction of For example, alpha (α) used as a basic prediction parameter may be 1, and beta (β) used as a basic prediction parameter may be 0.

When the prediction parameters are derived according to the above-described embodiments, the encoder and the decoder may derive the prediction values of the pixels in the current chroma block 1630 based on the prediction parameter and the pixel values of the pixels in the luma reference block 1610 . have. As an example, the encoder and the decoder based on the prediction parameters 'a', 'k', and 'b' derived by Equation 27 and the pixel values of the pixels in the luma reference block 1610 shown in FIG. 16 , the current chroma block A prediction value of pixels within 1630 may be derived. In this case, a process of deriving prediction values of pixels in the current chroma block 1630 may be expressed by Equation 28 below, for example.

[Equation 28]

predSamples[x,y] = Clip1(((p1[x,y]*a)>>k)+b), (x,y=0~(nS-1))

Clip1(i) = Clip3(0,(1<<BitDepth)-1,i)

Here, 'predSamples[x,y]' may indicate a predicted value of a pixel in the current chroma block 1630 . Also, clip(x,y,z) can be x if z is less than x, y if z is greater than y, and z otherwise.

17 is a flowchart schematically illustrating an embodiment of a decoding process according to the present invention.

In the embodiment of FIG. 17 , P _Y may represent a pixel value of a neighboring luma pixel used for prediction of a current luma block. Also, N _Y may indicate a horizontal length and a vertical length of the current luma block, and P _LM may indicate a pixel value of a neighboring luma pixel used to generate luma reference information in the LM mode. Here, P _Y [x,y] and P _LM [x,y] may have the same value.

The embodiment of FIG. 17 is described based on the luma component (Y). However, the present invention is not limited thereto, and the embodiment of FIG. 17 may be applied to the chroma components (U, V) other than the luma component (Y) in the same or similar manner to the luma component. Also, in this case, PC _C may be used instead of the variable P _Y , N _C may be used instead of the variable N _Y , and the value assigned to cIdx may be 1 or 2 instead of 0. Here, cIdx may be an index indicating which component among Y, U, and V the color component of the current block corresponds. For example, if the color component of the current block is Y, 0 may be allocated to cIdx, if the color component of the current block is U, 1 may be allocated to cIdx, and if the color component of the current block is V, cIdx may be allocated to cIdx. 2 may be assigned. A relationship between variables applied to the embodiment of FIG. 17 according to color components may be represented by Table 5 below as an example.

[Table 5]

Here, P _C may represent a pixel value of a neighboring chroma pixel located in the periphery of the current chroma block. Also, N _C may indicate the horizontal length and vertical length of the current chroma block.

For example, the embodiment of FIG. 17 may be applied in the order of a Y component, a U component, and a V component to each color component constituting the color space.

Referring to FIG. 17, the encoder and the decoder are P _Y [x,y], (x=-1, y=-1~N _Y *2-1 and x=0~N _Y *2-1, y=- It is possible to mark by checking whether 1) is available (S1710). Here, the marking may refer to a process of storing or displaying information on availability. Since the embodiments in which the peripheral luma pixels are not available have been described above in the embodiment of FIG. 8 , a detailed description thereof will be omitted herein.

Referring back to FIG. 17 , the encoder and the decoder may determine whether the conditions A and B are satisfied for the current luma block ( S1720 ).

Here, condition A is a condition indicating that the LM mode can be applied to intra prediction of the current chroma block corresponding to the current luma block. As an example, condition A may be satisfied when 1 is assigned to the LM flag (chroma_pred_from_luma_enabled_flag). As described above, chroma_pred_from_luma_enabled_flag may indicate whether the LM mode can be supported for intra prediction of the current chroma block. In addition, condition B is a condition indicating that the current block is a luma component block (Y). For example, condition B may be satisfied when 0 is assigned to cIdx.

When condition A and condition B are satisfied, the encoder and the decoder may generate a P _LM .

Referring back to FIG. 17, when conditions A and B are satisfied, the encoder and the decoder are P _LM [x,y], (x=-1, y=-1~N _Y *2-1 and x=0~ It can be marked by checking whether N _Y *2-1, y=-1) is available (S1730).

At this time, since P _Y [x,y] and P _LM [x,y] can have the same value, if P _Y [x,y] is available, P _LM [x,y] is also marked as available and P _LM [x,y] may also be marked as unavailable if P _Y [x,y] is not available.

In addition, as an example, after the marking process for P _LM is performed, the encoder and decoder are available among pixels corresponding to P _LM [x,y], (x=-1, y=0~N _Y *2-1). It is also possible to replace the pixel value of the pixels determined to be one with the luma pixel value at the [x-1,y] position.

Referring back to FIG. 17 , the encoder and the decoder may determine whether condition C is satisfied with respect to neighboring luma pixels to be used for prediction of the current luma block ( S1740 ). Here, condition C is at least one of P _Y [x,y], (x=-1, y=-1~N _Y *2-1 and x=0~N _Y *2-1, y=-1) A condition indicating that the pixel of ' is not available.

When condition C is satisfied, that is, when at least one pixel among neighboring luma pixels to be used for prediction of the current luma block is not available, the encoder and the decoder perform P _Y [x,y], (x=-1, y= A padding process may be performed on pixels that are not available among -1 to N _Y *2-1 and x=0 to N _Y *2-1, y=-1) (S1750). Since the embodiments of the padding process have been described above, a detailed description thereof will be omitted herein.

Referring back to FIG. 17 , the encoder and the decoder may determine whether conditions A, B, and D are satisfied with respect to neighboring luma pixels to be used for generating luma reference information ( S1760 ). Here, condition D is at least one of P _LM [x,y], (x=-1, y=-1~N _Y *2-1 and x=0~N _Y *2-1, y=-1) A condition indicating that the pixel of ' is not available.

When conditions A, B and D are satisfied, that is, when at least one pixel among the surrounding luma pixels to be used for generating luma reference information is not available, the encoder and the decoder perform P _LM [x,y], (x=-1 , y=-1 to N _Y *2-1 and x=0 to N _Y *2-1, y=-1), a padding process may be performed on pixels that are not available (S1770). Since the embodiments of the padding process have been described above, a detailed description thereof will be omitted herein.

After the above-described padding process is performed, the encoder and the decoder may generate a reconstructed block for the current luma block by performing prediction and decoding processes on the current luma block based on the neighboring luma pixels ( S1780 ). When the embodiment of FIG. 17 is applied to a chroma component (U or V), the encoder and the decoder generate a prediction block for the current chroma block based on neighboring luma pixels, pixels in the reconstructed current luma block, and neighboring chroma pixels. and may generate a reconstructed block for the current chroma block based on the generated prediction block.

FIG. 18 is a diagram for explaining an embodiment of a process of deriving luma reference information and prediction parameters for a current chroma block when a decoding process is performed as in the embodiment of FIG. 17 .

18 shows a luma component block 1810 that is a prediction target, neighboring luma pixels 1820, P _Y used for prediction of the luma component block, a reconstructed current luma block 1830, P _RecY , and luma reference information. It shows the used peripheral luma pixels 1840 , P _LM , the current chroma block 1850 and the peripheral chroma pixels 1860 , P _C . 18, the size of the luma component block 1810 to be predicted and the reconstructed current luma block 1830, P _RecY may be 8x8, and the size of the current chroma block 1850 may be 4x4.

As described above, the embodiment of FIG. 17 may be applied to the chroma components (U, V) other than the luma component (Y). When the prediction mode of the current chroma block 1850 is the LM mode, after the padding process is performed, the encoder and the decoder may derive luma reference information and prediction parameters for the current chroma block 1850 .

The encoder and decoder may derive prediction values of pixels in the luma component block 1810 to be predicted based on the neighboring luma pixels 1820 and P _Y . In this case, the encoder and the decoder may generate a reconstructed block of a luma component based on the derived prediction value, which may correspond to the reconstructed current luma block 1830 (P _RecY ). In addition, the encoder and the decoder may determine the neighboring luma pixels 1840 and P _LM to be used for generating the luma reference information based on the neighboring luma pixels 1820 and P _Y used for prediction of the luma component block 1810 . have.

The encoder and decoder may generate luma reference information to be used for prediction of the current chroma block 1850 based on the neighboring luma pixels 1840 , P _LM and the reconstructed current luma block 1830 , P _RecY . In FIG. 18 , since the sizes of the reconstructed current luma block 1830 and the current chroma block 1850 are different from each other, the encoder and the decoder have the same size (4x4) as the current chroma block 1850 and the luma reference block and its corresponding surroundings. Luma reference pixels may be generated. At this time, since the size of the current chroma block 1850 is smaller than the size of the reconstructed current luma block 1830 , the encoder and the decoder are used in pixels within the reconstructed current luma block 1830 and adjacent luma pixels 1840 . The luma reference information may be generated by performing down-sampling on the .

For example, a downsampling process may be performed on neighboring luma pixels located at the upper end of the restored current luma block 1830 by Equation 29 below.

[Equation 29]

P _Y '[x,-1] = (P _LM [2x-1,-1] + 2*P _LM [2x,-1] + P _LM [2x+1,-1] + 2) >> 2,

(x = 0~N _C -1)

Here, P _Y ′ may indicate a pixel value of a pixel corresponding to the luma reference information (eg, a pixel in a luma reference block or a neighboring luma reference pixel). Also, P _LM may indicate a pixel value of a neighboring luma pixel 1840 to be used for generating luma reference information. And, N _C may represent the horizontal length and vertical length of the current chroma block 1850 .

For example, a down-sampling process may be performed on neighboring luma pixels located on the left side of the restored current luma block 1830 by Equation 30 below.

[Equation 30]

P _Y '[-1,y] = (P _LM [-1,2y] + P _LM [-1,2y+1]) >> 1, (y = 0~N _C -1)

For pixels in the restored current luma block 1830 , for example, a downsampling process may be performed according to Equation 31 below. A luma reference block corresponding to the current chroma block 1850 may be generated by Equation 31 below.

[Equation 31]

P _Y '[x,y] = (P _RecY [2x,2y] + P _RecY [2x,2y+1]) >> 1, (x,y = 0~N _C -1)

Here, P _RecY may represent a pixel value of a pixel in the current luma block 1830 .

When the luma reference information is generated by the above-described process, the encoder and the decoder may derive prediction parameters for the current chroma block 1850 based on the luma reference information and the chroma component information. As an example, the encoder and the decoder may derive a prediction parameter based on pixel values of the neighboring luma reference pixels 1840 and pixel values of the neighboring chroma pixels 1860 . Equation 32 below may represent a part of a process of deriving a prediction parameter.

[Equation 32]

Here, P _C may indicate a pixel value of a neighboring chroma pixel 1860 to be used for generating luma reference information, and 'Bitdepth' may indicate a bit depth of a chroma component pixel.

Implementation of the process of deriving k2, a1, a2, k1, a1s, a2s, a3, α (alpha), k and β (beta) based on k3, L, C, LL, and LC derived by Equation 32 Since the example is similar to the embodiment of Equation 27, a detailed description thereof will be omitted herein. However, in this case, in the embodiment of Equation 27, nS may be replaced with N _C , p1 may be replaced with P _Y ′, and p2 may be replaced with PC _C .

When the prediction parameter is derived, the encoder and the decoder may derive prediction values of the pixels in the current chroma block 1850 based on the prediction parameter and the pixel values of the pixels in the luma reference block 1830 . Since the embodiment of the process of deriving the prediction values of the pixels in the current chroma block 1850 is similar to the embodiment of Equation 28, a detailed description thereof will be omitted herein. However, in this case, in the embodiment of Equation 28, nS may be replaced with N _C and p1 may be replaced with P _RecY .

19A and 19B are diagrams for explaining another embodiment of a decoding process according to the present invention.

19A shows an embodiment of a decoding process according to the present invention. Also, FIG. 19B shows that one CU 1990 is divided into a plurality of TUs based on a quad-tree structure.

19A and 19B (hereinafter referred to as FIG. 19 ), P _k may be P _Y , P _U or P _V , and N _k may be N _Y , N _U or N _V . Here, P _Y may represent a pixel value of a neighboring luma pixel used for prediction of the current luma block. In addition, P _U and P _V may represent pixel values of neighboring chroma pixels used for prediction of the current chroma block. Here, P _U and P _V may be represented by PC _C . In addition, N _Y may represent the horizontal length and vertical length of the current luma block, and N _U and N _V may represent the horizontal length and vertical length of the current chroma block. Here, N _U and N _V may be represented by N _C . In the embodiment of FIG. 19 , P _LM may indicate a pixel value of a neighboring luma pixel used to generate luma reference information in the LM mode. Here, P _Y [x,y] and P _LM [x,y] may have the same value.

In the embodiment of FIG. 19, k may be Y, U, or V, so the embodiment of FIG. 19 may be applied to both the luma component (Y) and the chroma component (U, V). A relationship between variables applied to the embodiment of FIG. 19 according to color components may be shown in Table 6 below as an example.

[Table 6]

Here, cIdx may be an index indicating which color component of the current block corresponds to among Y, U(Cb), and V(Cr). For example, if the color component of the current block is Y, 0 may be assigned to cIdx, if the color component of the current block is U(Cb), 1 may be assigned to cIdx, and the color component of the current block may be V( Cr), 2 may be assigned to cIdx.

Also, the embodiment of FIG. 19 may be applied in the order of a Y component, a U(Cb) component, and a V(Cr) component to each color component constituting a color space in one CU, for example.

Referring to FIG. 19B , one CU 1990 may be divided into a plurality of TUs 1, 2, 3, 4, 5, 6, 7 based on a quad tree structure. In this case, the encoder and the decoder may first perform the decoding process according to the embodiment of FIG. 19A on Y component blocks in the CU 1990 . In addition, the encoder and the decoder may sequentially perform the decoding process according to the embodiment of FIG. 19A on the U component blocks and the V component blocks in the CU 1990 . For example, the decoding process according to the embodiment of FIG. 19 is [Y1->Y2->...->Y7->U1->U2->...->U7->V1->V2...- >V7]. Here, as an example, Y1 may mean that decoding is performed on the luma component of block 1 in FIG. 19B .

Referring to FIG. 19A , the encoder and the decoder are P _k [x,y], (x=-1, y=-1~N _k *2-1 and x=0~N _k *2-1, y=- It is possible to mark by checking whether 1) is available (S1910). As described above, the marking may refer to a process of storing or displaying information on availability. Since the embodiments in which the peripheral luma pixels are not available have been described above in the embodiment of FIG. 8 , a detailed description thereof will be omitted herein.

Referring back to FIG. 19A , the encoder and the decoder may determine whether condition A, condition B, and condition C are satisfied ( S1920 ).

Here, condition A is a condition indicating that the LM mode can be applied to intra prediction of the current chroma block. As an example, condition A may be satisfied when 1 is assigned to the LM flag (chroma_pred_from_luma_enabled_flag). As described above, chroma_pred_from_luma_enabled_flag may indicate whether the LM mode can be supported for intra prediction of the current chroma block. In addition, condition B is a condition indicating that the intra prediction mode of the current block is the LM mode. Here, as an example, the LM mode may be represented by Intra_FromLuma. In addition, condition C is a condition indicating that the current block is a chroma component block (U(Cb) or V(Cr)). For example, condition C may be satisfied when 1 or 2 is assigned to cIdx.

When condition A, condition B, and condition C are satisfied, the encoder and the decoder may generate a P _LM .

Referring back to FIG. 19A, when condition A, condition B, and condition C are satisfied, the encoder and the decoder are P _LM [x,y], (x=-1, y=-1~N _Y *2-1 and x =0~N _Y *2-1, y=-1) may be checked by checking whether it is available and marked (S1930).

In this case, as an example, after the marking process on the P _LM is performed, the encoder and the decoder set the pixels corresponding to P _LM [x,y], (x=-1, y=0~N _Y *2-1). The pixel values of the pixels determined to be available among them may be replaced with the pixel value of the [x-1,y] position, that is, the pixel value of P _LM [x-1,y].

Also at this time, according to an embodiment, the encoder and the decoder may check whether the P _LM is available based on the availability of the neighboring chroma pixels corresponding to the neighboring luma pixels, that is, _PU or P _V . At this time, whether _PU or P _V is available may be already marked in step S1910. For example, the encoder and the protector may determine whether the P _LM is available based on whether a pixel corresponding to P _U is available.

For example, if the surrounding _chroma pixel P _U [x,y] is marked as unavailable for x=-1, y=0~N _U *2-1, the encoder and decoder ,2y] and P _LM [x,2y+1] may also be determined to be unavailable. On the other hand, if the surrounding chroma pixel P _U [x,y] is marked as available for x=-1, y=0~N _U *2-1, the encoder and the decoder are the neighboring luma pixels P _LM [x,2y] and P _LM [x,2y+1] may also be determined to be available.

In this case, the encoder and the decoder allocate the luma pixel values located in [x-1,2y] and [x-1,2y+1] to P _LM [x,2y] and P _LM [x,2y+1], respectively. can do. Accordingly, when the neighboring luma pixel is a pixel on the left pixel line located adjacent to the left side of the current luma block and it is determined that the neighboring luma pixel is available, the pixel value of the luma pixel located adjacent to the left side of the neighboring luma pixel is the neighboring luma pixel. It may be determined as a pixel value of a luma pixel.

In addition, if the surrounding _chroma pixel P _U [x,y] is marked as unavailable for x=0~N _U *2-1, y=-1, the encoder and the decoder ] and P _LM [2x+1,y] can also be determined to be unavailable. On the other hand, if the surrounding chroma pixel P _U [x,y] is marked as available for x=0~N _U *2-1, y=-1, the encoder and the decoder are the neighboring luma pixels P _LM [2x,y] and P _LM [2x+1,y] may also be determined to be available.

In this case, the encoder and the decoder may allocate the luma pixel values located at [2x,y] and [2x+1,y] to P _LM [2x,y] and P _LM [2x+1,y], respectively. Accordingly, when the neighboring luma pixel is a pixel on the upper pixel line located adjacent to the top of the current luma block and it is determined that the neighboring luma pixel is available, the pixel value of the luma pixel at the same position as the neighboring luma pixel is the neighboring luma pixel may be determined by the pixel value of .

Also, if the surrounding chroma pixel P _U [x,y] is marked as unavailable for x=-1, y=-1, the encoder and decoder assume that the surrounding luma pixel P _LM [x,y] is also not available. can judge On the other hand, if the surrounding chroma pixel P _U [x,y] is marked as available for x=-1 and y=-1, the encoder and decoder can determine that the surrounding luma pixel P _LM [x,y] is also available. can

In this case, the encoder and the decoder may allocate the luma pixel value located at [x,y] to P _LM [x,y]. Accordingly, when the position of the neighboring luma pixel is [-1,-1] and it is determined that the neighboring luma pixel is available, the pixel value of the luma pixel at the same position as the neighboring luma pixel is determined as the pixel value of the neighboring luma pixel. can

In the above-described embodiment, a process of checking whether P _LM is available based on _PU has been described, but the present invention is not limited thereto. For example, whether P _LM is available may be checked based on P _V .

Referring back to FIG. 19A , the encoder and the decoder may determine whether the condition D is satisfied ( S1940 ). Here, the condition D is at least one of P _k [x,y], (x=-1, y=-1~N _k *2-1 and x=0~N _k *2-1, y=-1) A condition indicating that the pixel of ' is not available.

When condition D is satisfied, the encoder and the decoder perform P _k [x,y], (x=-1, y=-1~N _k *2-1 and x=0~N _k *2-1, y= A padding process may be performed on pixels that are not available among -1) (S1950). Since the embodiments of the padding process have been described above, a detailed description thereof will be omitted herein.

Referring back to FIG. 19A , the encoder and the decoder may determine whether condition A, condition B, condition C, and condition E are satisfied ( S1960 ). Here, the condition E is at least one of P _LM [x,y], (x=-1, y=-1~N _Y *2-1 and x=0~N _Y *2-1, y=-1) A condition indicating that the pixel of ' is not available.

When condition A, condition B, condition C, and condition E are satisfied, the encoder and decoder perform P _LM [x,y], (x=-1, y=-1~N _Y *2-1 and x=0~ A padding process may be performed on pixels that are not available among N _Y *2-1, y=-1) (S1970). Since specific embodiments of the padding process have been described above, a detailed description thereof will be omitted herein.

After the above-described padding process is performed, the encoder and the decoder perform prediction and decoding processes on the current luma block and/or the current chroma block to obtain a reconstructed block for the current luma block and/or a reconstructed block for the current chroma block. can be created (S1980).

In this case, LM mode prediction may be applied to the current chroma block. Since the prediction process in the LM mode is similar to that in the embodiment of FIG. 18 , a detailed description thereof will be omitted herein.

In the above-described embodiments, the methods are described on the basis of a flowchart as a series of steps or blocks, but the present invention is not limited to the order of steps, and some steps may occur in a different order or at the same time as other steps as described above. can In addition, those of ordinary skill in the art will recognize that the steps shown in the flowchart are not exclusive, other steps may be included, or that one or more steps of the flowchart may be deleted without affecting the scope of the present invention. You will understand.

The above-described embodiments include examples of various aspects. It is not possible to describe every possible combination for representing the various aspects, but one of ordinary skill in the art will recognize that other combinations are possible. Accordingly, it is intended that the present invention cover all other substitutions, modifications and variations falling within the scope of the following claims.

Claims (9)

Determining whether left peripheral luma pixels adjacent to the left of the luma reference block are available among the peripheral luma pixels of the luma reference block corresponding to the current chroma block and whether upper peripheral luma pixels adjacent to the top of the luma reference block are available to do;
determining peripheral luma reference pixels according to whether the left peripheral luma pixels are available and whether the top peripheral luma pixels are available;
determining neighboring chroma reference pixels corresponding to the neighboring luma reference pixels from among neighboring chroma pixels located around the current chroma block;
deriving a prediction parameter for the current chroma block based on the neighboring luma reference pixels and the neighboring chroma reference pixels;
deriving a prediction value of a pixel in the current chroma block based on the pixel in the luma reference block and the prediction parameter; and
based on the prediction value, reconstructing a pixel in the current chroma block;
The left peripheral luma pixels include a first left peripheral luma pixel immediately adjacent to the left side of the luma reference block and a second left peripheral luma pixel immediately adjacent to the left side of the first left peripheral luma pixel,
When the left peripheral luma pixels are available, a downsampled peripheral luma reference pixel determined based on the first left peripheral luma pixel and the second left peripheral luma pixel and a peripheral chroma corresponding to the downsampled peripheral luma reference pixel The image decoding method according to claim 1, wherein the prediction parameter for the current chroma block is derived based on a reference pixel.
The method of claim 1,
The determining of the peripheral luma reference pixels comprises:
when the left peripheral luma pixels are not available, the left peripheral luma pixels are not included in the peripheral luma reference pixels;
When the upper peripheral luma pixels are not available, the upper peripheral luma pixels are not included in the peripheral luma reference pixels.
The method of claim 1,
In the step of deriving the prediction parameter,
The peripheral luma pixels are included in pixel groups having a predetermined size,
For each of the pixel groups, a peripheral luma sample at a predetermined position among the peripheral luma pixels is selected;
the peripheral luma pixels are downsampled based on the peripheral luma sample of the selected predetermined position;
The image decoding method according to claim 1, wherein the prediction parameter is derived based on the downsampled neighboring luma pixels.
The method of claim 1,
The upper peripheral luma pixels include a first upper peripheral luma pixel immediately adjacent to an upper end of the luma reference block and a second upper peripheral luma pixel immediately adjacent to an upper end of the first upper peripheral luma pixel. Decryption method.
The method of claim 1,
The determining of the peripheral luma reference pixels comprises:
When the left peripheral luma pixels are available, the peripheral luma reference pixels are determined according to the left peripheral luma pixels and the lower left peripheral luma pixels.
The method of claim 1,
The determining of the peripheral luma reference pixels comprises:
If the upper peripheral luma pixels are available, the peripheral luma reference pixels are determined according to the upper peripheral luma pixels and the right peripheral luma pixel.
According to claim 1,
The video decoding method comprises:
obtaining LM flag information indicating whether an LM mode can be applied to intra prediction of a chroma block in a current sequence including the current chroma block; and
When the LM flag information indicates that the LM mode is applied to intra prediction of the chroma block in the current sequence, whether the LM mode is applied to the current chroma block according to intra prediction mode information indicating the intra prediction mode of the current chroma block Video decoding method further comprising the step of determining whether.
Determining whether left peripheral luma pixels adjacent to the left of the luma reference block are available among the peripheral luma pixels of the luma reference block corresponding to the current chroma block and whether upper peripheral luma pixels adjacent to the top of the luma reference block are available to do;
determining peripheral luma reference pixels according to whether the left peripheral luma pixels are available and whether the top peripheral luma pixels are available;
determining neighboring chroma reference pixels corresponding to the neighboring luma reference pixels from among neighboring chroma pixels located around the current chroma block;
deriving a prediction parameter for the current chroma block based on the neighboring luma reference pixels and the neighboring chroma reference pixels;
deriving a prediction value of a pixel in the current chroma block based on the pixel in the luma reference block and the prediction parameter; and
encoding a pixel in the current chroma block based on the prediction value;
The left peripheral luma pixels include a first left peripheral luma pixel immediately adjacent to the left side of the luma reference block and a second left peripheral luma pixel immediately adjacent to the left side of the first left peripheral luma pixel,
When the left peripheral luma pixels are available, a downsampled peripheral luma reference pixel determined based on the first left peripheral luma pixel and the second left peripheral luma pixel and a peripheral chroma corresponding to the downsampled peripheral luma reference pixel The video encoding method according to claim 1, wherein the prediction parameter for the current chroma block is derived based on a reference pixel.
A computer-readable recording medium storing a bitstream, comprising:
The bitstream includes image data decoded by an image decoding method,
The video decoding method comprises:
Determining whether left peripheral luma pixels adjacent to the left of the luma reference block are available among the peripheral luma pixels of the luma reference block corresponding to the current chroma block and whether upper peripheral luma pixels adjacent to the top of the luma reference block are available to do;
determining peripheral luma reference pixels according to whether the left peripheral luma pixels are available and whether the top peripheral luma pixels are available;
determining neighboring chroma reference pixels corresponding to the neighboring luma reference pixels from among neighboring chroma pixels located around the current chroma block;
deriving a prediction parameter for the current chroma block based on the neighboring luma reference pixels and the neighboring chroma reference pixels;
deriving a prediction value of a pixel in the current chroma block based on the pixel in the luma reference block and the prediction parameter; and
based on the prediction value, reconstructing a pixel in the current chroma block;
The left peripheral luma pixels include a first left peripheral luma pixel immediately adjacent to the left side of the luma reference block and a second left peripheral luma pixel immediately adjacent to the left side of the first left peripheral luma pixel,
When the left peripheral luma pixels are available, a downsampled peripheral luma reference pixel determined based on the first left peripheral luma pixel and the second left peripheral luma pixel and a peripheral chroma corresponding to the downsampled peripheral luma reference pixel based on a reference pixel, the prediction parameter for the current chroma block is derived.

Images