BR122023022608A2 - DECODING APPARATUS FOR DECODING IMAGES, IMAGE CODING APPARATUS FOR CODING IMAGES AND APPARATUS FOR TRANSMITTING DATA TO AN IMAGE INFORMATION - Google Patents

Abstract

A method for image decoding in accordance with the present document may comprise the steps of: deriving an intraprediction mode of a chroma block as a cross-component linear model (CCLM) mode; updating the intraprediction mode of the chroma block based on an intraprediction mode of a luma block corresponding to the chroma block; and determining a set of LFNST including LFNST matrices based on the updated intraprediction mode, wherein the updated intraprediction mode is derived as an intraprediction mode corresponding to a specific position in the luma block, and the specific position is defined based on a chroma block color format.

Description

FUNDAMENTALS Field of Technique

[001] The present disclosure relates to an image coding technique and, more particularly, to a method and apparatus for coding an image based on the transform in an image coding system.

Related Technique

[002] Currently, the demand for high-resolution and high-quality images/videos, such as 4K, 8K or more ultra-high definition (UHD) images/videos, is increasing in various fields. As image/video data becomes of higher resolution and quality, the amount of information transmitted or the number of bits increases compared to conventional image data. Therefore, when image data is transmitted using a medium such as a conventional wired/wireless broadband line or image/video data is stored using an existing storage medium, the transmission cost and storage cost are increased.

[003] Furthermore, nowadays, the interest and demand for immersive media such as virtual reality (VR), artificial reality (AR) content or hologram, or the like, is increasing, and the transmission of images/videos with image characteristics different from those of real images such as a game image is increasing.

[004] Consequently, there is a need for a highly efficient image/video compression technique to effectively compress and transmit or store and reproduce high-resolution, high-quality image/video information with various features, as described above.

SUMMARY

[005] A technical aspect of the present disclosure is to provide a method and apparatus for increasing image coding efficiency.

[006] Another technical aspect of the present disclosure is to provide a method and apparatus for increasing efficiency in coding an LFNST index.

[007] Yet another technical aspect of the present disclosure is to provide a method and apparatus for increasing the efficiency of a second transform by encoding an LFNST index.

[008] Yet another technical aspect of the present disclosure is to provide an image coding method and an image coding apparatus for deriving a set of LFNST transforms using an intra mode for a luma block in a CCLM mode.

[009] According to an embodiment of the present disclosure, an image decoding method performed by a decoding apparatus is provided. The method may include: deriving an intraprediction mode for a chroma block as a cross-component linear model (CCLM) mode; updating the intraprediction mode for the chroma block based on an intraprediction mode for a luma block corresponding to the chroma block; determine a set of LFNST including LFNST matrices based on the updated intraprediction mode, wherein the updated intraprediction mode can be derived as an intraprediction mode corresponding to a specific position in the luma block, and wherein the specific position can be defined based on a chroma block color format.

[010] The specific position may be a central position of the luma block.

[011] The specific position can be defined as ((xTbY + (nTbW * SubWidthC)/2), (yTbY + (nTbH * SubHeightC)/2)), xTbY and yTbY can denote upper left coordinates of the luma block, nTbW and nTbH can denote a width and a height of the chroma block, and SubWidthC and SubHeightC can denote variables corresponding to the color format.

[012] When the color format is 4:2:0, SubWidthC and SubHeightC can be 2, and when the color format is 4:2:2, SubWidthC can be 2 and SubHeightC can be 1.

[013] When the intraprediction mode corresponding to the specific position is a MIP mode, the updated intraprediction mode may be an intraplanar mode.

[014] When the intraprediction mode corresponding to the specific position is an IBC mode, the updated intraprediction mode may be an intraDC mode.

[015] When the intraprediction mode corresponding to the specific position is a palette mode, the updated intraprediction mode may be an intraDC mode.

[016] According to an embodiment of the present disclosure, an image coding method performed by a coding apparatus is provided. The method may include: deriving an intraprediction mode for a chroma block as a cross-component linear model (CCLM) mode; derive prediction samples for the chroma block based on the CCLM mode; derive residual samples for the chroma block based on the prediction samples; updating the intraprediction mode for the chroma block based on an intraprediction mode for a luma block corresponding to the chroma block determining a set of LFNST including LFNST matrices based on the updated intraprediction mode; and deriving modified transform coefficients for the chroma block based on the residual samples and the LFNST matrix, wherein the updated intraprediction mode can be derived as an intraprediction mode corresponding to a specific position in the luma block, and the specific position can be defined based on a chroma block color format.

[017] According to yet another embodiment of the present disclosure, a digital storage medium may be provided that stores image data including encoded image information and a bit stream generated in accordance with an image coding method performed by an apparatus coding.

[018] According to yet another embodiment of the present disclosure, a digital storage medium may be provided that stores image data including encoded image information and a bit stream to cause a decoding apparatus to perform the method of decoding image.

[019] According to the present disclosure, it is possible to increase the overall efficiency of image/video compression.

[020] According to the present disclosure, it is possible to increase efficiency in coding an LFNST index.

[021] According to the present disclosure, it is possible to increase the efficiency of a second transform by encoding an LFNST index.

[022] According to the present disclosure, it is possible to provide an image coding method and an image coding apparatus for deriving a set of LFNST transforms using an intra mode for a luma block in a CCLM mode.

[023] The effects that can be obtained through specific examples of the present disclosure are not limited to the effects listed above. For example, there may be various technical effects that a person skilled in the related art can understand or derive from the present disclosure. Accordingly, the specific effects of the present disclosure are not limited to those explicitly described in the present disclosure and may include various effects that can be understood or derived from the technical features of the present disclosure.

BRIEF DESCRIPTION OF THE DRAWINGS

[024] Figure 1 is a diagram that schematically describes a configuration of a video/image coding apparatus to which the present document can be applied.

[025] Figure 2 is a diagram that schematically describes a configuration of a video/image decoding apparatus to which the present document can be applied.

[026] Figure 3 schematically illustrates a multiple transform technique according to an embodiment of the present disclosure.

[027] Figure 4 exemplifies intradirectional modes of 65 prediction directions.

[028] Figure 5 is a diagram illustrating an RST in accordance with an embodiment of the present disclosure.

[029] Figure 6 is a diagram illustrating an output data arrangement sequence of a direct primary transform into a one-dimensional vector according to an example.

[030] Figure 7 is a diagram illustrating an output data arrangement sequence of a direct secondary transform in a two-dimensional block according to an example.

[031] Figure 8 is a diagram illustrating a block format to which LFNST is applied.

[032] Figure 9 is a diagram illustrating an output data arrangement of a direct LFNST according to an example.

[033] Figure 10 illustrates zero-out in a block to which a 4 x 4 LFNST is applied according to an example.

[034] Figure 11 illustrates zero-out in a block to which an 8 x 8 LFNST is applied according to an example.

[035] Figure 12 illustrates a CCLM applicable when deriving an intraprediction mode for a chroma block according to an embodiment.

[036] Figure 13 illustrates an image decoding method according to an example.

[037] Figure 14 illustrates an image coding method according to an example.

[038] Figure 15 schematically illustrates an example of a video/image coding system to which the present disclosure is applicable.

[039] Figure 16 illustrates the structure of a content streaming system to which the present disclosure is applied.

DESCRIPTION OF EXAMPLE MODALITIES

[040] Although the present disclosure may be susceptible to various modifications and includes various embodiments, its specific embodiments have been shown in the drawings by way of example and will now be described in detail. However, this is not intended to limit the present disclosure to the specific embodiments disclosed herein. The terminology used herein is intended only to describe specific embodiments and is not intended to limit the technical idea of the present disclosure. Singular forms may include plural forms unless the context clearly indicates otherwise. Terms such as “include” and “have” are intended to indicate that there are features, numbers, steps, operations, elements, components or combinations thereof used in the following description and, therefore, should not be understood as the possibility of existence or addition of one or more different features, numbers, steps, operations, elements, components or combinations thereof are excluded in advance.

[041] Meanwhile, each component in the drawings described in this document is illustrated independently for convenience of description as to characteristic functions different from each other and, however, it does not mean that each component is realized by separate hardware or software. For example, any two or more of these components can be combined to form a single component, and any single component can be divided into multiple components. The embodiments in which the components are combined and/or divided will fall within the scope of the patent right of the present disclosure, as long as they do not depart from the essence of the present disclosure.

[042] Next, the preferred embodiments of the present disclosure will be explained in more detail by referring to the attached drawings. Furthermore, the same reference signs are used for the same components in the drawings, and repeated descriptions for the same components will be omitted.

[043] This document refers to video/image coding. For example, the method/example disclosed in this document may be related to a VVC (Versatile Video Coding) standard (ITU-T Rec. H.266), a next generation video/image coding standard after VVC, or other related video coding standards (e.g. HEVC (High Efficiency Video Coding) standard (ITU-T Rec. H.265), EVC (Essential Video Coding) standard, AVS2 standard, etc..).

[044] In this document, a variety of modalities related to video/image coding may be provided and, unless otherwise specified, the modalities may be combined with each other and carried out.

[045] In this document, a video can mean a set of a series of images over time. Generally a picture means a unit that represents an image in a specific time zone, and a slice/block is a unit that constitutes a part of the picture. The slice/block may include one or more coding tree units (CTUs). An engraving can be made up of one or more slices/blocks. An engraving can be made up of one or more groups of blocks. A block group can include one or more blocks.

[046] A pixel or a pel can mean a smaller unit that constitutes an engraving (or image). Additionally, 'sample' can be used as a term corresponding to a pixel. A swatch can generally represent a pixel or a value of a pixel, and can represent just a pixel/pixel value of a luma component or just a pixel/pixel value of a chroma component. Alternatively, the sample may refer to a pixel value in the spatial domain, or when this pixel value is converted to the frequency domain, it may refer to a transform coefficient in the frequency domain.

[047] A unit may represent the basic image processing unit. The unit can include at least one specific region and information related to the region. A unit may include one luma block and two chroma blocks (e.g. cb, cr). The unit and a term such as a block, an area or the like can be used in place of each other according to the circumstances. In a general case, an M x N block may include a set (or a matrix) of samples (or sample matrices) or transform coefficients consisting of M columns and N rows.

[048] In this document, the terms “/” and “,” should be interpreted to indicate “and/or”. For example, the expression “A/B” can mean “A and/or B”. Furthermore, “A, B” can mean “A and/or B”. Furthermore, “A/B/C” can mean “at least one of A, B and/or C”. Furthermore, “A/B/C” can mean “at least one of A, B and/or C”.

[049] Furthermore, in the document, the term “or” must be interpreted to indicate “and/or”. For example, the expression “A or B” may include 1) A only, 2) B only, and/or 3) both A and B. In other words, the term “or” in this document should be interpreted to indicate “further or alternatively".

[050] In the present disclosure, “at least one of A and B” may mean “A only”, “B only” or “both A and B”. Furthermore, in the present disclosure, the expression “at least one of A or B” or “at least one of A and/or B” may be interpreted as “at least one of A and B”.

[051] Furthermore, in the present disclosure, “at least one of A, B and C” may mean “A only”, “B only”, “C only” or “any combination of A, B and C”. Furthermore, “at least one of A, B or C” or “at least one of A, B and/or C” can mean “at least one of A, B and C”.

[052] Furthermore, a parenthesis used in the present disclosure may mean “for example”. Specifically, when indicated as “prediction (intraprediction)”, it may mean that “intraprediction” is proposed as an example of “prediction”. In other words, the “prediction” of the present disclosure is not limited to “intra-prediction”, and “intra-prediction” may be proposed as an example of “prediction”. Furthermore, when indicated as “prediction (i.e., intraprediction)”, it can also mean that “intraprediction” is proposed as an example of “prediction”.

[053] Technical features individually described in a figure in the present disclosure can be individually implemented or can be simultaneously implemented.

[054] Figure 1 is a diagram that schematically illustrates a configuration of a video/image coding apparatus applicable to the embodiments of this document. Next, what is termed as the video coding apparatus may include the image coding apparatus.

[055] Referring to Figure 1, the coding apparatus 100 may include and be configured with an image partitioner 110, a predictor 120, a residual processor 130, an entropy encoder 140, an adder 150, a filter 160, and a memory 170. The predictor 120 may include an interpredictor 121 and an intrapredictor 122. The residual processor 130 may include a transformer 132, a quantizer 133, a dequantizer 134, and an inverse transformer 135. The residual processor 130 may further include a subtractor 131. The adder 150 may be termed as a reconstructed block reconstructor or generator. The image partitioner 110, the predictor 120, the residual processor 130, the entropy encoder 140, the adder 150, and the filter 160, which were described above, may be configured by one or more hardware components (e.g., chipsets of encoder or processors) according to an embodiment. Furthermore, memory 170 may include a decoded write buffer (DPB), and may be comprised of a digital storage medium. The hardware component may further include memory 170 as an internal/external component.

[056] The image partitioner 110 may partition an input image (or a picture or a frame) input to the coding apparatus 100 into one or more processing units. As an example, the processing unit may be called a coding unit (CU). In this case, starting with a coding tree unit (CTU) or the largest coding unit (LCU), the coding unit can be recursively partitioned according to the quaternary binary tree ternary tree structure (QTBTTT). For example, a coding unit can be divided into a plurality of coding units of a deeper depth based on the quaternary tree structure, the binary tree structure and/or the ternary structure. In this case, for example, the quaternary tree structure can be applied first and the binary tree structure and/or the ternary structure can be applied later. Alternatively, the binary tree structure can be applied first. The coding procedure according to the present disclosure can be performed based on the final coding unit that is no longer partitioned. In this case, the maximum coding unit can be directly used as the final coding unit based on the coding efficiency according to the image characteristic. Alternatively, the coding unit can be recursively divided into coding units of even greater depth as needed, so that the coding unit of an optimal size can be used as a final coding unit. Here, the coding procedure may include procedures such as prediction, transformation and reconstruction, which will be described later. As another example, the processing unit may further include a prediction unit (PU) or a transform unit (TU). In this case, the prediction unit and the transform unit can be divided or partitioned from the final coding unit described above. The prediction unit may be a sample prediction unit and the transform unit may be a unit for deriving a transform coefficient and/or a unit for deriving a residual signal from a transform coefficient.

[057] The unit and a term such as a block, an area or the like can be used in place of each other according to the circumstances. In a general case, an M x N block can represent a set of samples or transform coefficients consisting of M columns and N rows. The sample may generally represent a pixel or a value of a pixel, and may represent only a pixel/pixel value of a luma component, or only a pixel/pixel value of a chroma component. Sample can be used as a term corresponding to a pixel or a skin of an engraving (or image).

[058] In coding apparatus 100, a prediction signal (predicted block, prediction sample matrix) output from interpredictor 121 or intrapredictor 122 may be subtracted from an input image signal (original block, original sample matrix ) to generate a residual signal (residual block, residual sample matrix) and the generated residual signal is transmitted to the transformer 132. In this case, as shown, a subtraction unit of a prediction signal (predicted block, sample matrix of prediction) of the input image signal (original block, original sample matrix) in the coding apparatus 100 may be called subtractor 131. The predictor may perform prediction on a processing target block (hereinafter referred to as 'current block') and can generate a predicted block including prediction samples for the current block. The predictor can determine whether intraprediction or interprediction is applied on a current block or CU basis. As discussed later in the description of each prediction mode, the predictor may generate various prediction-related information, such as prediction mode information, and transmit the generated information to the entropy encoder 140. The prediction-related information may be encoded in the encoder of entropy 140, and be output in the form of a stream of bits.

[059] The intrapredictor 122 can predict the current block by referring to samples in the current engraving. Said samples can be located in the neighborhood or far from the current block according to the prediction mode. In intraprediction, the prediction modes may include a plurality of non-directional modes and a plurality of directional modes. Non-directional modes may include, for example, a DC mode and a planar mode. The directional mode may include, for example, 33 directional prediction modes or 65 directional prediction modes according to the degree of detail of the prediction direction. However, this is just an example, and more or less directional prediction modes can be used depending on a configuration. The intrapredictor 122 may determine the prediction mode applied to the current block using the prediction mode applied to the neighboring block.

[060] The interpredictor 121 may derive a predicted block for the current block based on a reference block (reference sample array) specified by a motion vector in a reference engraving. At this time, in order to reduce the amount of motion information transmitted in the between prediction mode, motion information can be predicted in a block, sub-block or sample basis based on the correlation of motion information between the neighboring block and the current block. The motion information may include a motion vector and a reference engraving index. The motion information may further include inter-prediction direction information (L0 prediction, L1 prediction, Bi prediction, etc.). In the case of interprediction, the neighboring block may include a spatial neighboring block existing in the current picture and a temporal neighboring block existing in the reference picture. The reference picture including the reference block and the reference picture including the temporal neighboring block may be the same or different from each other. The temporal neighboring block can be called collocated reference block, collocated CU (colCU) and the like, and the reference engraving including the temporal neighboring block can be called collocated engraving (colPic). For example, the interpredictor 121 may configure a candidate list of motion information based on neighboring blocks and generate information indicating which candidate is used to derive a motion vector and/or a reference engraving index of the current block. Interprediction can be performed based on several prediction modes. For example, in the case of a jump mode and a merge mode, the interpredictor 121 may use movement information from the neighboring block as movement information from the current block. In skip mode, unlike blending mode, the residual signal may not be transmitted. In the case of motion information prediction mode (motion vector prediction, MVP), the motion vector of the neighboring block can be used as a motion vector predictor, and the motion vector of the current block can be indicated by signaling of a motion vector difference.

[061] The predictor 120 may generate a prediction signal based on various prediction methods to be described below. For example, for prediction in a block, the predictor can apply intraprediction or interprediction, and also apply intraprediction and interprediction at the same time. The latter can be called combined inter and intra prediction (CIIP). Furthermore, the predictor can be based on an intra-block copy (IBC) prediction mode or a palette mode to perform prediction on a block. IBC prediction mode or palette mode can be used for image/video coding of game content or the like, such as screen content coding (SCC). IBC basically performs prediction on a current picture, but it can be performed in a similar way to interprediction in that it derives a reference block on a current picture. That is, the IBC may use at least one of the interprediction techniques described in this document. Palette mode can be considered as an example of intracoding or intraprediction. When palette mode is applied, a sample value in an engraving can be flagged based on information from a palette index and a palette table.

[062] The prediction signal generated through the predictor (including the interpredictor 121 and/or the intrapredictor 122) can be used to generate a reconstructed signal or to generate a residual signal. Transformer 132 can generate transform coefficients by applying a transform technique to the residual signal. For example, the transform technique may include at least one of a discrete cosine transform (DCT), a discrete sine transform (DST), a Karhunen-Loève transform (KLT), a graph-based transform (GBT), or a conditionally nonlinear transform (CNT). Here, the GBT means transform obtained from a graph when the relationship information between pixels is represented by the graph. The CNT mean transform obtained based on a prediction signal generated using all previously reconstructed pixels. Furthermore, the transform process can be applied to square pixel blocks of the same size or can be applied to non-square blocks of varying sizes.

[063] The quantizer 133 can quantify the transform coefficients and transmit them to the entropy encoder 140, and the entropy encoder 140 can encode the quantized signal (information about the quantized transform coefficients) and output the encoded signal in a stream of bits. The information about the quantized transform coefficients can be termed as residual information. The quantizer 133 can rearrange the block-type quantized transform coefficients into a one-dimensional vector form based on a coefficient sweep order and generate information about the quantized transform coefficients based on the quantized transform coefficients of the one-dimensional vector form. The entropy encoder 140 can realize various encoding methods, such as, for example, exponential Golomb, context-adaptive variable-length coding (CAVLC), context-adaptive binary arithmetic coding (CABAC), and the like. The entropy encoder 140 may encode information necessary for video/image reconstruction other than quantized transform coefficients (e.g., values of syntax elements, etc.) together or separately. The encoded information (e.g., encoded video/image information) may be transmitted or stored on a unit basis of a network abstraction layer (NAL) in the form of a bit stream. The video/image information may further include information about various parameter sets, such as an adaptation parameter set (APS), an engraving parameter set (PPS), a sequence parameter set (SPS), a video parameters (VPS) or similar. Additionally, video/image information may also include general restriction information. In this document, information and/or syntax elements that are transmitted/signaled to the decoding apparatus from the encoding apparatus may be incorporated into video/image information. The video/image information may be encoded via the above-described coding procedure and incorporated into the bit stream. The bit stream can be transmitted over a network or stored on a digital storage medium. Here, the network may include a transmission network, a communication network and/or the like, and the digital storage medium may include various storage media such as USB, SD, CD, DVD, Blu-ray, HDD, SSD and similar. A transmitter (not shown) that transmits an output signal from the entropy encoder 140 and/or a store (not shown) that stores it may be configured as an internal/external element of the coding apparatus 100, or the transmitter may be included in entropy encoder 140.

[064] The quantized transform coefficients output from the quantizer 133 can be used to generate a prediction signal. For example, by applying dequantization and inverse transform to transform coefficients quantized through dequantizer 134 and inverse transformer 135, the residual signal (residual block or residual samples) can be reconstructed. Adder 155 adds the reconstructed residual signal to a prediction signal output of interpredictor 121 or intrapredictor 122 so that a reconstructed signal (reconstructed etching, reconstructed block, reconstructed sample matrix) can be generated. When there is no residue for a processing target block, as in the case where skip mode is applied, the predicted block can be used as a reconstructed block. The adder 150 may be called a rebuilder or a reconstructed block generator. The generated reconstruction signal can be used for intra-prediction of a next processing target block in the current block and, as described later, can be used for inter-prediction of a next write through filtering.

[065] Meanwhile, in the process of encoding and/or reconstructing the engraving, chroma scale luma mapping (LMCS) can be applied.

[066] Filter 160 can improve subjective/objective video quality by applying filtering to the reconstructed signal. For example, filter 160 may generate a modified reconstructed picture by applying various filtering methods to the reconstructed picture and may store the modified reconstructed picture in memory 170, specifically in the DPB of memory 170. The various filtering methods may include, for example, filtering deblocking, adaptive sample shift, an adaptive loop filter, a two-sided filter or the like. As discussed further in the description of each filtering method, filter 160 may generate various filtering-related information and transmit the generated information to entropy encoder 140. Filtering information may be encoded in entropy encoder 140 and output at bit stream form.

[067] The modified reconstructed picture that has been transmitted to the memory 170 can be used as a reference picture in the interpredictor 121. Through this, the coding apparatus can avoid prediction incompatibility in the coding apparatus 100 and a decoding apparatus when the interprediction is applied and can also improve coding efficiency.

[068] DPB memory 170 may store the modified reconstructed picture for use as a reference picture in interpredictor 121. Memory 170 may store motion information of a block in the current picture from which the motion information was derived (or encoded) and/or block movement information in an already reconstructed engraving. The stored motion information may be transmitted to interpredictor 121 for use as motion information of a neighboring block or motion information of a temporal neighboring block. Memory 170 may store reconstructed samples of reconstructed blocks in the current engraving and transmit them to intrapredictor 122.

[069] Figure 2 is a diagram that schematically illustrates a configuration of a video/image decoding apparatus applicable to embodiments of this document.

[070] Referring to Figure 2, the decoding apparatus 200 may include and be configured with an entropy decoder 210, a residual processor 220, a predictor 230, an adder 240, a filter 250, and a memory 260. The predictor 230 may include an interpredictor 232 and an intrapredictor 231. The residual processor 220 may include a dequantizer 221 and an inverse transformer 221. The entropy decoder 210, the residual processor 220, the predictor 230, the adder 240, and the filter 250, which have been described above, may be configured by one or more hardware components (e.g., decoder chipsets or processors) according to one embodiment. Furthermore, memory 260 may include a decoded write buffer (DPB), and may be comprised of a digital storage medium. The hardware component may further include memory 260 as an internal/external component.

[071] When a bit stream including video/image information is input, the decoding apparatus 200 may reconstruct an image corresponding to a process by which the video/image information was processed in the encoding apparatus of Figure 1. For example , the decoding apparatus 200 may derive units/blocks based on the block partition information obtained from the bit stream. The decoding apparatus 200 may perform decoding using a processing unit applied to the encoding apparatus. Therefore, the decoding processing unit may be, for example, an encoding unit, which may be partitioned along the quadruple tree structure, the binary tree structure, and/or the ternary tree structure of a encoding or a larger encoding unit. One or more transform units may be derived from the coding unit. And, the reconstructed image signal decoded and output through the decoding apparatus 200 can be reproduced through a player.

[072] The decoding apparatus 200 may receive an output signal from the encoding apparatus of Figure 1 in the form of a bit stream and the received signal may be decoded via the entropy decoder 210. For example, the entropy decoder 210 can analyze the bit stream to derive information (e.g., video/image information) necessary for image reconstruction (or engraving reconstruction). The video/image information may further include information about various parameter sets, such as an adaptation parameter set (APS), an engraving parameter set (PPS), a sequence parameter set (SPS), a video parameters (VPS) or similar. Additionally, video/image information may also include general restriction information. The decoding apparatus may further decode a picture based on parameter set information and/or general constraint information. In the present disclosure, signaled/received information and/or syntax elements, which will be described later, can be decoded through the decoding procedure and obtained from the bit stream. For example, the entropy decoder 210 may decode information in the bitstream based on a coding method such as exponential Golomb coding, CAVLC, CABAC, or the like, and may output a value of a syntax element necessary for image reconstruction and quantized values of a transform coefficient with respect to a residue. More specifically, a CABAC entropy decoding method may receive a bin corresponding to each syntax element in a bitstream, determine a context model using decoding target syntax element information and decoding information from neighboring target blocks, and of decoding, or symbol/bin information decoded in a previous step, predict the bin generation probability according to the given context model and perform arithmetic decoding of the bin to generate a symbol corresponding to each syntax element value. Here, the CABAC entropy decoding method can update the context model using information from a decoded symbol/bin to a context model of the next symbol/bin after determining the context model. The prediction-related information between the information decoded in the entropy decoder 210 may be provided to the predictor (interpredictor 232 and intrapredictor 231), and residual values, i.e., quantified transform coefficients, on which entropy decoding was performed in the decoder entropy 210, and associated parameter information may be input to the residual processor 220. The residual processor 220 may derive a residual signal (residual block, residual samples, residual sample array). Furthermore, information about filtering between information decoded in the entropy decoder 210 may be provided to the filter 250. Meanwhile, a receiver (not shown) that receives an output signal from the encoding apparatus may further constitute the decoding apparatus 200 as an internal/external element, and the receiver may be a component of the entropy decoder 210. Meanwhile, the decoding apparatus according to the present document may be called a video/image/recording coding apparatus, and the decoding apparatus Decoding can be classified into an information decoder (video/picture/picture information decoder) and a sample decoder (video/picture/picture sample decoder). The information decoder may include the entropy decoder 210 and the sample decoder may include at least one of the dequantizer 221, the inverse transformer 222, the adder 240, the filter 250, the memory 260, the interpredictor 232 and the intrapredictor 231 .

[073] The dequantizer 221 can output transform coefficients by dequantizing the quantized transform coefficients. The dequantizer 221 can rearrange the quantized transform coefficients into the form of a two-dimensional block. In this case, the rearrangement may perform the rearrangement based on a coefficient sweep order that was performed in the coding apparatus. The dequantizer 221 may perform dequantization on the quantized transform coefficients using the quantization parameter (e.g., quantization step size information) and obtain transform coefficients.

[074] The dequantizer 222 obtains a residual signal (residual block, residual sample matrix) by inverse transform transform coefficients.

[075] The predictor can perform prediction on the current block, and generate a predicted block including prediction samples for the current block. The predictor may determine whether intra-prediction or inter-prediction is applied to the current block based on information in the prediction output of the entropy decoder 210 and specifically may determine an intra-/inter-prediction mode.

[076] The predictor 220 may generate a prediction signal based on various prediction methods to be described below. For example, for prediction in a block, the predictor can apply intraprediction or interprediction, and also apply intraprediction and interprediction at the same time. The latter can be called combined inter and intra prediction (CIIP). Furthermore, the predictor can be based on an intra-block copy (IBC) prediction mode or a palette mode to perform prediction on a block. IBC prediction mode or palette mode can be used for image/video coding of game content or the like, such as screen content coding (SCC). IBC basically performs prediction on a current picture, but it can be performed in a similar way to interprediction in that it derives a reference block on a current picture. That is, the IBC can use at least one of the interprediction techniques described in this document. Palette mode can be considered as an example of intracoding or intraprediction. When palette mode is applied, information in a palette table and a palette index can be included in the video/image information and flagged.

[077] The intrapredictor 231 can predict the current block by referring to the samples in the current engraving. Said samples can be located in the neighborhood or far from the current block according to the prediction mode. In intraprediction, the prediction modes may include a plurality of non-directional modes and a plurality of directional modes. The intrapredictor 231 may determine the prediction mode applied to the current block using the prediction mode applied to the neighboring block.

[078] The interpredictor 232 may derive a predicted block for the current block based on a reference block (reference sample array) specified by a motion vector in a reference engraving. At this time, in order to reduce the amount of motion information transmitted in the between prediction mode, motion information can be predicted in a block, sub-block or sample basis based on the correlation of motion information between the neighboring block and the current block. The motion information may include a motion vector and a reference engraving index. The motion information may further include inter-prediction direction information (L0 prediction, L1 prediction, Bi prediction, etc.). In the case of interprediction, the neighboring block may include a spatial neighboring block existing in the current picture and a temporal neighboring block existing in the reference picture. For example, the interpredictor 232 may configure a candidate list of motion information based on neighboring blocks and derive a motion vector and/or a reference engraving index of the current block based on the received candidate selection information. Interprediction may be performed based on various prediction modes, and the prediction information may include information that indicates an interprediction mode for the current block.

[079] The adder 240 adds the obtained residual signal to a prediction signal (predicted block, predicted sample matrix) output from the predictor (interpredictor 232 and/or intrapredictor 231), so that a reconstructed signal (reconstructed recording, block reconstructed, reconstructed sample matrix) can be generated. When there is no residue for a processing target block, as in the case where skip mode is applied, the predicted block can be used as a reconstructed block.

[080] The adder 240 can be called a reconstructor or reconstructed block generator. The generated reconstructed signal can be used to intra-predict a next processing target block in the current block and, as described later, can be output through filtering or be used to inter-predict a next write.

[081] Meanwhile, in the engraving decoding process, chroma scaled luma mapping (LMCS) can be applied.

[082] Filter 250 can improve subjective/objective video quality by applying filtering to the reconstructed signal. For example, filter 250 may generate a modified reconstructed picture by applying various filtering methods to the reconstructed picture and may transmit the modified reconstructed picture to memory 260, specifically to the DPB of memory 260. The various filtering methods may include, for example, filtering deblocking, adaptive sample shift, an adaptive loop filter, a two-sided filter or the like.

[083] The reconstructed (modified) picture that was stored in the DPB of memory 260 can be used as a reference picture in interpredictor 232. Memory 260 can store motion information of a block in the current picture, from which motion information was derived (or decoded) and/or block movement information in an already reconstructed engraving. The stored motion information may be transmitted to interpredictor 160 for use as motion information of a neighboring block or motion information of a temporal neighboring block. Memory 260 may store reconstructed samples of reconstructed blocks in the current engraving and transmit them to intrapredictor 231.

[084] In this document, the embodiments described in the filter 160, the interpredictor 121 and the intrapredictor 122 of the coding apparatus 100 can be applied in a similar or corresponding way to the filter 250, the interpredictor 232 and the intrapredictor 231 of the decoding apparatus 200 .

[085] As described above, prediction is performed to increase compression efficiency when performing video coding. Through this, a predicted block including prediction samples for a current block, which is an encoding target block, can be generated. Here, the predicted block includes prediction samples in a space domain (or pixel domain). The predicted block can be derived indentively in the encoding apparatus and the decoding apparatus, and the encoding apparatus can increase the image coding efficiency by signaling to the decoding apparatus not the original sample value of an original block itself, but the residual information (residual information) between the original block and the predicted block. The decoding apparatus may derive a residual block including residual samples based on the residual information, generate a reconstructed block including reconstructed samples by adding the residual block to the predicted block, and generate a reconstructed picture including reconstructed blocks.

[086] Residual information can be generated through transform and quantization procedures. For example, the coding apparatus may derive a residual block between the original block and the predicted block, derive transform coefficients by performing a transform procedure on residual samples (residual sample matrix) included in the residual block, and derive quantized transform coefficients by performing a procedure for quantifying the transform coefficients, so that it can signal the residual information associated with the decoding device (through a bit stream). Here, the residual information may include value information, position information, a transform technique, transform kernel, a quantization parameter, or the like of the quantized transform coefficients. The decoding apparatus may perform a quantization/dequantization procedure and derive the residual samples (or residual sample block) based on the residual information. The decoding apparatus may generate a reconstructed block based on a predicted block and the residual block. The encoding apparatus may derive a residual block by dequantization/inverse transform of quantized transform coefficients for reference for interprediction of a next picture and may generate a reconstructed picture based thereon.

[087] Figure 3 schematically illustrates a multiple transform technique according to an embodiment of the present disclosure.

[088] With reference to Figure 3, a transformer may correspond to the transformer in the coding apparatus of Figure 1, and an inverse transformer may correspond to the inverse transformer in the coding apparatus of Figure 1 above, or to the inverse transformer in the decoding apparatus of Figure 1. Figure 2.

[089] The transformer can derive transform (primary) coefficients by performing a primary transform based on residual samples (residual sample matrix) in a residual block (S410). This primary transform can be called the kernel transform. Here, the primary transform can be based on multiple transform selection (MTS) and when a multiple transform is applied as the primary transform, it can be termed as a multi-core transform.

[090] The multi-core transform can represent an additional transform method using the discrete cosine transform (DCT) type 2 and the discrete sine transform (DST) type 7, DCT type 8 and/or DST type 1. That is , the multi-core transform The transform can represent a transform method for transforming a residual signal (or residual block) of a space domain into transform coefficients (or primary transform coefficients) of a frequency domain based on a plurality of transform kernels selected from DCT type 2, DST type 7, DCT type 8 and DST type 1. Here, the primary transform coefficients can be called temporary transform coefficients from the transformer's point of view.

[091] In other words, when the conventional transform method is applied, the transform coefficients can be generated by applying the transform from a space domain to a frequency domain to a residual signal (or residual block) based on DCT type 2 Contrary to this, when the multi-core transform is applied, the transform coefficients (or primary transform coefficients) can be generated by applying the transform from a space domain to a frequency domain for a residual signal (or residual block). based on DCT type 2, DCT type 7, DCT type 8 and/or DCT type 1. Here, DCT type 2, DST type 7, DCT type 8 and DST type 1 can be called the type of transform, transform core or transform core. These types of DCT/DST transforms can be defined based on basic functions.

[092] When multi-core transform is performed, a vertical transform kernel and a horizontal transform kernel for a target block can be selected from among the transform kernels, a vertical transform can be performed on the target block based on the vertical transform kernel, and a horizontal transform can be performed on the target block based on the horizontal transform kernel. Here, the horizontal transform may indicate a transform in the horizontal components of the target block and the vertical transform may indicate a transform in the vertical components of the target block. The vertical transform kernel/horizontal transform kernel may be adaptively determined based on a prediction mode and/or a transform index for the target block (CU or sub-block) including a residual block.

[093] Furthermore, according to an example, if the primary transform is performed by applying the MTS, a mapping relationship for transform cores can be defined by setting specific basis functions to predetermined values and combining basis functions to be applied in the vertical transform or horizontal transform. For example, when the horizontal transform kernel is expressed as trTypeHor and the vertical direction transform kernel is expressed as trTypeVer, a trTypeHor or trTypeVer value of 0 can be set to DCT2, a trTypeHor or trTypeVer value of 1 can be set to DST7 and a trTypeHor or trTypeVer value of 2 can be set to DCT8.

[094] In this case, the MTS index information may be encoded and signaled to the decoding apparatus to indicate any one of a plurality of transform kernel sets. For example, an MTS index of 0 might indicate that the trTypeHor and trTypeVer values are 0, an MTS index of 1 might indicate that the trTypeHor and trTypeVer values are 1, an MTS index of 2 might indicate that the trTypeHor value is 2 and the value trTypeVer. is 1, an MTS index of 3 may indicate that the trTypeHor value is 1 and the trTypeVer value is 2, and an MTS index of 4 may indicate that both the trTypeHor and trTypeVer values are 2.

[095] In an example, the transform kernel sets according to the MTS index information are illustrated in the following table. Table 1

[096] The transformer can derive modified (secondary) transform coefficients by performing the secondary transform based on the (primary) transform coefficients (S420). The primary transform is a transform from a spatial domain to a frequency domain, and the secondary transform refers to the transform into a more compressive expression using an existing correlation between the (primary) transform coefficients. The secondary transform may include a non-separable transform. In this case, the secondary transform can be called the non-separable secondary transform (NSST) or the mode-dependent non-separable secondary transform (MDNSST). The non-separable secondary transform may represent a transform that generates modified transform coefficients (or secondary transform coefficients) for a residual signal by secondary transform, based on a non-separable transform matrix, (primary) transform coefficients derived through the transform primary. At this time, the vertical transform and the horizontal transform cannot be applied separately (or the horizontal and vertical transforms may not be applied independently) to the transform (primary) coefficients, but the transforms can be applied at once based on no separable transformed matrix. In other words, the non-separable secondary transform can represent a transform method in which the vertical direction and the horizontal direction are not applied separately to the (primary) transform coefficients and, for example, two-dimensional signals (transform coefficients) are rearranged for a one-dimensional signal through a certain determined direction (e.g., first row direction or first column direction), and then the modified transform coefficients (or secondary transform coefficients) are generated based on the non-separable transform matrix. For example, according to a first row order, the M x N blocks are arranged in a row in the order of a first row, a second row, ... and an nth row. According to a first column order, the M x N blocks are arranged in a row in the order of a first column, a second column, . and an nth column. The non-separable secondary transform can be applied to an upper left region of a block configured with (primary) transform coefficients (hereafter it can be referred to as a transform coefficient block). For example, if the width (W) and height (H) of the transform coefficient block are all equal to or greater than 8, an 8 x 8 non-separable secondary transform can be applied to an upper-left 8 x 8 region of the coefficient of transform block. Furthermore, if the width (W) and height (H) of the transform coefficient block are all equal to or greater than 4, and the width (W) or height (H) of the transform coefficient block is less than 8, then a 4 x 4 non-separable secondary transform can be applied to an upper left min(8,W) x min(8,H) region of the transform coefficient block. However, the embodiment is not limited to this and, for example, even if only the condition that the width (W) or height (H) of the transform coefficient block is equal to or greater than 4 is satisfied, the secondary transform non-separable 4 x 4 can be applied to the upper left min(8,W)xmin(8,H) region of the transform coefficient block.

[097] Specifically, for example, if a 4 x 4 input block is used, the non-separable secondary transform can be performed as follows.

[098] The 4 x 4 X input block can be represented as follows. Equation 1

[099] If X is represented in the form of a vector, vector A' can be represented as below. Equation 2

[0100] In Equation 2, the vector X is a one-dimensional vector obtained by rearranging the two-dimensional block X of Equation 1 according to first-line order.

[0101] In this case, the secondary non-separable transform can be calculated as below. Equation 3

[0102] In this equation, F represents a transform coefficient vector and T represents a 16 x 16 transform matrix (non-separable).

[0103] Through Equation 3 above, a 16 x 1 transform coefficient vector F can be derived, and the F can be derived and can be rearranged into a 4 x 4 block through a sweep order (horizontal, vertical, diagonal and the like). However, the calculation described above is an example, and the hypercube-Givens transform (HyGT) or similar can be used for the calculation of the non-separable secondary transform in order to reduce the computational complexity of the non-separable secondary transform.

[0104] Meanwhile, in the non-separable secondary transform, a transform kernel (or transform kernel, transform type) can be selected to be mode dependent. In this case, the mode may include the intraprediction mode and/or the interprediction mode.

[0105] As described above, the non-separable secondary transform can be performed based on an 8 x 8 transform or a 4 x 4 transform determined based on the width (W) and height (H) of the transform coefficient block. The 8 x 8 transform refers to a transform that is applicable to an 8 x 8 region included in the transform coefficient block when W and H are equal to or greater than 8, and the 8 x 8 region can be an 8 x region 8 top left in the transform coefficient block. Similarly, the 4 x 4 transform refers to a transform that is applicable to a 4 x 4 region included in the transform coefficient block when W and H are equal to or greater than 4, and the 4 x 4 region can be a 4 x 4 region top left in the transformed coefficient block. For example, an 8 x 8 transform kernel matrix can be a 64 x 64/16 x 64 transform kernel matrix, and a 4 x 4 transform kernel matrix can be a 16 x 16/8 x 16 matrix.

[0106] Here, to select a mode-dependent transform kernel, two transform-non-separable secondary transform kernels defined for a non-separable secondary transform can be configured for the 8 x 8 transform and the 4 x 4 transform, and cannot be four transform sets. That is, four transform sets can be configured for the 8 x 8 transform and four transform sets can be configured for the 4 x 4 transform. In this case, each of the four transform sets for the 8 x 8 transform can include two 8 x 8 transform cores and each of the four transform sets for the 4 x 4 transform may include two 4 x 4 transform cores.

[0107] However, as the size of the transform, that is, the size of a region to which the transform is applied, may be, for example, a size other than 8 x 8 or 4 x 4, the number of sets may be n, and the number of transform kernels in each set can be k.

[0108] The transform set can be referred to as an NSST set or an LFNST set. A specific set among the transform sets can be selected, for example, based on the intraprediction mode of the current block (CU or sub-block). A low-frequency non-separable transform (LFNST) can be an example of a reduced non-separable transform, which will be described later and represents a non-separable transform for a low-frequency component.

[0109] For reference, for example, the intraprediction mode may include two non-directional (or non-angular) intraprediction modes and 65 directional (or angular) intraprediction modes. The non-directional intraprediction modes may include a planar intraprediction mode of No. 0 and a DC intraprediction mode of No. 1, and the directional intraprediction modes may include 65 intraprediction modes of No. 2 to 66. However, this is an example, and this document can be applied even when the number of intraprediction modes is different. Meanwhile, in some cases, intraprediction mode No. 67 can be used additionally, and intraprediction mode No. 67 can represent a linear model (LM) mode.

[0110] Figure 4 exemplarily shows intradirectional modes of 65 prediction directions.

[0111] Referring to Figure 4, based on the intraprediction mode 34 with a left-up diagonal prediction direction, the intraprediction modes can be divided into intraprediction modes with horizontal directionality and intraprediction modes with vertical directionality. In Figure 4, H and V denote horizontal directionality and vertical directionality, respectively, and the numbers - 32 to 32 indicate displacements by 1/32 units in a sample grid position. These numerals can represent an offset to a mode index value. Intraprediction modes 2 to 33 have horizontal directionality and intraprediction modes 34 to 66 have vertical directionality. Strictly speaking, the intraprediction mode 34 can be considered as neither horizontal nor vertical, but can be classified as belonging to horizontal directionality in determining a transform set of a secondary transform. This is because the input data is transposed to be used for a symmetric vertical direction mode based on the intraprediction mode 34, and a method of aligning input data for a horizontal mode is used for the intraprediction mode 34. Transpose input data means that rows and columns of two-dimensional M x N block data are switched into N x M data. The intraprediction mode 18 and the intraprediction mode 50 may represent a horizontal intraprediction mode and a vertical intraprediction mode, respectively , and intraprediction mode 2 can be termed as a right-up diagonal intraprediction mode because intraprediction mode 2 has a reference pixel and performs prediction in the right upward direction. Likewise, the intraprediction mode 34 can be termed as a right descending diagonal intraprediction mode and the intraprediction mode 66 can be termed as a left descending diagonal intraprediction mode.

[0112] According to an example, the four transform sets according to the intraprediction mode can be mapped, for example, as shown in the following table. Table 2

[0113] As shown in Table 2, any of the four sets of transforms, that is, lfnstTrSetIdx, can be mapped to any of the four indices, that is, from 0 to 3, according to the intraprediction mode.

[0114] When it is determined that a specific set is used for the non-separable transform, one of the k transform kernels in the specific set may be selected via a secondary non-separable transform index. An encoding apparatus may derive a non-separable secondary transform index indicating a specific transform kernel based on a rate distortion (RD) check and may signal the non-separable secondary transform index to a decoding apparatus. The decoding apparatus may select one of the k transform kernels in the specific set based on the non-separable secondary transform index. For example, the index value of LFNST 0 may refer to a first non-separable secondary transform kernel, the index value of LFNST 1 may refer to a second non-separable secondary transform kernel, and the index value of LFNST 2 may refer to a third core non-separable secondary transform. Alternatively, the LFNST index value 0 may indicate that the first non-separable secondary transform is not applied to the target block, and the LFNST index values 1 to 3 may indicate the three transform kernels.

[0115] The transformer can perform the non-separable secondary transform based on the selected transform kernels and can obtain modified (secondary) transform coefficients. As described above, the modified transform coefficients can be derived as quantized transform coefficients via the quantizer and can be encoded and signaled to the decoding apparatus and transferred to the dequantizer/inverse transformer in the encoding apparatus.

[0116] Meanwhile, as described above, if the secondary transform is omitted, the (primary) transform coefficients, which are an output of the primary (separable) transform, can be derived as quantized transform coefficients through the quantizer as described above , and can be encoded and signaled to the decoding apparatus and transferred to the dequantizer/inverse transformer in the encoding apparatus.

[0117] The inverse transformer can perform a series of procedures in the reverse order to that in which they were performed in the transformer described above. The inverse transformer can receive transformer coefficients (dequantized) and derive transform coefficients (primary) by performing a secondary (inverse) transform (S450), and can obtain a residual block (residual samples) by performing a primary (inverse) transform on the transformer coefficients. transformed (primaries) (S460). In this context, the primary transform coefficients can be called modified transform coefficients from the point of view of the inverse transformer. As described above, the encoding apparatus and the decoding apparatus can generate the reconstructed block based on the residual block and the predicted block and can generate the reconstructed engraving based on the reconstructed block.

[0118] The decoding apparatus may further include a secondary inverse transform determinant (or an element for determining whether to apply a secondary inverse transform) and a secondary inverse transform determinant (or an element for determining a secondary inverse transform). . The secondary inverse transform application determinant can determine whether to apply a secondary inverse transform. For example, the secondary inverse transform may be an NSST, an RST, or an LFNST, and the secondary inverse transform apply determinant may determine whether to apply the secondary inverse transform based on a secondary transform flag obtained by analyzing the bit stream. In another example, the secondary inverse transform application determinant may determine whether to apply the secondary inverse transform based on a transform coefficient of a residual block.

[0119] The secondary inverse transform determinant can determine a secondary inverse transform. In this case, the secondary inverse transform determinant can determine the secondary inverse transform applied to the current block based on a set of LFNST transform (NSST or RST) specified according to an intraprediction mode. In one embodiment, a secondary transform determination method may be determined depending on a primary transform determination method. Various combinations of primary transforms and secondary transforms can be determined according to the intraprediction mode. Furthermore, in one example, the secondary inverse transform determinant may determine a region to which a secondary inverse transform is applied based on the current block size.

[0120] Meanwhile, as described above, if the secondary (inverse) transform is omitted, the (dequantized) transform coefficients can be received, the primary (separable) inverse transform can be performed, and the residual block (residual samples) can be obtained. As described above, the encoding apparatus and the decoding apparatus can generate the reconstructed block based on the residual block and the predicted block and can generate the reconstructed engraving based on the reconstructed block.

[0121] Meanwhile, in the present disclosure, a reduced secondary transform (RST) in which the size of a transform matrix (kernel) is reduced can be applied in the NSST concept to reduce the amount of computation and memory required for the transform non-separable secondary.

[0122] Meanwhile, the transform kernel, the transform matrix and the coefficient constituting the transform kernel matrix, that is, the kernel coefficient or the matrix coefficient, described in the present disclosure, can be expressed in 8 bits. This may be a condition for implementation in both the decoding apparatus and the encoding apparatus and may reduce the amount of memory required to store the transform kernel with a performance degradation that can be reasonably accommodated compared to the existing 9-bit or 10-bit . Additionally, expressing the kernel matrix in 8 bits can allow a small multiplier to be used and may be more suitable for single instruction multiple data (SIMD) instructions used for optimal software implementation.

[0123] In the present specification, the term “RST” may mean a transform that is performed on residual samples for a target block based on a transform matrix whose size is reduced according to a reduced factor. In the case of performing the reduced transform, the amount of computation required for the transform can be reduced due to a reduction in the size of the transform matrix. In other words, RST can be used to solve the problem of computational complexity that occurs in the non-separable transform or in the transform of a large block.

[0124] RST may be termed as various terms, such as reduced transform, reduced secondary transform, reduction transform, simplified transform, simple transform and the like, and the name that RST may be termed as is not limited to the listed examples. Alternatively, since RST is mainly performed in a low-frequency region by including a non-zero coefficient in a transform block, it can be termed as a Low-Frequency Non-Separable Transform (LFNST). The transform index can be termed as an LFNST index.

[0125] Meanwhile, when the secondary inverse transform is performed based on RST, the inverse transformer 135 of the coding apparatus 100 and the inverse transformer 222 of the decoding apparatus 200 may include a reduced secondary inverse transformer that derives modified transform coefficients. based on the inverse RST of the transform coefficients and an inverse primary transformer that derives residual samples for the target block based on the inverse primary transform of the modified transform coefficients. The inverse primary transform refers to the inverse transform of the primary transform applied to the residue. In the present disclosure, deriving a transform coefficient based on a transform may refer to deriving a transform coefficient by applying the transform.

[0126] Figure 5 is a diagram illustrating an RST in accordance with an embodiment of the present disclosure.

[0127] In the present disclosure, a “target block” may refer to a current block to be encoded, a residual block, or a transform block.

[0128] In RST according to an example, an N-dimensional vector can be mapped to an R-dimensional vector located in another space, so that the reduced transform matrix can be determined, where R is less than N. may mean the square of the length of a side of a block to which the transform is applied, or the total number of transform coefficients corresponding to a block to which the transform is applied, and the reduced factor may mean an R/N value. The reduced factor can be termed as a reduced factor, reduction factor, simplified factor, simple factor or other various terms. Meanwhile, R can be termed as a reduced coefficient, but according to the circumstances, the reduced factor can mean R. Also, according to the circumstances, the reduced factor can mean the N/R value.

[0129] In an example, the reduced factor or reduced coefficient may be signaled via a bit stream, but the example is not limited to this. For example, a predefined value for the reduced factor or the reduced coefficient may be stored in each encoding apparatus 100 and decoding apparatus 200, and in this case the reduced factor or the reduced coefficient may not be signaled separately.

[0130] The size of the reduced transform matrix according to an example may be R x N less than N x N, the size of a conventional transform matrix and may be defined as in Equation 4 below. Equation 4

[0131] The matrix T in the Reduced Transform block shown in Figure 5(a) may mean the TRxN matrix of Equation 4. As shown in Figure 5(a), when the reduced transform matrix TRxN is multiplied by residual samples for the target block, the transform coefficients for the target block can be derived.

[0132] In an example, if the size of the block to which the transform is applied is 8 x 8 and R = 16 (i.e. R/N = 16/64 = 1/4), then the RST according to the Figure 5(a) can be expressed as a matrix operation as shown in Equation 5 below. In this case, the memory and multiplication calculation can be reduced to approximately 1/4 by the reduced factor.

[0133] In the present disclosure, a matrix operation can be understood as an operation of multiplying a column vector by a matrix, arranged to the left of the column vector, to obtain a column vector. Equation 5

[0134] In Equation 5, r1 to r64 may represent residual samples for the target block and may specifically be transform coefficients generated by applying a primary transform. As a result of calculating Equation 5, the ci transform coefficients for the target block can be derived and a ci derivation process can be as in Equation 6. Equation 6

[0135] As a result of calculating Equation 6, the transform coefficients c1 to cR for the target block can be derived. That is, when R = 16, the transform coefficients c1 to c16 for the target block can be derived. If, instead of RST, a regular transform is applied and a transform matrix of size 64 x 64 (N x N) is multiplied to residual samples of size 64 x 1 (N x 1), only 16 (R) coefficients of transform will be derived for the target block because RST was applied, although 64 (N) transform coefficients will be derived for the target block. Once the total number of transform coefficients for the target block is reduced from N to R, the amount of data transmitted by the encoding apparatus 100 to the decoding apparatus 200 decreases, so that the transmission efficiency between the encoding apparatus 100 and the decoding apparatus encoding 100 and decoding apparatus 200 may be improved.

[0136] When considered from the point of view of the transform matrix size, the size of the regular transform matrix is 64 x 64 (N x N), but the size of the reduced transform matrix is reduced to 16 x 64 (R x N), then the memory usage in the case of performing the RST can be reduced by an R/N ratio when compared to the case of performing the regular transform. Furthermore, when compared to the number of N x N multiplication calculations in the case of using the regular transformed matrix, the use of the reduced transformed matrix can reduce the number of multiplication calculations by the ratio R/N (R x N).

[0137] In one example, transformer 132 of coding apparatus 100 may derive transform coefficients for the target block by performing the primary transform and the RST-based secondary transform on residual samples for the target block. These transform coefficients may be transferred to the inverse transformer of the decoding apparatus 200 and the inverse transformer 222 of the decoding apparatus 200 may derive the modified transform coefficients based on the reduced inverse secondary transform (RST) for the transform coefficients and may derive residual samples for the target block based on the inverse primary transform for the modified transform coefficients.

[0138] The size of the TRxN inverse RST matrix according to an example is N x R smaller than the N x N size of the regular inverse transform matrix and is in a transpose relationship with the TRxN reduced transform matrix shown in Eq. 4.

[0139] The Tt matrix in Reduced Inv. Transform block shown in Figure 5(b) can mean the inverse RST matrix TRxNT (the superscript T means transpose). When the inverse RST matrix TRxNT is multiplied by the transform coefficients for the target block as shown in Figure 5(b), the modified transform coefficients for the target block or the residual samples for the current block can be derived. The inverse RST matrix TRxNT can be expressed as (TRxNT)NxR.

[0140] More specifically, when the inverse RST is applied as the secondary inverse transform, the modified transform coefficients for the target block can be derived when the inverse RST matrix TRxNT multiplied by the transform coefficients for the target block. Meanwhile, the inverse RST can be applied as the inverse primary transform, and in this case, the residual samples for the target block can be derived when the inverse RST matrix TRxNT is multiplied by the transform coefficients for the target block.

[0141] In an example, if the size of the block to which the inverse transform is applied is 8 x 8 and R = 16 (i.e. R/N = 16/64 = 1/4), then the RST according to Figure 5(b) can be expressed as a matrix operation as shown in Equation 7 below. Equation 7 <<DRAW-CODE>

[0142] Na may represent the modified transform coefficients for the target block or the residual samples for the target block may be derived and the ri derivation process may be as in Equation 8. Equation 8

[0143] As a result of calculating Equation 8, r1 to rN representing the modified transform coefficients for the target block or the residual samples for the target block can be derived. When considered from the point of view of the size of the inverse transform matrix, the size of the regular inverse transform matrix is 64 x 64 (N x N), but the size of the reduced inverse transform matrix is reduced to 64 x 16 (R x N), then the memory usage in a case of performing the inverse RST can be reduced by an R/N ratio when compared to a case of performing the regular inverse transform. Furthermore, when compared to the number of N x N multiplication calculations in a regular inverse transform matrix use case, the use of the reduced inverse transform matrix can reduce the number of multiplication calculations by the ratio R/N (N x R).

[0144] A transform set configuration shown in Table 2 can also be applied to an 8 x 8 RST. That is, the 8 x 8 RST can be applied according to a set of transforms in Table 2. As a set of transform includes two or three transforms (kernels) according to an intraprediction mode, it can be configured to select one of up to four transforms including the one in a case where no secondary transform is applied. In a transform where no secondary transform is applied, one can consider applying an identity matrix. Assuming that the indices 0, 1, 2 and 3 are respectively assigned to the four transforms (for example, the index 0 can be allocated to a case where an identity matrix is applied, i.e., a case where no secondary transform is applied), a transform index or an LFNST index as a syntax element can be signaled for each transform coefficient block, thus designating a transform to be applied. That is, for a top left 8 x 8 block, through the transform index, it is possible to designate an 8 x 8 RST in an RST configuration, or designate an 8 x 8 lfnst when LFNST is applied. lfnst 8 x 8 and RST 8 x 8 refer to transforms applicable to an 8 x 8 region included in the transform coefficient block when W and H of the target block to be transformed are equal to or greater than 8, and the region 8x8 can be an upper left 8x8 region in the transform coefficient block. Similarly, a 4 x 4 lfnst and a 4 x 4 RST refer to transforms applicable to a 4 x 4 region included in the transform coefficient block when W and H of the target block are equal to or greater than 4, and the 4x4 region can be a top left 4x4 region in the transform coefficient block.

[0145] According to one embodiment of the present disclosure, for a transform in an encoding process, only 48 pieces of data can be selected and a maximum transform kernel matrix of 16 x 48 can be applied thereto, instead of apply a 16 x 64 transform kernel matrix to 64 pieces of data forming an 8 x 8 region. Here, “maximum” means that m has a maximum value of 16 in an m x 48 transform kernel matrix to generate m coefficients . That is, when an RST is performed by applying an m x 48 transform kernel matrix (m < 16) to an 8 x 8 region, 48 data are entered and m coefficients are generated. When m is 16, 48 data are entered and 16 coefficients are generated. That is, assuming that 48 data pieces form a 48 x 1 vector, a 16 x 48 matrix and a 48 x 1 vector are multiplied sequentially, thus generating a 16 x 1 vector. Here, the 48 pieces of data that form the 8 x region 8 can be arranged accordingly, thus forming the 48 x 1 vector. For example, a 48 x 1 vector can be constructed based on 48 pieces of data that constitute a region excluding the lower right 4 x 4 region between the 8 x 8 regions .Here, when a matrix operation is performed by applying a maximum transform kernel matrix of 16 x 48, 16 modified transform coefficients are generated and the 16 modified transform coefficients can be arranged in a top left 4 x 4 region accordingly. with a scan order, and a top-right 4x4 region and a bottom-left 4x4 region can be padded with zeros.

[0146] For an inverse transform in a decoding process, the transposed matrix of the previous transform kernel matrix can be used. That is, when an inverse RST or LFNST is performed in the inverse transform process performed by the decoding apparatus, the input coefficient data to which the inverse RST is applied is configured into a one-dimensional vector according to a predetermined arrangement order and a modified coefficient vector obtained by multiplying the one-dimensional vector and a corresponding inverse RST matrix to the left of the one-dimensional vector can be arranged in a two-dimensional block according to a predetermined arrangement order.

[0147] In summary, in the transform process, when an RST or LFNST is applied to an 8 x 8 region, a matrix operation of 48 transform coefficients on the upper left, upper right and lower left regions of the 8 x 8 regions excluding the lower right region between the transform coefficients in the 8 x 8 region and a 16 x 48 transform kernel matrix. For the matrix operation, the 48 transform coefficients are inserted into a one-dimensional matrix. When the matrix operation is performed, 16 modified transform coefficients are derived, and the modified transform coefficients can be arranged in the upper left region of the 8 x 8 region.

[0148] On the contrary, in the inverse transform process, when an inverse RST or LFNST is applied to an 8 x 8 region, 16 transform coefficients corresponding to an upper left region of the 8 x 8 region among the transform coefficients in the 8 x 8 region can be inserted into a one-dimensional matrix according to a scan order and can be subjected to a matrix operation with a 48 x 16 transform kernel matrix. That is, the matrix operation can be expressed as (48 x 16 matrix) * (16 x 1 transform coefficient vector) = (48 x 1 modified transform coefficient vector). Here, an n x 1 vector can be interpreted as having the same meaning as an n x 1 matrix and can therefore be expressed as an n x 1 column vector. Furthermore, *denotes matrix multiplication. When the matrix operation is performed, 48 modified transform coefficients can be derived and the 48 modified transform coefficients can be arranged in the upper left, upper right and lower left regions of the 8 x 8 region excluding a lower right region.

[0149] When a secondary inverse transform is based on an RST, the inverse transformer 135 of the coding apparatus 100 and the inverse transformer 222 of the decoding apparatus 200 may include a stepped-down secondary inverse transformer for deriving modified transform coefficients based on a Inverse RST on transform coefficients and an inverse primary transformer to derive residual samples for the target block based on an inverse primary transform on the modified transform coefficients. The inverse primary transform refers to the inverse transform of a primary transform applied to a residue. In the present disclosure, deriving a transform coefficient based on a transform may refer to deriving the transform coefficient by applying the transform.

[0150] The non-separated transform described above, the LFNST, will be described in detail below. The LFNST may include a forward transform by the encoding apparatus and an inverse transform by the decoding apparatus.

[0151] The coding apparatus receives a result (or part of a result) derived after applying a primary transform (core) as input and applies a direct secondary transform (secondary transform). Equation 9

[0152] In Equation 9, x and y are inputs and outputs of the secondary transform, respectively, and G is a matrix representing the secondary transform, and the base vectors of the transform are composed of column vectors. In the case of an inverse LFNST, when the dimension of the transform matrix G is expressed as [number of rows x number of columns], in the case of a direct LFNST, the transpose of the matrix G becomes the dimension of GT.

[0153] For the inverse LFNST, the dimensions of the matrix G are [48 x 16], [48 x 8], [16 x 16], [16 x 8], and the matrix [48 x 8] and the [16 ]x8] are partial matrices that sampled 8 transform basis vectors from the left of the [48 x 16] matrix and the [16 x 16] matrix, respectively.

[0154] On the other hand, for direct LFNST, the dimensions of the GT matrix are [16 x 48], [8 x 48], [16 x 16], [8 x 16], and the [8 x 48] matrix and matrix [8 x 16] are partial matrices obtained by sampling 8 transform basis vectors from the top of matrix [16 x 48] and matrix [16 x 16], respectively.

[0155] Therefore, in the case of direct LFNST, a vector [48 x 1] or [16 x 1] is possible as an input x, and a vector [16 x 1] or a vector [8 x 1] is possible as a y output. In video encoding and decoding, the output of the direct primary transform is two-dimensional (2D) data, so to construct vector [48 x 1] or vector [16 x 1] as the input x, a one-dimensional vector must be constructed by arranging appropriately the 2D data that is the output of the direct transform.

[0156] Figure 6 is a diagram illustrating an output data arrangement sequence of a direct primary transform into a one-dimensional vector according to an example. The left diagrams of (a) and (b) of Figure 6 show the sequence for constructing a [48 x 1] vector and the right diagrams of (a) and (b) of Figure 6 show the sequence for constructing a vector [16 x 1]. In the case of LFNST, a one-dimensional vector x can be obtained by sequentially arranging the 2D data in the same order as in (a) and (b) of Figure 6.

[0157] The arrangement direction of the direct primary transform output data can be determined according to an intraprediction mode of the current block. For example, when the intraprediction mode of the current block is in the horizontal direction with respect to the diagonal direction, the output data of the direct primary transform can be arranged in the order of (a) in Figure 6, and when the intraprediction mode of the block current is in the vertical direction with respect to the diagonal direction, the output data of the direct primary transform can be arranged in the order of (b) in Figure 6.

[0158] According to an example, an arrangement order different from the arrangement orders of (a) and (b) Figure 6 can be applied, and to derive the same result (vector y) as when the arrangement orders of ( a) and (b) Figure 6 is applied, the column vectors of the matrix G can be rearranged according to the arrangement order. In other words, it is possible to rearrange the column vectors of G so that each element that constitutes the vector x is always multiplied by the same base vector of the transform.

[0159] Since the output y derived through Equation 9 is a one-dimensional vector, when two-dimensional data is required as input data in the process of using the result of the direct secondary transform as input, for example, in the process of performing quantization or residual encoding, the output vector y of Equation 9 must be properly organized as 2D data again.

[0160] Figure 7 is a diagram illustrating an output data arrangement sequence of a direct secondary transform in a two-dimensional block according to an example.

[0161] In the case of LFNST, the output values can be arranged in a 2D block according to a predetermined scan order. (a) of Figure 7 shows that when the output y is a [16 x 1] vector, the output values are arranged in 16 positions of the 2D block according to a diagonal scan order. (b) of Figure 7 shows that when the output y is a vector [8 x 1], the output values are arranged in 8 positions of the 2D block according to the diagonal scan order, and the remaining 8 positions are filled with zeros. X in (b) of Figure 7 indicates that it is zero-filled.

[0162] According to another example, since the order in which the output vector y is processed in performing quantization or residual coding can be predefined, the output vector y cannot be arranged in the 2D block as shown in Figure 7. However, in case of residual coding, data coding can be performed in 2D block units (e.g. 4 x 4) as CG (Coefficient Group), and in this case, the data is organized according to a specific order as in the diagonal scan order of Figure 7.

[0163] Meanwhile, the decoding apparatus may configure the one-dimensional input vector y by arranging the two-dimensional data output through a dequantization or similar process according to a predefined scan order for the inverse transform. The input vector y can be output as the output vector x by the following equation. Equation 10

[0164] In the case of inverse LFNST, an output vector x can be derived by multiplying an input vector y, which is a [16 x 1] vector or an [8 x 1] vector, by a matrix G. For LFNST Inversely, the output vector x can be a [48 x 1] vector or a [16 x 1] vector.

[0165] The output vector x is arranged in a two-dimensional block according to the order shown in Figure 6 and is organized as two-dimensional data, and this two-dimensional data becomes input data (or a part of the input data) of the inverse primary transform.

[0166] Consequently, the inverse secondary transform is the opposite of the direct secondary transform process as a whole, and in the case of the inverse transform, unlike the direct direction, the inverse secondary transform is applied first and then the inverse primary transform is applied.

[0167] In inverse LFNST, one of the 8 [48 x 16] matrices and 8 [16 x 16] matrices can be selected as the transform matrix G. Whether to apply the matrix [48 x 16] or the [16 x 16 ]] depends on the size and shape of the block.

[0168] Furthermore, 8 matrices can be derived from four sets of transforms as shown in Table 2 above, and each set of transforms can consist of two matrices. Which set of transforms to use among the 4 sets of transforms is determined according to the intraprediction mode, and more specifically, the set of transforms is determined based on the value of the extended intraprediction mode considering the wide-angle intraprediction ( WAIP). Which matrix to select from the two matrices that constitute the selected transform set is derived through index signaling. More specifically, 0, 1 and 2 are possible as the transmitted index value, 0 can indicate that LFNST is not applied, and 1 and 2 can indicate either of the two transform matrices that constitute a transform set selected based on the mode intraprediction value.

[0169] Meanwhile, as described above, which transform matrix of the [48 x 16] matrix and the [16 x 16] matrix is applied to the LFNST is determined by the size and shape of the transform target block.

[0170] Figure 8 is a diagram illustrating a block format to which LFNST is applied. (a) of Figure 8 shows 4 x 4 blocks, (b) shows 4 x 8 and 8 x 4 blocks, (c) shows 4 x N or N x 4 blocks where N is 16 or more, (d) shows 8 x 8, (e) shows M x N blocks where M > 8, N > 8, and N > 8 or M > 8.

[0171] In Figure 8, blocks with thick edges indicate regions in which LFNST is applied. For the blocks in Figures 10 (a) and (b), the LFNST is applied to the top left 4 x 4 region and for the block in Figure 8 (c), the LFNST is applied individually to the two top left 4 x 4 regions. are arranged continuously. In (a), (b) and (c) of Figure 8, since the LFNST is applied in 4 x 4 region units, this LFNST will be referred to as “LFNST 4 x 4” from now on. Based on the matrix dimension for G, a [16 x 16] or [16 x 8] matrix can be applied.

[0172] More specifically, the [16 x 8] matrix is applied to the 4 x 4 block (4 x 4 TU or 4 x 4 CU) of Figure 8 (a) and the [16 x 16] matrix is applied to the blocks in (b) and (c) of Figure 8. This is to adjust the computational complexity for the worst case for 8 multiplications per sample.

[0173] With respect to (d) and (e) of Figure 8, the LFNST is applied to the upper left 8 x 8 region and this LFNST is hereinafter referred to as “LFNST 8 x 8”. As the corresponding transform matrix, a [48 x 16] matrix or a [48 x 8] matrix can be applied. In the case of direct LFNST, since the vector [48 x 1] (vector x in Equation 9) is entered as input data, all sample values from the upper left 8 x 8 region are not used as input values of the Direct LFNST. That is, as can be seen from the left order of Figure 6 (a) or the left order of Figure 6 (b), the vector [48 x 1] can be constructed based on samples belonging to the 3 blocks 4 x 4 remaining, leaving the bottom right 4 x 4 block as is.

[0174] The matrix [48 x 8] can be applied to an 8 x 8 block (8 x 8 TU or 8 x 8 CU) in Figure 8 (d), and the matrix [48 x 16] can be applied to the block 8 x 8 in Figure 8(e). This also serves to adjust the computational complexity for the worst case for 8 multiplications per sample.

[0175] Depending on the shape of the block, when the corresponding direct LFNST (LFNST 4 x 4 or LFNST 8 x 8) is applied, 8 or 16 output data (vector y in Equation 9, vector [8 x 1] or [16 x 1]) is generated. In direct LFNST, the number of output data is equal to or less than the number of input data due to the characteristics of the GT matrix.

[0176] Figure 9 is a diagram illustrating an arrangement of output data from a direct LFNST according to an example, and shows a block in which output data from the direct LFNST is arranged according to a block format.

[0177] The shaded area in the upper left part of the block shown in Figure 9 corresponds to the area where the direct LFNST output data is located, positions marked with 0 indicate samples filled with 0 values, and the remaining area represents regions that do not are changed by direct LFNST. In the area not changed by LFNST, the output data of the direct primary transform remains unchanged.

[0178] As described above, since the dimension of the applied transform matrix varies according to the shape of the block, the number of output data also varies. Like Figure 9, direct LFNST output data may not completely fill the upper left 4 x 4 block. In the case of (a) and (d) of Figure 9, a [16 x 8] matrix and a [48 x 8] matrix are applied to the block indicated by a thick line or a partial region within the block, respectively, and a vector [8 x 1] as output of the forward LFNST is generated. That is, according to the scanning order shown in (b) of Figure 7, only 8 output data can be filled as shown in (a) and (d) of Figure 9, and 0 can be filled in the remaining 8 positions. In the case of the LFNST applied block of Figure 8(d), as shown in Figure 9(d), two upper right and lower left 4 x 4 blocks adjacent to the upper left 4 x 4 block are also filled with 0 values.

[0179] As described above, basically, when signaling the LFNST index, it is specified whether to apply the LFNST and the transform matrix to be applied. As shown in Figure 9, when LFNST is applied, since the number of output data of direct LFNST may be equal to or less than the number of input data, a region filled with a zero value occurs as follows.

[0180] 1) As shown in (a) of Figure 9, samples from the 8th position and later in the scan order in the upper left 4 x 4 block, that is, samples from the 9th to the 16th.

[0181] 2) As shown in (d) and (e) of Figure 9, when the [48 x 16] matrix or the [48 x 8] matrix is applied, two 4 x 4 blocks adjacent to the upper 4 x 4 block left or the second and third 4 x 4 blocks in scan order.

[0182] Therefore, if there is non-zero data checking areas 1) and 2), it is certain that LFNST is not applied, so the corresponding LFNST index flag can be omitted.

[0183] According to an example, for example, in the case of LFNST adopted in the VVC standard, since the LFNST index signaling is performed after residual coding, the coding apparatus can know whether non-zero data exists (significant coefficients) for all positions within the TU or CU block through residual coding. Accordingly, the encoding apparatus can determine whether to perform signaling on the LFNST index based on the existence of non-zero data and the decoding apparatus can determine whether the LFNST index is parsed. When non-zero data does not exist in the area designated in 1) and 2) above, LFNST index signaling is performed.

[0184] However, for the adopted LFNST, the following simplification methods can be applied.

[0185] (i) According to an example, the number of output data for direct LFNST may be limited to a maximum of 16.

[0186] In the case of (c) of Figure 8, the 4 x 4 LFNST can be applied to two 4 x 4 regions adjacent to the upper left, respectively, and in this case, a maximum of 32 LFNST output data can be generated when the number of output data for direct LFNST is limited to a maximum of 16, in the case of 4 x N/N x 4 (N > 16) blocks (TU or CU), 4 x 4 LFNST is applied only to a 4 x 4 top left region, LFNST can be applied only once to all blocks in Figure 8. With this, the implementation of image coding can be simplified.

[0187] (ii) According to an example, zero-out can be additionally applied to a region to which LFNST is not applied. In this document, zero-out can mean filling values of all positions belonging to a given region with value 0. That is, zero-out can be applied to a region that is not changed due to LFNST and maintains the result of the direct primary transform. As described above, since the LFNST is divided into LFNST 4 x 4 and LFNST 8 x 8, the zero-out can be divided into two types ((ii)-(A) and (ii)-(B)) as he follows.

[0188] (ii)-( A) When LFNST 4 x 4 is applied, a region to which LFNST 4 x 4 is not applied can be reset. Figure 11 is a diagram illustrating zero-out in a block to which LFNST 4 x 4 is applied according to an example.

[0189] As shown in Figure 10, with respect to a block to which LFNST 4 x 4 is applied, that is, for all blocks in (a), (b) and (c) of Figure 9, the entire region to the which LFNST is not applied can be filled with zeros.

[0190] (d) of Figure 10 shows that zero-out is performed on a remaining block to which the 4 x 4 LFNST is not applied when the maximum number of direct LFNST output data pieces is limited to 16 in accordance with An example.

[0191] (ii)-(B) When an 8 x 8 LFNST is applied, a region to which the 8 x 8 LFNST is not applied can be reset. Figure 11 illustrates zero-out in a block to which an 8 x 8 LFNST is applied according to an example.

[0192] As shown in Figure 11, with respect to a block to which the LFNST 8 x 8 is applied, that is, for all blocks in (d) and (e) of Figure 9, the entire region to which the LFNST is not applied can be filled with zeros.

[0193] (iii) Due to the zero-out presented in (ii) above, the area filled with zeros may not be the same when LFNST is applied. Thus, it is possible to verify whether non-zero data exist according to the zero-out proposed in (ii) in a wider area than the LFNST case in Figure 9.

[0194] For example, when (ii)-(B) is applied, whether non-zero data exists can be checked in up to a region additionally filled with 0s in Figure 11 in addition to the regions filled with zeros in Figure (d) and ( e) of Figure 9, and then an LFNST index can be signaled only when there is no non-zero data.

[0195] even if the zero-out proposed in (ii) is applied, it is possible to check whether non-zero data exists in the same way as the existing LFNST index signaling. That is, after checking whether non-zero data exists in the zero-filled block in Figure 9, the LFNST index flag can be applied. In this case, the encoding device only performs zero output and the decoding device does not assume zero output, i.e., only checking whether non-zero data exists only in the area explicitly marked as 0 in FIG. 9, can perform LFNST index analysis.

[0196] Various embodiments in which combinations of the simplification methods ((i), (ii)-(A), (ii)-(B), (iii)) for the LFNST are applied can be derived. Obviously, combinations of the above simplification methods are not limited to the following embodiment and any combination can be applied to LFNST.

Modality

[0197] - Limit the number of output data for direct LFNST to a maximum of 16 ^ (i)

[0198] - When LFNST 4 x 4 is applied, all areas in which LFNST 4 x 4 is not applied are zero-out ^ (ii)-(A)

[0199] - When LFNST 8 x 8 is applied, all areas in which LFNST 8 x 8 is not applied are reset to zero ^ (ii)-(B)

[0200] - After checking if the non-zero data there is also the existing area filled with zero value and the area filled with zeros due to additional zero-outs ((ii)-(A), (ii)-(B) ), the LFNST index is signaled only when non-zero data does not exist ^ (iii)

[0201] In the embodiment, when LFNST is applied, a region in which non-zero output data may exist is limited to the interior of an upper left 4 x 4 region. Specifically, in (a) of Figure 10 and (a) of Figure 11, an eighth position in a scan order is the last position where non-zero data can exist, and in (b) and (d) of Figure 10 and (b) from Figure 11, a sixteenth position in the scan order (i.e., a lower right position of a upper left 4 x 4 block) is the last position where non-zero data can exist.

[0202] Therefore, when LFNST is applied, after checking whether non-zero data exists in a position where the residual encoding process is not allowed (in a position beyond the last position), it can be determined whether the index of LFNST is flagged.

[0203] In the case of the zero-out method proposed in (ii), since the number of data finally generated when the primary transform and LFNST are applied, the amount of computation required to perform the entire transform process can be reduced . That is, when LFNST is applied, since zero-out is applied to direct primary transform output data existing in a region in which LFNST is not applied, there is no need to generate data for the region that becomes zero-out during performing the direct primary transform. Thus, it is possible to reduce the amount of computation required to generate the corresponding data. The additional effects of the zero-out method proposed in (ii) are summarized below.

[0204] First, as described above, the amount of computation required to perform the entire transform process is reduced.

[0205] In particular, when (ii)-(B) is applied, the amount of calculation for the worst case is reduced, so that the transformation process can be speeded up. In other words, in general, a large amount of computation is required to perform a large primary transform. By applying (ii)-(B), the number of data derived as a result of performing direct LFNST can be reduced to 16 or less. Furthermore, as the size of the entire block (TU or CU) increases, the effect of reducing the amount of transform operation increases further.

[0206] Second, the amount of computation required for the entire transform process can be reduced, thereby reducing the energy consumption required to perform the transform.

[0207] Third, the latency involved in the transform process is reduced.

[0208] The secondary transform, such as the LFNST, adds a computational amount to the existing primary transform, thereby increasing the overall delay time involved in executing the transform. In particular, in the case of intraprediction, since reconstructed data from neighboring blocks are used in the prediction process, during encoding, an increase in latency due to a secondary transform leads to an increase in latency until reconstruction. This can lead to an increase in overall intraprediction coding latency.

[0209] However, if the zero-out suggested in (ii) is applied, the delay time for performing the primary transform can be greatly reduced. When LFNST is applied, the delay time for the entire transform is maintained or reduced , so that the coding apparatus can be implemented more simply.

[0210] Meanwhile, in conventional intraprediction, a coding target block is considered as a coding unit, and coding is performed without partitioning it. However, ISP (Intra Sub-Partitions) coding refers to performing intra-prediction coding with the coding target block being partitioned in a horizontal or vertical direction. In this case, a reconstructed block can be generated by performing encoding/decoding on partitioned block units, and the reconstructed block can be used as a reference block of the next partitioned block. According to an example, in ISP coding, a coding block can be partitioned into two or four sub-blocks and encoded, and in ISP, intraprediction is performed on a sub-block by referring to the reconstructed pixel value of a sub-block located adjacent to the left or upper side of it. Hereinafter, the term “encoding” may be used as a concept including both the encoding performed by the encoding apparatus and the decoding performed by the decoding apparatus.

[0211] The ISP partitions a block predicted as intra luma into two or four sub-partitions in the vertical or horizontal direction according to the size of the block. For example, the minimum block size that ISP can be applied to is 4x8 or 8x4. If the block size is greater than 4x8 or 8x4, the block will be partitioned into four sub-partitionings.

[0212] When ISP is applied, the sub-blocks are encoded sequentially according to the partition type, such as, horizontally or vertically, from left to right, or top to bottom, and the encoding for the next sub-block is block can be performed after performing a restoration process through inverse transform and intraprediction for a sub-block. To the leftmost or topmost sub-block, as in a conventional intraprediction method, the reconstructed pixel of the coding block that has already been encoded is referred. Furthermore, if the previous sub-block is not adjacent to each side of an internal sub-block that follows it, to derive reference pixels adjacent to a corresponding side, as in the conventional intraprediction method, the reconstructed pixel of an already block coded adjacent encoding is referred to.

[0213] In an ISP encoding mode, all sub-blocks may be encoded with the same intraprediction mode, and a flag indicating whether to use ISP encoding and a flag indicating a direction (horizontal or vertical) in which the partition runs can be flagged. Here, the number of sub-blocks can be adjusted to 2 or 4 depending on the block format. When the size (width x height) of a sub-block is less than 16, partitioning into the sub-blocks may not be allowed, or a restriction may be placed to not apply the ISP's own encoding.

[0214] Meanwhile, in the case of ISP prediction mode, a coding unit is partitioned into two or four partition blocks, i.e., sub-blocks, and predicted, and the same intra-prediction mode is applied to the two or four blocks so partitioned partition blocks.

[0215] As described above, both a horizontal direction (if an M x N coding unit with a horizontal length and a vertical length of M and N, respectively, is divided in the horizontal direction, it is divided into M x blocks (N /2) when divided into two, and into M x (N/4) blocks when divided into four) and a vertical direction (if the M x N coding unit is divided in the vertical direction, it will be divided into blocks (M/ 2)xN when divided into two, and divided into (M/4)xN blocks when divided into four) are possible as partition direction. When partitioned in the horizontal direction, the partition blocks are encoded in a top-to-bottom order, and when partitioned in the vertical direction, the partition blocks are encoded in a left-to-right order. The currently encoded partition block can be predicted by reference to the reconstructed pixel values of the top (left) partition block in the case of horizontal (vertical) direction partition.

[0216] The transform can be applied to the residual signal generated by the ISP prediction method in partition block units. The MTS (Multiple Transform Selection) technology based on the DST-7/DCT-8 combination, as well as the existing DCT-2, can be applied to the primary transform (core transform or primary transform) based on the forward direction and a LFNST (Low Frequency Non-Separable Transform) can be applied to a transform coefficient generated according to the primary transform to generate a final modified transform coefficient.

[0217] That is, LFNST can also be applied to partition blocks divided by applying the ISP prediction mode, and the same intraprediction mode is applied to partition blocks divided as described above. Therefore, by selecting the derived LFNST set based on the intraprediction mode, the derived LFNST set can be applied to all partition blocks. That is, the same intraprediction mode is applied to all partition blocks, and therefore the same set of LFNST can be applied to all partition blocks.

[0218] Meanwhile, according to an example, LFNST can only be applied to transform blocks with horizontal and vertical lengths of 4 or more. Therefore, when the horizontal or vertical length of the partition block partitioned according to the ISP prediction method is less than 4, LFNST will not be applied and the LFNST index will not be flagged. Furthermore, when LFNST is applied to each partition block, the corresponding partition block can be considered as a transform block. Obviously, when the ISP prediction method is not applied, LFNST can be applied to the coding block.

[0219] The application of LFNST to each partition block is described in detail below.

[0220] According to an example, after applying direct LFNST to an individual partition block, and after leaving only up to 16 coefficients (8 or 16) in the upper left 4 x 4 region according to the scan order of the transform coefficient, zero-out filling all remaining positions and regions with a value of 0 can be applied.

[0221] Alternatively, according to an example, when the length of one side of the partition block is 4, LFNST is applied only to the upper left 4 x 4 region, and when the length of all sides of the partition block, that is, the width and height are 8 or more, LFNST can be applied to the remaining 48 coefficients, except for a 4 x 4 lower-right region in an 8 x 8 upper-left region.

[0222] Alternatively, according to an example, in order to adjust the worst-case computational complexity for 8 multiplications per sample, when each partition block is 4 x 4 or 8 x 8, only 8 transform coefficients can be output after the application of direct LFNST. That is, if the partition block is 4 x 4, an 8 x 16 matrix can be applied as the transform matrix, and if the partition block is 8 x 8, an 8 x 48 matrix can be applied as the transform matrix.

[0223] Meanwhile, in the current VVC standard, LFNST index signaling is performed in units of coding units. Consequently, when ISP prediction mode is used and LFNST is applied to all partition blocks, then the same LFNST index value can be applied to the corresponding partition blocks. That is, when the LFNST index value is transmitted once at the coding unit level, the corresponding LFNST index can be applied to all partition blocks in the coding unit. As described above, the LFNST index value can have values of 0, 1, and 2, 0 indicates a case where LFNST is not applied, and 1 and 2 indicate two transform matrices present in an LFNST set when LFNST is applied. applied.

[0224] As described above, the LFNST set is determined by the intra-prediction mode, and since all partition blocks in the coding unit are predicted in the same intra-prediction mode in the case of the ISP prediction mode, the partition blocks partition may refer to the same set of LFNST.

[0225] As another example, LFNST index signaling is still performed in units of coding units, but in the case of ISP prediction mode, without determining whether or not to apply LFNST uniformly to all partition blocks, Whether to apply the signaled LFNST index value at the coding unit level for each partition block or not to apply LFNST can be determined through a separate condition. Here, the separate condition can be signaled in the form of a flag for each partition block through the bit stream, and when the value of the flag is 1, the LFNST index value signaled at the coding unit level can be applied and when the flag value is 0, LFNST cannot be applied.

[0226] Next, a method for maintaining computational complexity for the worst case when LFNST is applied to ISP mode will be described.

[0227] In the case of ISP mode, to keep the number of multiplications per sample (or per coefficient, or per position) at a certain value or less when LFNST is applied, the application of LFNST can be restricted. Depending on the size of the partition block, the number of multiplications per sample (either per coefficient or per position) can be kept to 8 or less by applying LFNST as follows.

[0228] 1. When the horizontal length and vertical length of the partition block are equal to or greater than 4, the same method as the worst-case calculation complexity adjustment method for LFNST in the current VVC standard can be applied.

[0229] That is, when the partition block is a 4 x 4 block, instead of a 16 x 16 matrix, in the forward sense, an 8 x 16 matrix obtained by sampling the first 8 rows of a 16 x 16 matrix can be applied, and in the opposite direction, a 16 x 8 matrix obtained by sampling the 8 left columns of a 16 x 16 matrix can be applied. Furthermore, when the partition block is an 8 x 8 block, in the forward direction, instead of a 16 x 48 matrix, an 8 x 48 matrix obtained by sampling the first 8 rows of a 16 x 48 matrix can be applied and in the reverse direction, instead of a 48 x 16 matrix, a 48 x 8 matrix obtained by sampling the left 8 columns of a 48 x 16 matrix can be applied.

[0230] In the case of a 4 x N or N x 4 block (N > 4), when the direct transform is performed, 16 coefficients generated after applying a 16x16 matrix only in the upper left 4 x 4 block are arranged in the region 4 x 4 upper left, and the other regions can be filled with 0 values. Furthermore, when performing the inverse transform, 16 coefficients located in the upper left 4 x 4 block can be arranged in scan order to configure an input vector and , then 16 output data can be generated by multiplying the 16 x 16 matrix. The generated output data can be arranged in the upper left 4 x 4 region, and the remaining regions except the upper left 4 x 4 region can be filled with zeros.

[0231] In the case of an 8 x N or N x 8 (N > 8) block, when the direct transform is performed, 16 coefficients generated after applying the 16 x 48 matrix only to the ROI region in the upper 8 x 8 block left (remaining regions excluding the lower right 4 x 4 block from the upper left 8 x 8 block) can be arranged in the upper left 4 x 4 area and the other regions can be filled with 0 values. Furthermore, when performing the inverse transform, 16 coefficients located in the upper left 4 x 4 block can be arranged in scan order to configure an input vector, and then 48 output data can be generated by multiplying the 48 x 16 matrix. in the ROI region and the other regions can be filled with 0 values.

[0232] As another example, in order to keep the number of multiplications per sample (or per coefficient, or per position) at a certain value or less, the number of multiplications per sample (or per coefficient, or per position) based on in ISP coding unit size instead of ISP partition block size can be kept at 8 or less. If there is only one block among the ISP partition blocks, which satisfies the condition under which LFNST is applied, the complexity calculation for the worst-case LFNST can be applied based on the size of the corresponding coding unit, rather than of the partition block size. For example, when a luma coding block for a given coding unit is partitioned into 4 partition blocks of size 4 x 4 and encoded by the ISP, and when there is no non-zero transform coefficient for two partition blocks between them, the other two partition blocks can be configured respectively to generate 16 transform coefficients instead of 8 (based on the encoder).

[0233] Next, a method of signaling the LFNST index in the case of ISP mode will be described.

[0234] As described above, the LFNST index can have values of 0, 1 and 2, where 0 indicates that LFNST is not applied and 1 and 2 respectively indicate one of the two LFNST kernel matrices included in the selected LFNST set . LFNST is applied based on the LFNST kernel matrix selected by the LFNST index. A method of transmitting the LFNST index in the current VVC standard will be described below.

[0235] 1. An LFNST index may be transmitted once for each coding unit (CU), and in the case of a dual tree, individual LFNST indexes may be signaled to a luma block and a chroma block, respectively.

[0236] 2. When the LFNST index is unsigned, the LFNST index value is inferred as a default value of 0. The case in which the LFNST index value is inferred as 0 is as follows.

[0237] A. In the case of a mode in which no transform is applied (e.g. transform jumping, BDPCM, lossless coding, etc.)

[0238] B. When the primary transform is not DCT-2 (DST7 or DCT8), that is, when the transform in the horizontal direction or the transform in the vertical direction is not DCT-2.

[0239] C. When the horizontal length or the vertical length for the luma block of the coding unit exceeds the size of the maximum transformable luma transform, for example, when the size of the maximum transformable luma transform is 64, and when the size for the luma block of the encoding block is equal to 128 x 16, LFNST cannot be applied.

[0240] In the case of the dual tree, it is determined whether or not the size of the maximum luma transform is exceeded for each coding unit for the luma component and the coding unit for the chroma component. That is, it checks for the luma block whether the size of the maximum transformable luma transform is exceeded, and it checks for the chroma block whether the horizontal/vertical length of the corresponding luma block for the color format and the size of the luma transform maximum transformable exceeds the size are exceeded. For example, when the color format is 4:2:0, the horizontal/vertical length of the corresponding luma block is twice that of the corresponding chroma block, and the transform size of the corresponding luma block is twice that of the corresponding chroma block. As another example, when the color format is 4:4:4, the horizontal/vertical length and transform size of the corresponding luma block are the same as the corresponding chroma block.

[0241] A 64-length transform or a 32-length transform may mean a transform applied to the width or height with a length of 64 or 32, respectively, and “transform size” may mean 64 or 32 as the corresponding length.

[0242] In the case of a single tree, after checking whether a horizontal length or a vertical length of a luma block exceeds the maximum size of the transformable luma block, if it does, the LFNST index flag may be omitted.

[0243] D. The LFNST index can be transmitted only when the horizontal length and vertical length of the coding unit are equal to or greater than 4.

[0244] In the case of a double tree, the LFNST index can be signaled only when the horizontal and vertical lengths of a corresponding component (i.e., a luma or chroma component) are equal to or greater than 4.

[0245] In the case of a single tree, the LFNST index may be signaled when the horizontal and vertical lengths for the luma component are equal to or greater than 4.

[0246] E. If the position of the last non-zero coefficient is not a DC position (upper left position of the block), and if the position of the last non-zero coefficient is not a DC position, in the case of a luma block of a double tree type, the LFNST index is transmitted. In the case of a double tree chroma block, if any of the positions of the last non-zero coefficient for Cb and the position of the last non-zero coefficient for Cr is not a DC position, the corresponding LNFST index is transmitted.

[0247] In the case of the simple tree type, if the position of the last non-zero coefficient of any of the luma components, Cb component and Cr component is not the DC position, the LFNST index is transmitted.

[0248] Here, if a coded block flag (CBF) value indicating whether or not there is a transform coefficient for a transform block is 0, the position of the last non-zero coefficient for the corresponding transform block is not checked to determine whether or not the LFNST index is signaled. That is, when the corresponding CBF value is 0, since no transform is applied to the corresponding block, the position of the last non-zero coefficient may not be considered when checking the signaling condition of the LFNST index.

[0249] For example, 1) in the case of a double-tree type and a luma component, if the corresponding CBF value is 0, the LFNST index is not signaled, 2) in the case of a double-tree type and chroma component, if the CBF value for Cb is 0 and the CBF value for Cr is 1, only the position of the last non-zero coefficient for Cr is checked and the corresponding LFNST index is transmitted, 3) in the case of a single tree type, the position of the last nonzero coefficient is checked only for components with a CBF value of 1 each of luma, Cb, and Cr.

[0250] F. When it is confirmed that the transform coefficient exists at a position other than a position where the LFNST transform coefficient may exist, the LFNST index flag may be omitted. In the case of a 4 x 4 transform block and an 8 x 8 transform block, the LFNST transform coefficients can exist in eight positions of the DC position according to the sweep order of the transform coefficient in the VVC pattern, and the remaining positions are filled with zeros. Furthermore, when the 4 x 4 transform block and the 8 x 8 transform block are not, the LFNST transform coefficients can exist at sixteen positions of the DC position according to the scan order of the transform coefficient in the VVC pattern. , and the remaining positions are filled with zeros.

[0251] Thus, if non-zero transform coefficients exist in the region that is to be filled with the zero value after residual coding progresses, the LFNST index flag may be omitted.

[0252] Meanwhile, the ISP mode can also be applied to the luma block only, or it can be applied to both the luma block and the chroma block. As described above, when ISP prediction is applied, the corresponding coding unit can be divided into two or four partition and prediction blocks, and a transform can be applied to each of the partition blocks. Therefore, also when determining a condition for signaling the LFNST index in coding unit units, it is necessary to take into account the fact that LFNST can be applied to the respective partition blocks. Furthermore, when the ISP prediction mode is applied only to a specific component (for example, a luma block), the LFNST index must be flagged considering the fact that only the component is divided into partition blocks. The LFNST index signaling methods available in ISP mode are summarized below.

[0253] 1. An LFNST index may be transmitted once for each coding unit (CU), and in the case of a dual tree, individual LFNST indexes may be signaled to a luma block and a chroma block, respectively.

[0254] 2. When the LFNST index is unsigned, the LFNST index value is inferred as a default value of 0. The case in which the LFNST index value is inferred as 0 is as follows.

[0255] A. In the case of a mode in which no transform is applied (e.g. transform jumping, BDPCM, lossless coding, etc.)

[0256] B. When the horizontal length or the vertical length for the luma block of the coding unit exceeds the size of the maximum transformable luma transform, for example, when the size of the maximum transformable luma transform is 64, and when the size for the luma block of the encoding block is equal to 128 x 16, LFNST cannot be applied.

[0257] Whether or not to flag the LFNST index can be determined based on the size of the partition block rather than the coding unit. That is, if the horizontal or vertical length of the partition block for the corresponding luma block exceeds the size of the maximum transformable luma transform, the LFNST index flag can be omitted and the LFNST index value can be inferred to be 0.

[0258] In the case of the dual tree, it is determined whether or not the size of the maximum luma transform is exceeded for each coding unit or partition block for the luma component and the coding unit or partition block for the chroma component. That is, if the horizontal and vertical lengths of the coding unit or partition block for luma are compared with the maximum size of the luma transform, respectively, and at least one of them is greater than the maximum size of the luma transform, the LFNST is not applied, and in the case of a coding unit or partition block for chroma, the horizontal/vertical length of the corresponding luma block for the color format and the size of the maximum transformable luma transform are compared. For example, when the color format is 4:2:0, the horizontal/vertical length of the corresponding luma block is twice that of the corresponding chroma block, and the transform size of the corresponding luma block is twice that of the corresponding chroma block. As another example, when the color format is 4:4:4, the horizontal/vertical length and size of the transform and the corresponding luma block are the same as the corresponding chroma block.

[0259] In the case of a single tree, after checking whether a horizontal length or a vertical length for a luma block (coding unit or partition block) exceeds the maximum size of the transformable luma transform block, if it exceeds, the flag of the LFNST index can be omitted.

[0260] C. If the LFNST included in the current VVC standard is applied, the LFNST index can be transmitted only when the horizontal length and vertical length of the partition block are equal to or greater than 4.

[0261] If the LFNST for the 2 x M (1 x M) or M x 2 (M x 1) block is applied in addition to the LFNST included in the current VVC standard, the LFNST index can be transmitted only when the block size partition is equal to or greater than a 2 x M (1 x M) or M x 2 (M x 1) block. Here, the expression “the P x Q block is equal to or greater than the R x S block” means that P > R and Q > S.

[0262] In summary, the LFNST index can be transmitted only when the partition block is equal to or greater than the minimum size to which the LFNST is applicable. In the case of a dual tree, the LFNST index can be signaled only when the partition block for the luma or chroma component is equal to or greater than the minimum size to which the LFNST is applicable. In the case of a single tree, the LFNST index can be signaled only when the partition block for the luma component is equal to or greater than the minimum size to which the LFNST is applicable.

[0263] In this document, the expression “the M x N block is greater than or equal to the K x L block” means that M is greater than or equal to K and N is greater than or equal to L. The expression “the M x N block is greater than the block K x L” means that M is greater than or equal to K and N is greater than or equal to L, and that M is greater than K or N is greater than L. The expression “the block M x N smaller or equal to block K x L” means that M is less than or equal to K and N is less than or equal to L, while the expression “block M x N is less than block K x L” means that M is less than or equal to K and N is less than or equal to L, and that M is less than K or N is less than L.

[0264] D. If the position of the last non-zero coefficient is not a DC position (upper left block position), and if the position of the last non-zero coefficient is not a DC position in any of all the blocks of partition In the case of a double-tree luma block, the LFNST index is transmitted. In the case of a dual tree type and a chroma block, if at least one of the positions of the last non-zero coefficient of all partition blocks for Cb (if ISP mode is not applied to the chroma component, the number of partition blocks is considered one) and the position of the last non-zero coefficient of all partition blocks for Cr (if ISP mode is not applied to the chroma component, the number of partition blocks is considered one) is not a Position DC, the corresponding LNFST index can be transmitted.

[0265] In the case of the simple tree type, if the position of the last non-zero coefficient of any of all partition blocks for the luma component, the Cb component and the Cr component is not the DC position, the corresponding LFNST index can be transmitted.

[0266] Here, if the value of the coded block flag (CBF) indicating whether there is a transform coefficient for each partition block is 0, the position of the last non-zero coefficient for the corresponding partition block is not checked to determine whether or not the LFNST index is flagged. That is, when the corresponding CBF value is 0, since no transform is applied to the corresponding block, the position of the last non-zero coefficient for the corresponding partition block is not considered when checking the signaling condition of the LFNST index .

[0267] For example, 1) in the case of a dual-tree type and a luma component, if the corresponding CBF value for each partition block is 0, the partition block is excluded when determining whether or not to signal the LFNST index, 2 ) in the case of a dual tree type and a chroma component, if the CBF value for Cb is 0 and the CBF value for Cr is 1 for each partition block, only the position of the last coefficient non-zero for Cr is checked to determine whether or not to flag the LFNST index, 3) in the case of the simple tree type, it is possible to determine whether or not to flag the LFNST index by checking the position of the last non-zero coefficient only for blocks with a CBF value of 1 for all partition blocks of the luma component, the Cb component, and the Cr component.

[0268] In the case of ISP mode, the image information can also be configured so that the position of the last non-zero coefficient is not checked, and one embodiment of the same is as follows.

[0269] i. In the case of ISP mode, LFNST index signaling can be allowed without checking the position of the last non-zero coefficient for both the luma block and the chroma block. That is, even if the position of the last non-zero coefficient for all partition blocks is the DC position or the corresponding CBF value is 0, LFNST index signaling can be allowed.

[0270] ii. In the case of ISP mode, checking the position of the last non-zero coefficient can be omitted only for the luma block, and in the case of the chroma block, checking the position of the last non-zero coefficient can be performed in the manner described above . For example, in the case of a double-tree type and a luma block, LFNST index flagging is allowed without checking the position of the last non-zero coefficient, and in the case of a double-tree type and a chroma block, either or is not flagged a corresponding LFNST index can be determined by checking whether or not a DC position exists for the position of the last non-zero coefficient in the manner described above.

[0271] iii. In case of ISP mode and simple tree type, method i or ii can be applied. That is, in the case of ISP mode and when the number i is applied to the simple tree type, it is possible to omit checking the position of the last non-zero coefficient for both the luma block and the chroma block and allow signaling of LFNST index. Alternatively, applying section ii, for the partition blocks for the luma component, checking the position of the last non-zero coefficient is omitted, and for the partition blocks for the chroma component (if ISP is not applied for the component chroma, the number of partition blocks can be taken as 1), the position of the last non-zero coefficient is checked in the manner described above, thus determining whether or not to flag the LFNST index.

[0272] E. When it is confirmed that the transform coefficient exists at a position other than a position where the LFNST transform coefficient can exist even for a partition block among all partition blocks, signaling of the LFNST index can be omitted.

[0273] For example, in the case of a 4 x 4 partition block and an 8 x 8 partition block, the LFNST transform coefficients may exist in eight positions of the DC position according to the scan order of the transform coefficient in the VVC pattern, and the remaining positions are filled with zeros. Furthermore, if it is equal to or greater than 4x4 and is neither a 4x4 partition block nor an 8x8 partition block, the LFNST transform coefficients can exist in 16 positions of the DC position according to the order of the transform coefficient in the VVC pattern, and all remaining positions are filled with zeros.

[0274] Thus, if non-zero transform coefficients exist in the region that is to be filled with the zero value after residual coding progresses, the LFNST index flag may be omitted.

[0275] Meanwhile, in the case of ISP mode, the length condition is displayed independently for the horizontal direction and the vertical direction, and DST-7 is applied instead of DCT-2 without signaling for the MTS index. It is determined whether the horizontal or vertical length is greater than or equal to 4 and less than or equal to 16, and a primary transform kernel is determined according to the determination result. Therefore, in the case of ISP mode, when LFNST can be applied, the following transform combination configuration is possible.

[0276] 1. When the LFNST index is 0 (including the case where the LFNST index is inferred to be 0), the primary transform decision condition at the time of the ISP included in the current VVC standard can be followed. In other words, one can check whether the length condition (being equal to or greater than 4 or equal to or less than 16) is satisfied independently for the horizontal and vertical directions, respectively, and if it is satisfied, DST-7 can be applied instead of DCT-2 for primary transform, while if not satisfied, DCT-2 can be applied.

[0277] 2. For a case where the LFNST index is greater than 0, the following two configurations may be possible as a primary transform.

[0278] A. DCT-2 can be applied in horizontal and vertical directions.

[0279] B. The primary transform decision condition at the time of the ISP included in the current VVC standard can be followed. In other words, one can check whether the length condition (being equal to or greater than 4 or equal to or less than 16) is satisfied independently for the horizontal and vertical directions respectively, and if it is satisfied, DST-7 can be applied instead of DCT-2, while if not satisfied, DCT-2 can be applied.

[0280] In ISP mode, the image information can be configured so that the LFNST index is transmitted for each partition block instead of for each coding unit. In this case, in the LFNST index signaling method described above, it can be considered that only one partition block exists on a drive on which the LFNST index is transmitted, and it can be determined whether or not to signal the LFNST index.

[0281] Meanwhile, the signaling order of the LFNST index and the MTS index will be described below.

[0282] According to an example, the LFNST index signaled in the residual encoding may be encoded after the encoding position for the last non-zero coefficient position, and the MTS index may be encoded immediately after the LFNST index. In the case of this configuration, the LFNST index can be signaled for each transform unit. Alternatively, even if unsigned in residual coding, the LFNST index can be coded after coding to the last significant coefficient position, and the MTS index can be coded after the LFNST index.

[0283] The syntax of residual encoding according to an example is as follows. Table 3

[0284] The meanings of the main variables shown in Table 3 are as follows.

[0285] 1. cbWidth, cbHeight: the width and height of the current coding block

[0286] 2. log2TbWidth, log2TbHeight: The base-2 log value for the width and height of the current transform block may be reduced, reflecting zero-out, to an upper-left region in which a non-zero coefficient may exist.

[0287] 3. sps_lfnst_enabled_flag: a flag indicating whether LFNST is enabled or not, if the flag value is 0, it indicates that LFNST is not enabled, and if the flag value is 1, it indicates that LFNST is enabled. It is defined in the Sequence Parameter Set (SPS).

[0288] 4. CuPredMode[chType][x0][y0]: a prediction mode corresponding to the variable chType and position (x0, y0), chType can have values of 0 and 1, where 0 indicates a luma component and 1 indicates a chroma component. The position (x0, y0) indicates a position in the engraving, and MODE_INTRA (intraprediction) and MODE_INTER (interprediction) are possible as a value of CuPredMode[chType][x0][y0].

[0289] 5. IntraSubPartitionsSplit[x0][y0]: the contents of position (x0, y0) are the same as No. 4. Indicates which ISP partition at position (x0, y0) is applied, ISP_NO_SPLIT indicates which unit of coding corresponding to position (x0, y0) is not divided into partition blocks.

[0290] 6. intra_mip_flag[x0][y0]: the content of position (x0, y0) is the same as No. 4 above. The intra_mip_flag is a flag that indicates whether a matrix-based intraprediction (MIP) prediction mode is applied or not. If the flag value is 0, it indicates that MIP is not enabled, and if the flag value is 1, it indicates that MIP is enabled.

[0291] 7. cIdx: the value of 0 indicates luma, and the values of 1 and 2 indicate Cb and Cr which are respectively chroma components.

[0292] 8. treeType: indicates single tree and double tree, etc. (SINGLE_TREE: single tree, DUAL_TREE_LUMA: double tree for luma component, DUAL_TREE_CHROMA: double tree for chroma component)

[0293] 9. tu_cbf_cb[x0][y0]: the content of position (x0, y0) is the same as no. 4. Indicates the coded block flag (CBF) for the Cb component. If its value is 0, it means that no non-zero coefficients are present in the corresponding transform unit for the Cb component, and if its value is 1, it indicates that non-zero coefficients are present in the corresponding transform unit for the Cb component. CB.

[0294] 10. lastSubBlock: Indicates a position in the scanning order of a sub-block (Coefficient Group (CG)) in which the last non-zero coefficient is located. 0 indicates a sub-block in which the DC component is included and, if greater than 0, it is not a sub-block in which the DC component is included.

[0295] 11. lastScanPos: Indicates the position where the last significant coefficient is in the scan order within a sub-block. If a sub-block includes 16 positions, values from 0 to 15 are possible.

[0296] 12. lfnst_idx[x0][y0]: LFNST index syntax element to be parsed. If not parsed, it is inferred as a value of 0. That is, the default value is set to 0, indicating that LFNST is not applied.

[0297] 13. LastSignificantCoeffX, LastSignificantCoeffY: Indicate the x and y coordinates where the last significant coefficient is found in the transform block. The x coordinate starts at 0 and increases from left to right, and the y coordinate starts at 0 and increases from top to bottom. If the values of both variables are 0, it means that the last significant coefficient is located in DC.

[0298] 14. cu_sbt_flag: A flag that indicates whether the subblock transform (SBT) included in the current VVC pattern is enabled. If the flag value is 0, it indicates that SBT is not enabled, and if the flag value is 1, it indicates that SBT is enabled.

[0299] 15.sps_explicit_mts_inter_enabled_flag, sps_explicit_mts_intra_enabled_flag: Flags that indicate whether or not explicit MTS is applied to inter CU and intra CU, respectively. If a corresponding flag value is 0, it indicates that MTS is not enabled for an inter or intra CU, and if the corresponding flag value is 1, it indicates that MTS is enabled.

[0300] 16. tu_mts_idx[x0][y0]: MTS index syntax element to be parsed. If not analyzed, it is inferred as a value of 0. That is, the default value is set to 0, indicating that the DCT-2 is enabled for both horizontal and vertical directions.

[0301] As shown in Table 3, in the case of a single tree, it is possible to determine whether or not to signal the LFNST index using only the last significant coefficient position condition for luma. That is, if the position of the last significant coefficient is not DC and the last significant coefficient exists in the upper left sub-block (CG), for example, a 4 x 4 block, then the LFNST index is flagged. In this case, in the case of the 4 x 4 transform block and the 8 x 8 transform block, the LFNST index is flagged only when the last significant coefficient exists in positions 0 to 7 in the upper left sub-block.

[0302] In the case of dual tree, the LFNST index is signaled independently of each of luma and chroma, and in the case of chroma, the LFNST index may be signaled by applying the last significant coefficient position condition only to the component of CB. The corresponding condition cannot be checked for the Cr component, and if the CBF value for Cb is 0, the LFNST index can be flagged by applying the last significant coefficient position condition to the Cr component.

[0303] 'Min(log2TbWidth, log2TbHeight) >= 2' of Table 8 can be expressed as “Min(tbWidth, tbHeight) >= 4”, and 'Min(log2TbWidth, log2TbHeight) >= 4' can be expressed as “ Min(tbWidth, tbHeight) >= 16”.

[0304] In Table 3, log2ZoTbWidth and log2ZoTbHeight mean log values whose base width and height are 2 (base-2) for the upper left region where the last significant coefficient can exist by zero-out, respectively.

[0305] As shown in Table 3, the log2ZoTbWidth and log2ZoTbHeight values can be updated in two places. The first is before the MTS index or LFNST index value is analyzed, and the second is after the MTS index is analyzed.

[0306] The first update is before the MTS index value (tu_mts_idx[x0][y0]) is parsed, so log2ZoTbWidth and log2ZoTbHeight can be set independently of the MTS index value.

[0307] After the MTS index is analyzed, log2ZoTbWidth and log2ZoTbHeigh are set to an MTS index greater than 0 (DST-7/DCT-8 combination). When DST-7/DCT-8 is applied independently in each of the horizontal and vertical directions in the primary transform, there can be up to 16 significant coefficients per row or column in each direction. That is, after applying DST-7/DCT-8 with a length of 32 or greater, up to 16 transform coefficients can be derived for each left or top row or column. Thus, in a 2D block, when DST-7/DCT-8 is applied in both the horizontal and vertical directions, significant coefficients can exist in an upper left region of only 16 x 16. [386] Furthermore, when the DCT-2 is independently applied in each of the horizontal and vertical directions in the current primary transform, there can be up to 32 significant coefficients per row or column in each direction. That is, when applying DCT-2 with a length of 64 or greater, up to 32 transform coefficients can be derived for each left or top row or column. Thus, in a 2D block, when DCT-2 is applied in both the horizontal and vertical directions, significant coefficients may exist in a 32 x 32 top left region.

[0308] Furthermore, when DST-7/DCT-8 is applied on one side and DCT-2 is applied on the other side in the horizontal and vertical directions, 16 significant coefficients may exist in the former direction and 32 significant coefficients may exist in the last direction. For example, in the case of a 64 x 8 transform block, if DCT-2 is applied in the horizontal direction and DST-7 is applied in the vertical direction (may occur in a situation where implicit MTS is applied), a Significant coefficient may exist in up to a 32 x 8 upper left region.

[0309] If, as shown in Table 8, log2ZoTbWidth and log2ZoTbHeight are updated in two places, that is, before analyzing the MTS index, the ranges of last_sig_coeff_x_prefix and last_sig_coeff_y_prefix can be determined by log2ZoTbWidth and log2ZoTbHeight as shown in the table below. Table 4

[0310] Furthermore, in this case, the maximum values of last_sig_coeff_x_prefix and last_sig_coeff_y_prefix can be set by reflecting the log2ZoTbWidth and log2ZoTbHeight values in the binarization process for last_sig_coeff_x_prefix and last_sig_coeff_y_prefix. Table 5

[0311] According to an example, in a case where ISP mode and LFNST are applied, when the signaling of Table 3 is applied, the specification text can be configured as shown in Table 6. Compared with Table 3 , the LFNST index signaling condition only in a case excluding ISP mode (IntraSubPartitionsSplit[x0][y0] = = ISP_NO_SPLIT in Table 3) is deleted.

[0312] In a single tree, when an LFNST index passed to a luma component (cIdx = 0) is reused for a chroma component, an LFNST index passed to a first ISP partition block in which there is a significant coefficient can be applied to a chroma transform block. Alternatively, even in a single tree, an LFNST index can be flagged for a chroma component separately from that for a luma component. A description of the variables in Table 6 is the same as in Table 3. Table 6

[0313] According to an example, the LFNST index and/or the MTS index may be signaled at a coding unit level. As described above, the LFNST index can have three values of 0, 1, and 2, where 0 indicates that LFNST is not applied, and 1 and 2 respectively indicate a first candidate and a second candidate out of two LFNST kernel candidates. included in a selected LFNST set. The LFNST index is encoded using truncated unary binarization and the values of 0, 1, and 2 can be encoded as bin strings of 0, 10, and 11, respectively.

[0314] According to an example, LFNST can be applied only when DCT-2 is applied in both the horizontal and vertical directions in the primary transform. Therefore, if the MTS index is signaled after signaling the LFNST index, the MTS index can be signaled only when the LFNST index is 0, and primary transform can be performed by applying DCT-2 in the horizontal direction and in the vertical direction without MTS index signaling when LFNST index is not 0.

[0315] The MTS index may have values of 0, 1, 2, 3, and 4, where 0, 1, 2, 3, and 4 may indicate that DCT-2/DCT-2, DST-7/DST- 7, DCT-8/DST-7, DST-7/DCT-8, DCT-8/DCT-8 are applied in the horizontal and vertical directions, respectively. Furthermore, the MTS index can be encoded using truncated unitary binarization, and the values of 0, 1, 2, 3, and 4 can be encoded as bin sequences of 0, 10, 110, 1110, and 1111, respectively.

[0316] LFNST index signaling at a coding unit level may be indicated as shown in the following table. The LFNST index can be flagged in the last part of a coding unit syntax table. Table 7

[0317] A variable LfnstDcOnly and a variable LfnstZeroOutSigCoeffFlag in Table 7 can be defined as shown below in Table 10.

[0318] The variable LfnstDcOnly equals 1 when all last significant coefficients are located in DC positions (top left positions) for transform blocks with a coded block flag (CBF, equal to 0 if there is at least one significant coefficient in a corresponding block, and equals 0 otherwise) of 1, and equals 0 otherwise. Specifically, in a double-tree luma case, the position of the last significant coefficient is checked with respect to a luma transform block, and in a double-tree chroma case, the position of the last significant coefficient is checked with respect to a transform block for Cb and a transform block for Cr. In the case of a single tree, the position of the last significant coefficient can be checked against the transform block for luma, Cb and Cr.

[0319] The variable LfnstZeroOutSigCoeffFlag is equal to 0 if there is a significant coefficient at a zero-out position when LFNST is applied, and is equal to 1 otherwise.

[0320] lfnst_idx[x0][y0] included in Table 7 and subsequent tables indicates an LFNST index for a corresponding coding unit and tu_mts_idx[x0][y0] indicates an MTS index for the corresponding coding unit.

[0321] According to an example, to encode the MTS index consecutively after the LFNST index at the coding unit level, a coding unit syntax table can be configured as shown in Table 8. Table 8

[0322] Comparing Table 8 with Table 7, a condition to check whether the value of tu_mts_idx[x0][y0] is 0 (i.e., check whether DCT-2 is applied in the horizontal direction and in the vertical direction) in a condition for flagging lfnst_idx[x0][y0] is changed to a condition for checking whether the value of transform_skip_flag[x0][y0] is 0 (!transform_skip_flag[x0][y0]). transform_skip_flag[x0][y0] indicates whether the coding unit is coded in transform skip mode in which a transform is skipped and the flag is flagged before the MTS index and the LFNST index. That is, because lfnst_idx[x0][y0] is flagged before the value of tu_mtx_idx[x0][y0] is flagged, only the condition for the value of transform_skip_flag[x0][y0] can be checked. [393] As shown in Table 8, a plurality of conditions are checked when encoding tu_mts_idx[x0][y0], and tu_mts_idx[x0][y0] is flagged only when the value of lfnst_idx[x0][y0] is 0 as described above.

[0323] tu_cbf_luma[x0][y0] is a flag that indicates whether there is a significant coefficient for a luma component, and cbWidth and cbHeight indicate the width and height of the coding unit for the luma component, respectively.

[0324] In Table 8, ( IntraSubPartitionsSplit[ x0 ][ y0 ] = = ISP_NO_SPLIT ) indicates that ISP mode is not applied and ( !cu_sbt_flag ) indicates that no SBT is applied.

[0325] According to Table 8, when the width and height of the coding unit for the luma component are 32 or less, tu_mts_idx[x0][y0] is flagged, that is, whether an MTS is applied is determined by the width and height of the encoding unit for the luma component.

[0326] According to another example, when the transform block (TU) occurs (e.g., when a maximum transform size is set to 32, a 64 x 64 coding unit is divided into four transform blocks of 32 x 32 and encoded), the MTS index can be signaled based on the size of each transform block. For example, when the width and height of a transform block are 32 or less, the same MTS index value can be applied to all transform blocks in a coding unit, thereby applying the same primary transform. Additionally, when the transform block occurs side by side, the value of tu_cbf_luma[x0][y0] in Table 14 can be a CBF value for a top-left transform block or can be set to 1 when a CBF value for even one transform block out of all transform blocks is 1.

[0327] According to an example, when the ISP mode is applied to the current block, LFNST may be applied, in which case Table 8 may be changed as shown in Table 9. Table 9 [394] As shown in Table 9, even in ISP mode, (IntraSubPartitionsSplitType != ISP_NO_SPLIT) lfnst_idx[ x0 ][ y0 ] can be configured to be flagged and the same LFNST index value can be applied to all blocks of ISP partition.

[0328] Furthermore, as shown in Table 9, since tu_mts_idx[x0][y0] can be signaled only in one mode excluding ISP mode, an index encoding part of MTS is the same as in Table 8

[0329] As shown in Table 8 and Table 9, when the MTS index is signaled immediately after the LFNST index, information about the primary transform cannot be known when performing residual coding. That is, the MTS index is flagged after residual encoding. Consequently, in a residual coding part, a part in which a zeroing is performed leaving only 16 coefficients for DST-7 or DCT-8 with a length of 32 can be changed as shown below in Table 10. Table 10

[0330] As shown in Table 10, in a process of determining log2ZoTbWidth and log2ZoTbHeight (where log2ZoTbWidth and log2ZoTbHeight respectively denote the base 2 log values of the width and height of an upper left region remaining after zeroing is performed), the checking the value of tu_mts_idx[ x0 ][ y0 ] can be omitted.

[0331] The binarization of last_sig_coeff_x_prefix and last_sig_coeff_y_prefix in Table 16 can be determined based on log2ZoTbWidth and log2ZoTbHeight as shown in Table 5.

[0332] Additionally, as shown in Table 10, a sps_mts_enable_flag check condition can be added when determining log2ZoTbWidth and log2ZoTbHeight in the residual encoding.

[0333] Meanwhile, according to another example, the coding unit syntax table, the transform unit syntax table and the residual coding syntax table are as follows. According to Table 11, the MTS index moves from the transform unit level to the coding unit level syntax and is flagged after the LFNST index flag. Additionally, the restriction that disallows LFNST when ISP is applied to the encoding unit has been removed. When ISP is applied to the coding unit, the restriction that disallows LFNST is removed, so that LFNST can be applied to all intraprediction blocks. Furthermore, both the MTS index and the LFNST index are conditionally signaled at the last part of the coding unit level. Table 11

[0334] In Table 11, MtsZeroOutSigCoeffFlag is initially set to 1, and this value can be changed in residual coding in Table 13. The value of a MtsZeroOutSigCoeffFlag variable is changed from 1 to 0 when a significant coefficient exists in a region ( LastSignificantCoeffX > 15 | | LastSignificantCoeffY > 15 ) to be padded with 0s by a zero-out, in which case the MTS index is not signaled as shown in Table 11.

[0335] Meanwhile, as shown in Table 11, when tu_cbf_luma[ x0 ][ y0 ] is 0, the encoding mts_idx[ x0 ][ y0 ] can be omitted. That is, when the CBF value of the luma component is 0, since the transform is not applied, there is no need to flag the MTS index, and thus the encoding of the MTS index can be omitted.

[0336] According to an example, the above technical feature can be implemented in another conditional syntax. For example, after the MTS is performed, a variable indicating whether a significant coefficient exists in a region other than the DC region of the current block can be derived, and when the variable indicates that the significant coefficient exists in the region excluding the DC region, the index from MTS can be flagged. In other words, the existence of the significant coefficient in the region other than the DC region of the current block indicates that the value of tu_cbf_luma[ x0][y0] is 1, and in this case, the MTS index can be signaled.

[0337] The variable may be expressed as MtsDcOnly, and after the variable MtsDcOnly is initially set to 1 at the coding unit level, the value is changed to 0 when it is determined that the significant coefficient is present in the region, except for the region DC of the current block at the residual coding level. When the MtsDcOnly variable is 0, the image information can be configured so that the MTS index is flagged.

[0338] When tu_cbf_luma[ x0 ][ y0 ] is 0, since the residual encoding syntax is not called at the transform unit level of Table 12, the initial value of 1 of the MtsDcOnly variable is maintained. In this case, because the MtsDcOnly variable is not changed to 0, the image information can be configured so that the MTS index is not flagged. In other words, the MTS index is not analyzed and flagged.

[0339] Meanwhile, the decoding apparatus can determine the color index cIdx from the transform coefficient to derive the variable MtsZeroOutSigCoeffFlag from Table 13. The color index cIdx of 0 means a luma component.

[0340] According to an example, because MTS can be applied only to the luma component of the current block, the decoding apparatus can determine whether the color index is luma by deriving the variable MtsZeroOutSigCoeffFlag to determine whether the MTS index should be analyzed.

[0341] The variable MtsZeroOutSigCoeffFlag is a variable that indicates whether zero-out is performed when MTS is applied. Indicates whether the transform coefficient exists in the upper left region where the last significant coefficient may exist due to zero-out after performing the MTS, that is, in the region other than the upper left 16 X 16 region. The variable MtsZeroOutSigCoeffFlag is initially defined as 1 at the coding unit level as shown in Table 11 (MtsZeroOutSigCoeffFlag = 1), and when the transform coefficient exists in the region other than the 16 X 16 region, its value can be changed from 1 to 0 in the residual coding level region as per shown in Table 13 (MtsZeroOutSigCoeffFlag = 0). When the value of the MtsZeroOutSigCoeffFlag variable is 0, the MTS index is not flagged.

[0342] As shown in Table 13, at the residual coding level, a non-zero region in which a non-zero transform coefficient may exist can be defined depending on whether the zero-out accompanying the MTS is realized or not, and Even in this case, the color index (cIdx) is 0, the non-zero region can be set to the top left 16 X 16 region of the current block.

[0343] Thus, when deriving the variable that determines whether the MTS index is analyzed, it is determined whether the color component is luma or chroma. However, because LFNST can be applied to both the luma component and the chroma component of the current block, the color component is not determined when deriving a variable to determine whether the LFNST index should be analyzed.

[0344] For example, Table 11 shows a variable LfnstZeroOutSigCoeffFlag that can indicate that zero-out is performed when LFNST is applied. The LfnstZeroOutSigCoeffFlag variable indicates whether there is a significant coefficient in the second region except the first region in the upper left corner of the current block. This value is initially set to 1, and when the significant coefficient is present in the second region, the value can be changed to 0. The LFNST index can be analyzed only when the value of the initially set variable LfnstZeroOutSigCoeffFlag is kept at 1. When determining and derive whether the value of the LfnstZeroOutSigCoeffFlag variable is 1. Since LFNST can be applied to both the luma component and the chroma component of the current block, the color index of the current block is not determined.

[0345] Figure 12 illustrates a CCLM applicable when deriving an intraprediction mode for a chroma block according to an embodiment.

[0346] In this specification, a “reference sample model” may refer to a set of neighboring reference samples of a current chroma block to predict the current chroma block. A reference sample template may be predefined and information about the reference sample template may be signaled from the encoding apparatus 100 to the decoding apparatus 200.

[0347] Referring to Figure 12, a set of shaded samples in a single row adjacent to a 4 x 4 block, which is a current chroma block, refers to a reference sample template. The reference sample template is configured in a single row of reference samples, while a reference sample region in a luma region corresponding to the reference sample template is configured in two rows, as shown in Figure 12.

[0348] In one embodiment, when intracoding a chroma image is performed in a Joint Exploration Test Model (JEM) used in the Joint Video Exploration Team (JVET), a cross-component linear model (CCLM) can to be used. CCLM is a method for predicting a pixel value of a chroma image from a pixel value of a reconstructed luma image and is based on a high correlation between a luma image and a chroma image.

[0349] CCLM prediction of Cb and Cr chroma images can be performed based on the following equation. Equation 11

[0350] Here, Predc (i, j) denotes a Cb or Cr chroma image to be predicted, RecL'(i, j) denotes a reconstructed luminance image adjusted to a chroma block size, and (i, j) denotes the coordinates of a pixel. In a 4:2:0 color format, because the size of a luminance image is twice that of a chroma image, RecL' with the size of the chroma block needs to be generated by downsampling, and therefore the chroma image Predc (i,j) , pixels of the luminance image to be used for the chroma image predc (i, j) can be employed considering both RecL(2i, 2j) and neighboring pixels. RecL'(i,j) can be termed as a downsampled luma sample.

[0351] For example, RecL'(i,j) can be derived using six neighboring pixels as shown in the following equation. Equation 12

[0352] α and β denote a cross-correlation and mean difference between a neighboring model of the Cb or Cr chroma block and a neighboring model of the luminance block in the shaded region in Figure 12. For example, α and β are represented by Equation 13 Equation 13

[0353] L(n) denotes neighboring reference samples and/or left neighboring samples of a chroma block corresponding to a current chroma image, C(n) denotes neighboring reference samples and/or left neighboring samples of the current chroma block for the which encoding is currently applied and (i, j) denotes a pixel position. Additionally, L(n) can denote downsampled upper neighbor samples and/or left neighbor samples of the current luma block. N can denote the total number of pixel pair values (luminance and chroma) used for calculating a CCLM parameter and can indicate a value that is twice the value of the width and height of the current chroma block.

[0354] Engravings can be divided into a sequence of coding tree units (CTUs). A CTU can correspond to a coding tree block (CTB). Alternatively, the CTU may include a coding tree block of luma samples and a coding tree block of corresponding chroma samples. A tree type can be classified as a single tree (SINGLE_TREE) or a double tree (DUAL_TREE) according to whether a luma block and a corresponding chroma block have an individual partition structure. A single tree may indicate that the chroma block has the same partition structure as the luma block, and a double tree may indicate that the chroma component block has a different partition structure than the luma block.

[0355] When an LFNST is applied to a chroma transform block according to an example, it is necessary to query the information in a placed luma transform block.

[0356] Existing descriptive report text about a relevant part is shown in the following table. Table 14 Inputs to this process are: - a luma location ( xTbY, yTbY ) specifying the top-left sample of the current luma transform block relative to the top-left luma sample of the current engraving, - a variable nTbW specifying the width of the block of current transform, - an nTbH variable specifying the height of the current transform block, - a cIdx variable specifying the color component of the current block, - an (nTbW)x(nTbH) d[ x ][ y ] array of transform coefficients scaled with x = 0..nTbW - 1, y = 0..nTbH - 1 When lfnst_idx is not equal to 0 and both nTbW and nTbH are greater than or equal to 4, the following applies: - When predModeIntra is equal to INTRA_LT_CCLM, INTRA_L_CCLM, or INTRA_T_CCLM, predModeIntra is derived as follows: - If intra_mip_flag[xTbY + nTbW/2][yTbY + nTbH/2] is equal to 1, predModeIntra is set equal to INTRA_PLANAR. - Otherwise, if CuPredMode [0][xTbY + nTbW/2][yTbY + nTbH/2] is equal to MODE_IBC or MODE_PLT, predModeIntra is set equal to INTRA_DC. - Otherwise, predModeIntra is set equal to IntraPredModeY[xTbY + nTbW/2][yTbY + nTbH/2].

[0357] As shown in Table 14, when a current intraprediction mode is a CCLM mode, the value of a predModeIntra variable for the chroma transform block is determined by taking an intraprediction mode value for the luma transform block co-located (part indicated in italics). The intraprediction mode value (predModeIntra value) from the luma transform block can be used later to determine a set of LFNST.

[0358] However, the input variables nTbW and nTbH as input values of this transform process denote the width and height of the current transform block. Thus, when the current block is a luma transform block, the variables nTbW and nTbH can denote the width and height of the luma transform block, and when the current block is a chroma transform block, the variables nTbW and nTbH denote the width and height of the chroma transform block.

[0359] Here, the variables nTbW and nTbH in the italicized part of Table 14 denote the width and height of the chroma transform block which do not reflect the color format and therefore do not accurately indicate a reference position of the chroma block. luma transform corresponding to the chroma transform block. Therefore, the italicized part of Table 14 can be modified as shown in the following table. Table 15

[0360] As shown in Table 15, nTbW and nTbH are changed to (nTbW * SubWidthC ) / 2 and ( nTbH * SubHeightC ) / 2, respectively. xTbY and yTbY can denote a luma position in the current picture (the top left sample of the current luma transform block relative to the top left luma sample of the current picture), and nTbW and nTbH can denote the width and height of the transform currently encoded block (a variable nTbW specifying the width of the current transform block and a variable nTbH specifying the height of the current transform block).

[0361] When the currently encoded transform block is a chroma transform block (Cb or Cr), nTbW and nTbH are the width and height of the chroma transform block, respectively. Consequently, when the currently encoded transform block is a chroma transform block (cIdx > 0), a reference position for a placed luma transform block needs to be obtained using the width and height of the luma transform block at the same time. obtain the reference position. In Table 15, SubWidthC and SubHeightC are values defined according to the color format (for example, 4:2:0, 4:2:2, or 4:4:4) and specifically a width ratio and a ratio in height between a luma component and a chroma component, respectively (see Table 16 below). So, in a case of the chroma transform block, (nTbW * SubWidthC) and (nTbH * SubHeightC) can be the width and height relative to the placed luma transform block, respectively.

[0362] Consequently, xTbY + ( nTbW * SubWidthC ) / 2 and yTbY + ( nTbH * SubHeightC ) / 2 denote the value of a central position in the luma transform block placed based on a top left position of the current engraving and can thus, precisely indicate the placed luma transform block. Table 16

[0363] In Table 15, a predModeIntra variable denotes an intraprediction mode value, and the value of the predModeIntra variable equal to INTRA_LT_CCLM, INTRA_L_CCLM, or INTRA_T_CCLM indicates that the current transform block is a chroma transform block. According to an example, in the current VVC standard, INTRA_LT_CCLM, INTRA_L_CCLM and INTRA_T_CCLM respectively correspond to mode values of 81, 82 and 83 among the intraprediction mode values. Therefore, as shown in Table 15, the reference position for the placed luma transform block needs to be obtained using the value of xTbY + (nTbW * SubWidthC) / 2 and the value of yTbY + (nTbH * SubHeightC) / 2.

[0364] As shown in Table 15, the value of the predModeIntra variable is updated in view of a variable intra_mip_flag[ xTbY + (nTbW * SubWidthC) / 2 ][ yTbY + (nTbH * SubHeightC)/2] and a variable CuPredMode [0 ][xTbY+(nTbW*SubWidthC)/2][yTbY+(nTbH*SubHeightC)/2].

[0365] intra_mip_flag is a variable that indicates whether the current transform block (or coding unit) is encoded by a matrix-based intraprediction (MIP) method, and intra_mip_flag[x][y] is a flag value that indicates Whether MIP is applied to a position corresponding to the (x, y) coordinates based on the luma component when a top left position in the current engraving is set to (0, 0). The x and y coordinates increase from left to right and top to bottom, respectively, and when the flag indicating whether MIP is applied is 1, the flag indicates that MIP is applied. When the flag indicating whether MIP is applied is 0, the flag indicates that MIP is not applied. MIP can only be applied to the luma block.

[0366] According to the modified part of Table 15, when the value of intra_mip_flag[ xTbY + ( nTbW * SubWidthC ) / 2 ][ yTbY + ( nTbH * SubHeightC ) / 2 ] in the placed luma transform block is 1, the predModeIntra value is set to a planar mode (INTRA_PLANAR).

[0367] The value of the variable CuPredMode[ 0 ][ xTbY + ( nTbW * SubWidthC ) / 2 ][ yTbY + ( nTbH * SubHeightC ) / 2 ] denotes a prediction mode value corresponding to coordinates ( xTbY + ( nTbW * SubWidthC ) / 2, yTbY + ( nTbH * SubHeightC ) / 2 ) when the top left position of the current engraving for the luma component is set (0, 0). The prediction mode value can have values MODE_INTRA, MODE_IBC, MODE_PLT, and MODE_INTER, which denote an intraprediction mode, an intrablock copy (IBC) prediction mode, a palette coding mode (PLT), and a interprediction, respectively. According to Table 15, when the value of variable CuPredMode[ 0 ][ xTbY + ( nTbW * SubWidthC ) / 2 ][ yTbY + ( nTbH * SubHeightC ) / 2 ] is MODE_IBC or MODE_PLT, the value of variable predModeIntra is set to a DC mode. In a case other than the two cases, the value of the variable predModeIntra is defined as IntraPredModeY[ xTbY + ( nTbW * SubWidthC ) / 2 ][ yTbY + ( nTbH * SubHeightC ) / 2 ] (value of the intraprediction mode corresponding to a center position in the placed luma transform block).

[0368] According to an example, the value of the predModeIntra variable can be updated once again based on the predModeIntra value updated in Table 15, considering whether wide-angle intraprediction is performed as shown in the following table. Table 17 8.4.5.2.6 Wide Angle Intraprediction Mode Mapping Process Inputs to this process are: 1. a predModeIntra variable specifying the intraprediction mode, 2. an nTbW variable specifying the width of the transform block, 3. an nTbH variable specifying the height of the transform block, - a cIdx variable specifying the color component of the current block. Output of this process is the modified intraprediction mode predModeIntra. The variables nW and nH are derived as follows: 4. If IntraSubPartitionsSplitType equals ISP_NO_SPLIT or cIdx does not equal 0, the following applies: nW = nTbW(8-111) nH = nTbH(8-112) 5 . Otherwise (IntraSubPartitionsSplitType is not equal to ISP_NO_SPLIT and cIdx is equal to 0), the following applies: nW = nCbW(8-113) nH = nCbH(8-114) The variable whRatio is set equal to Abs(Log2 (nW/nH)). For non-square blocks (nW is not equal to nH), the predModeIntra intraprediction mode is modified as follows: 6. If all of the following conditions are true, predModeIntra is set equal to (predModeIntra + 65). 7. nW is greater than nH 8. predModeIntra is greater than or equal to 2 9. predModeIntra is less than (whRatio > 1) ? (8 + 2 * whRatio) : 8 10. Otherwise, if all of the following conditions are true, predModeIntra will be set to (predModeIntra - 67). 11. nH is greater than nW 12. predModeIntra is less than or equal to 66 13. predModeIntra is greater than (whRatio > 1) ? (60 - 2 * whRatio) : 60

[0369] The input values of predModeIntra, nTbW, and nTbH in a mapping process shown in Table 17 are the same as the predModeIntra variable value updated in Table 15 and nTbW and nTbH referenced in Table 15, respectively.

[0370] In Table 17, nCbW and nCbH denote the width and height of a coding block corresponding to the transform block, respectively, and a variable IntraSubPartitionsSplitType indicates whether ISP mode is applied, where IntraSubPartitionsSplitType equal to ISP_NO_SPLIT indicates that the encoding unit is not ISP-partitioned (that is, ISP mode is not applied). The IntraSubPartitionsSplitType variable other than ISP_NO_SPLIT indicates that ISP mode is applied and therefore the encoding unit is partitioned into two or four partition blocks. In Table 17, cIdx is an index that indicates a color component. A cIdx value of 0 indicates a luma block and a cIdx value other than 0 indicates a chroma block. The predModeIntra value output through the mapping process in Table 17 is an updated value considering whether a wide-angle intraprediction (WAIP) mode is applied.

[0371] With respect to the predModeIntra value updated by Table 17, a set of LFNST can be determined through a mapping relationship shown in the following table. Table 18

[0372] In the above table, lfnstTrSetIdx denotes an index that indicates an LFNST set and has a value from 0 to 3, which indicates that a total of four LFNST sets are configured. Each LFNST set can include two transform kernels, i.e. LFNST kernels (transform kernels can be 16 x 16 matrices or 16 x 48 matrices based on a forward direction depending on a region to which an LFNST is applied) and a transform kernel to be applied between the two transform kernels can be specified by signaling the LFNST index. Additionally, you can specify whether LFNST should be applied via the LFNST index. In the current VVC standard, the LFNST index can have values of 0, 1 and 2, 0 indicates that no LFNST is applied and 1 and 2 indicate the two transform kernels, respectively.

[0373] The following drawings are provided to describe specific examples of the present disclosure. Since specific device designations or specific signal/message/field designations illustrated in the drawings are provided for illustration, the technical features of the present disclosure are not limited to specific designations used in the following drawings.

[0374] Figure 13 is a flow chart illustrating an operation of a video decoding apparatus in accordance with an embodiment of the present disclosure.

[0375] Each process disclosed in Figure 13 is based on some of the details described with reference to Figure 3 to Figure 12. Therefore, a description of specific details superimposed on the details described with reference to Figure 2 to Figure 12 will be omitted or will be done schematically .

[0376] The decoding apparatus 200 in accordance with one embodiment may obtain intraprediction mode information and an LFNST index from a bit stream (S1310).

[0377] Intraprediction mode information may include an intraprediction mode for a neighboring block (e.g., a left and/or top neighboring block) of a current block and a most likely mode index (MPM) indicating one of the candidates of MPM in an MPM List derived based on additional candidate modes or remaining intraprediction mode information indicating one of the remaining intraprediction modes not included in the MPM candidates.

[0378] Additionally, the intra mode information may include sps_cclm_enabled_flag flag information indicating whether a CCLM is applied to the current block and intra_chroma_pred_mode information about an intraprediction mode for a chroma component.

[0379] LFNST index information is received as syntax information, and syntax information is received as a binarized bin sequence including 0 and 1.

[0380] An LFNST index syntax element in accordance with the present embodiment may indicate whether an inverse LFNST or an inverse inseparable transform is applied and any of the transform kernel matrices included in a transform set and when the set of transform includes two transform kernel matrices, the transform index syntax element can have three values.

[0381] That is, according to one embodiment, the LFNST index syntax element values may include 0 indicating that no inverse LFNST is applied to a target block, 1 indicating a first transform kernel matrix between the LFNST matrices. transform kernel and 2 indicating a second transform kernel matrix between the transform kernel matrices.

[0382] The decoding apparatus 200 may decode information about quantized transform coefficients for the current block of the bit stream and may derive quantized transform coefficients for the target block based on information about the quantized transform coefficients for the current block. Information about the quantized transform coefficients for the target block may be included in a sequence parameter set (SPS) or a slice header and may include at least one of information about whether an RST is applied, information about a reduced factor , information about a minimum transform size to apply an RST, information about a maximum transform size to apply an RST, an inverse RST size, and information about a transform index indicating any of the transform kernel matrices included in a set of transformation.

[0383] The decoding apparatus 200 may derive transform coefficients by dequantizing residual information in the current block, i.e., the quantized transform coefficients, and may arrange the derived transform coefficients in a predetermined scan order.

[0384] Specifically, the derived transform coefficients can be arranged in 4 x 4 block units according to a reverse diagonal scan order, and the transform coefficients in a 4 x 4 block can also be arranged according to the order reverse diagonal scan. That is, the dequantized transform coefficients can be arranged according to a reverse scan order applied in a video codec, such as in VVC or HEVC.

[0385] The transform coefficient derived based on the residual information may be the dequantized transform coefficient as described above, or it may be quantized transform coefficients. That is, the transform coefficients can be any data to check for non-zero data in the current block regardless of quantization.

[0386] The decoding apparatus may derive an intraprediction mode for a chroma block as a CCLM mode based on the intraprediction mode information (S1320).

[0387] For example, the decoding apparatus may receive information about the intraprediction mode for the current chroma block through the bit stream, and may derive the CCLM mode as the intraprediction mode for the current chroma block based on the information in intraprediction mode.

[0388] The CCLM mode may include a top left CCLM mode, a top CCLM mode, or a left CCLM mode.

[0389] As described above, the decoding apparatus can derive a residual sample by applying an LFNST, which is a non-separable transform, or an MTS, which is a separable transform, and these transforms can be performed respectively based on the LFNST index indicating an LFNST kernel, that is, an LFNST matrix, and an MTS index indicating an MTS kernel.

[0390] For LFNST, a set of LFNST needs to be determined, and the set of LFNST has a mapping relationship with an intraprediction mode for the current block.

[0391] The decoding apparatus may update the intraprediction mode for the chroma block based on an intraprediction mode for a luma block corresponding to the chroma block for LFNST inverse of the chroma block (S1330).

[0392] According to an example, the updated intraprediction mode can be derived as an intraprediction mode corresponding to a specific position in the luma block, and the specific position can be defined based on a color format of the chroma block.

[0393] The specific position may be a central position of the luma block and may be represented by ((xTbY + (nTbW * SubWidthC)/2), (yTbY + (nTbH * SubHeightC)/2)).

[0394] In the center position, xTbY and yTbY denote top-left coordinates of the luma block, that is, a top-left position in a luma sample reference for a current transform block, nTbW and nTbH denote the width and height of the chroma block, and SubWidthC and SubHeightC correspond to variables corresponding to the color format. ((xTbY + (nTbW * SubWidthC)/2), (yTbY + (nTbH * SubHeightC)/2)) denotes a central position of a luma transform block, and IntraPredModeY[xTbY + (nTbW * SubWidthC)/2][ yTbY + (nTbH * SubHeightC)/2] denotes an intraprediction mode for the luma block for position.

[0395] SubWidthC and SubHeightC can be derived as shown in Table 16. That is, when the color format is 4:2:0, SubWidthC and SubHeighC are 2, and when the color format is 4:2:2, SubWidthC is 2 and SubHeightC is 1.

[0396] As shown in Table 15, to designate the specific position of the luma block corresponding to the chroma block, regardless of the color format, the color format is reflected in a variable that indicates the specific position.

[0397] According to an example, when the intraprediction mode for the luma block corresponding to the specific position is a matrix-based intraprediction mode (hereinafter, “MIP”), the decoding apparatus may set the updated intraprediction mode to an intraplanar mode.

[0398] The MIP mode can be referred to as affine linear weighted intraprediction (ALWIP) or matrix weighted intraprediction (MWIP). When MIP is applied to the current block, prediction samples for the current block can be derived by ii) performing a matrix-vector multiplication procedure i) using neighboring reference samples that have been subjected to an averaging procedure and iii ) also performing a horizontal procedure/vertical interpolation procedure.

[0399] Alternatively, according to an example, when the intraprediction mode corresponding to the specific position is an intrablock copy (IBC) mode or a palette mode, the decoding apparatus may set the updated intraprediction mode to an intraprediction mode. A.D.

[0400] IBC prediction mode or palette mode can be used to encode an image/video content including a game, e.g., screen content coding (SCC). IBC basically performs prediction within a current picture, but it can be performed in a similar way to interprediction in that a reference block is derived from the current picture. That is, the IBC may use at least one of the interprediction techniques described in the present disclosure. Palette mode can be considered as an example of intracoding or intraprediction. When palette mode is applied, a value of a sample in the engraving can be flagged based on information from a palette table and a palette index.

[0401] In summary, when the intraprediction mode for the center position is MIP mode, IBC mode, and palette mode, the intraprediction mode for the chroma block can be updated to a specific mode, such as intraplanar mode or intra DC mode.

[0402] When the intraprediction mode for the center position is not the MIP mode, the IBC mode, and the palette mode, the intraprediction mode for the chroma block can be updated to the intraprediction mode of the luma block in relative to the central position to reflect an association between the chroma block and the luma block.

[0403] The decoding apparatus may determine an LFNST set including LFNST matrices based on the updated intraprediction mode (S1340), and may derive transform coefficients for the chroma block based on an LFNST matrix derived from the LFNST set (S1350).

[0404] Any of the plurality of LFNST arrays may be selected based on the LFNST set and the LFNST index.

[0405] As shown in Table 18, the LFNST transform set is derived according to the intraprediction mode, and 81 to 83 indicating the CCLM mode in the intraprediction mode are omitted, because the LFNST transform set is derived using an intra mode value for a corresponding luma block in CCLM mode.

[0406] According to an example, as shown in Table 18, any of the four sets of LFNST can be determined according to the intraprediction mode for the current block, and a set of LFNST to be applied to the current chroma block as well can be determined.

[0407] The decoding apparatus may perform an inverse RST, e.g., an inverse LFNST, by applying the LFNST matrix to the dequantized transform coefficients, thereby deriving modified transform coefficients for the current chroma block.

[0408] The decoding apparatus may derive residual samples of the transform coefficients via a primary inverse transform (S1360). An MTS can be used in the primary inverse transform.

[0409] Furthermore, the decoding apparatus may generate reconstructed samples based on residual samples for the current block and prediction samples for the current block. The current block can be the current luma block or the current chroma block.

[0410] The following drawings are provided to describe specific examples of the present disclosure. Since specific device designations or specific signal/message/field designations illustrated in the drawings are provided for illustration, the technical features of the present disclosure are not limited to specific designations used in the following drawings.

[0411] Figure 14 is a flowchart illustrating an operation of a video coding apparatus in accordance with an embodiment of the present disclosure.

[0412] Each process disclosed in Figure 14 is based on some of the details described with reference to Figure 3 to Figure 12. Therefore, a description of specific details superimposed on the details described with reference to Figure 1 and Figure 3 to Figure 12 will be omitted or will be done schematically.

[0413] The coding apparatus 100 in accordance with one embodiment may derive an intraprediction mode for a chroma block as a CCLM mode (S1410).

[0414] For example, the encoding apparatus may determine the intraprediction mode for the current chroma block based on a rate distortion (RD) cost (or RDO). Here, the RD cost can be derived based on the sum of absolute differences (SAD). The coding apparatus may determine the CCLM mode as the intraprediction mode for the current chroma block based on the RD cost.

[0415] The CCLM mode may include a top left CCLM mode, a top CCLM mode, or a left CCLM mode.

[0416] The encoding apparatus may encode intraprediction mode information for the current chroma block, and the intraprediction mode information may be signaled via a bit stream. Prediction-related information about the current chroma block may include information about the intraprediction mode.

[0417] The encoding apparatus may derive prediction samples for the chroma block based on the CCLM mode (S1420).

[0418] According to one embodiment, the coding apparatus may derive residual samples for the chroma block based on the prediction samples (S1430).

[0419] According to one embodiment, the coding apparatus may derive transform coefficients for the chroma block based on a primary transform in the residual samples.

[0420] The primary transform can be performed through a plurality of transform kernels, in which case a transform kernel can be selected based on the intraprediction mode.

[0421] The encoding apparatus may update the intraprediction mode for the chroma block based on an intraprediction mode for a luma block corresponding to the chroma block for LFNST of the chroma block (S1440).

[0422] As shown in Table 15, the encoding apparatus may update the CCLM mode for the chroma block based on the intraprediction mode for the luma block corresponding to the chroma block (-When predModeIntra equals INTRA_LT_CCLM, INTRA_L_CCLM, or INTRA_T_CCLM , predModeIntra is derived as follows :).

[0423] According to an example, the updated intraprediction mode can be derived as an intraprediction mode corresponding to a specific position in the luma block, and the specific position can be defined based on a color format of the chroma block.

[0424] The specific position may be a central position of the luma block and may be represented by ((xTbY + (nTbW * SubWidthC)/2), (yTbY + (nTbH * SubHeightC)/2)).

[0425] In the center position, xTbY and yTbY denote top-left coordinates of the luma block, that is, a top-left position in a luma sample reference for a current transform block, nTbW and nTbH denote the width and height of the chroma block, and SubWidthC and SubHeightC correspond to variables corresponding to the color format. ((xTbY + (nTbW * SubWidthC)/2), (yTbY + (nTbH * SubHeightC)/2)) denotes a central position of a luma transform block, and IntraPredModeY[xTbY + (nTbW * SubWidthC)/2][ yTbY + (nTbH * SubHeightC)/2] denotes an intraprediction mode for the luma block for position.

[0426] SubWidthC and SubHeightC can be derived as shown in Table 16. That is, when the color format is 4:2:0, SubWidthC and SubHeightC are 2, and when the color format is 4:2:2, SubWidthC is 2 and SubHeightC is 1.

[0427] As shown in Table 15, to designate the specific position of the luma block corresponding to the chroma block, regardless of the color format, the color format is reflected in a variable indicating the specific position.

[0428] According to an example, when the intraprediction mode for the luma block corresponding to the specific position is a matrix-based intraprediction mode (hereinafter, “MIP”), the coding apparatus may set the updated intraprediction mode to an intraplanar mode.

[0429] The MIP mode can be referred to as affine linear weighted intraprediction (ALWIP) or matrix weighted intraprediction (MWIP). When MIP is applied to the current block, prediction samples for the current block can be derived by ii) performing a matrix-vector multiplication procedure i) using neighboring reference samples that have been subjected to an averaging procedure and iii ) also performing a horizontal procedure/vertical interpolation procedure.

[0430] Alternatively, according to an example, when the intraprediction mode corresponding to the specific position is an intrablock copy (IBC) mode or a palette mode, the coding apparatus may set the updated intraprediction mode to an intraprediction mode. A.D.

[0431] IBC prediction mode or palette mode can be used to encode an image/video content including a game, e.g., screen content coding (SCC). IBC basically performs prediction within a current picture, but it can be performed in a similar way to interprediction in that a reference block is derived from the current picture. That is, the IBC may use at least one of the interprediction techniques described in the present disclosure. Palette mode can be considered as an example of intracoding or intraprediction. When palette mode is applied, a value of a sample in the engraving can be flagged based on information from a palette table and a palette index.

[0432] In summary, when the intraprediction mode for the center position is MIP mode, IBC mode, and palette mode, the intraprediction mode for the chroma block can be updated to a specific mode, such as the intraplanar or intra DC mode.

[0433] When the intraprediction mode for the center position is not the MIP mode, the IBC mode and the palette mode, the intraprediction mode for the chroma block can be updated to the intraprediction mode of the luma block relative to to the central position to reflect an association between the chroma block and the luma block.

[0434] The coding apparatus may determine a set of LFNST including LFNST matrices based on the updated intraprediction mode (S1450), and may derive modified transform coefficients for the chroma block based on the residual samples and an LFNST matrix ( S1460).

[0435] The coding apparatus may determine the transform set based on a mapping relation in accordance with the intraprediction mode applied to the current block and may perform an LFNST, i.e., a non-separable transform, based on any of the two LFNST matrices included in the transform set.

[0436] As described above, a plurality of transform sets can be determined according to an intraprediction mode for a transform block to be transformed. A matrix applied to LFNST is the transpose of a matrix used in an inverse LFNST.

[0437] In an example, the LFNST matrix may be a non-square matrix in which the number of rows is less than the number of columns.

[0438] The encoding apparatus may derive quantized transform coefficients by performing quantitation based on the modified transform coefficients for the current chroma block and may encode and output image information including information about the quantized transform coefficients, information about the mode of intraprediction and an LFNST index that indicates the LFNST matrix (S1470).

[0439] Specifically, the coding apparatus 100 can generate information about the quantized transform coefficients and can encode the information generated about the quantized transform coefficients.

[0440] In an example, information about quantized transform coefficients may include at least one of information about whether LFNST is applied, information about a reduced factor, information about a minimum transform size to apply LFNST, and information about a maximum transform size to apply LFNST.

[0441] The encoding apparatus may encode, such as intra mode information, signal information indicating whether CCLM is applied to the current block, which is sps_cclm_enabled_flag, and information about an intraprediction mode for a chroma component, which is intra_chroma_pred_mode .

[0442] Information about the CCLM mode, which is intra_chroma_pred_mode, may indicate the top left CCLM mode, the top CCLM mode, or the left CCLM mode.

[0443] In the present disclosure, at least one of quantization/dequantization and/or transform/inverse transform may be omitted. When quantization/dequantization is omitted, a quantified transform coefficient can be termed as a transform coefficient. When the transform/inverse transform is omitted, the transform coefficient may be termed as a coefficient or a residual coefficient, or may be termed as a transform coefficient for consistency of expression.

[0444] Furthermore, in the present disclosure, a quantified transform coefficient and a transform coefficient may be referred to as a transform coefficient and a scaled transform coefficient, respectively. In this case, the residual information may include information about a transform coefficient(s) and the information about the transform coefficient(s) may be signaled using a residual encoding syntax. Transform coefficients can be derived based on residual information (or information about the transform coefficient(s), and scaled transform coefficients can be derived by inversely transforming (scaling) the transform coefficients. Residual samples can be derived based on the inverse transform (transform) of the scaled transform coefficients. These details may also be applied/expressed elsewhere in this disclosure.

[0445] In the embodiments described above, the methods are explained based on flowcharts through a series of steps or blocks, but the present disclosure is not limited to the order of the steps and a given step may be performed in a different order or step than described above, or concomitantly with another step. Furthermore, it may be understood by a person skilled in the art that the steps shown in a flowchart are not exclusive and that another step may be incorporated or one or more steps of the flowchart may be removed without affecting the scope of the present disclosure.

[0446] The methods described above in accordance with the present disclosure may be implemented as a form of software and an encoding apparatus and/or decoding apparatus in accordance with the disclosure may be included in a device for image processing, such as a TV, a computer, a smartphone, a set-top box, a display device or the like.

[0447] When the embodiments in the present disclosure are embodied by software, the methods described above may be incorporated as modules (processes, functions, or the like) to perform the functions described above. Modules can be stored in memory and can be executed by a processor. Memory can be inside or outside the processor and can be connected to the processor in several well-known ways. The processor may include an application-specific integrated circuit (ASIC), another chipset, logic circuit, and/or a data processing device. The memory may include a read-only memory (ROM), a random access memory (RAM), a flash memory, a memory card, a storage medium and/or other storage device. That is, the embodiments described in the present disclosure can be incorporated and executed on a processor, a microprocessor, a controller, or a chip. For example, the function units shown in each drawing may be incorporated and executed in a computer, a processor, a microprocessor, a controller, or a chip.

[0448] Furthermore, the decoding apparatus and the encoding apparatus to which the present disclosure is applied may be included in a multimedia transmission transceiver, a mobile communications terminal, a home theater video device, a digital cinema video camera, a surveillance camera, a video chat device, a real-time communication device such as video communication, a mobile transmission device, a storage medium, a video camera, a video on demand (VoD) service delivery device, an over the top (OTT) video device, an Internet device streaming service delivery device, a three-dimensional (3D) video device, a telephony and a medical video device, and can be used to process a video signal or a data signal. For example, the over the top (OTT) video device may include a game console, a Blu-ray player, an Internet-enabled TV, a home theater system, a smartphone, a Tablet PC, a video recorder digital (DVR) and similar.

[0449] Furthermore, the processing method to which the present disclosure is applied may be produced in the form of a program executed by a computer and stored on a computer-readable recording medium. Multimedia data with a data structure in accordance with the present disclosure may also be stored on a computer-readable recording medium. Computer-readable recording medium includes all types of storage devices and distributed storage devices on which computer-readable data is stored. Computer-readable recording media may include, for example, a Blu-ray Disc (BD), a universal serial bus (USB), a ROM, a PROM, an EPROM, an EEPROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk and an optical data storage device. Furthermore, the computer-readable recording medium includes embedded media in the form of a carrier wave (e.g., Internet transmission). Furthermore, a bit stream generated by the encoding method may be stored on a computer-readable recording medium or transmitted over a wired or wireless communication network. Additionally, embodiments of the present disclosure may be embodied as a computer program product by program codes and the program codes may be executed on a computer by embodiments of the present disclosure. Program codes can be stored on a computer-readable medium.

[0450] Figure 15 schematically illustrates an example of a video/image coding system to which the present disclosure is applicable.

[0451] Referring to Figure 15, the video/image coding system may include a first device (source device) and a second device (receiving device). The source device may provide information or encoded video/image data in the form of a file or stream to the receiving device via a digital storage medium or network.

[0452] The source device may include a video source, an encoding apparatus, and a transmitter. The receiving device may include a receiver, a decoding apparatus, and a renderer. The encoding apparatus may be called a video/image encoding apparatus, and the decoding apparatus may be called a video/image decoding apparatus. The transmitter may be included in the coding apparatus. The receiver may be included in the decoding apparatus. The renderer can include a screen, and the screen can be configured as a separate device or an external component.

[0453] The video source may obtain a video/image through a process of capturing, synthesizing, or generating a video/image. The video source may include a video/image capture device and/or a video/image generating device. The video/image capture device may include, for example, one or more cameras, video/image files including previously captured video/images, or the like. The video/image generating device may include, for example, a computer, a tablet and a smartphone and may (electronically) generate a video/image. For example, a virtual video/image may be generated using a computer or the like. In this case, the video/image capture process can be replaced by a related data generation process.

[0454] The encoding apparatus may encode an input video/image. The coding apparatus can perform a series of procedures such as prediction, transform, and quantization for compression and coding efficiency. The encoded data (encoded video/image information) can be output in the form of a bit stream.

[0455] The transmitter may transmit the encoded video/image information or data output in the form of a bit stream to the receiver of the receiving device via a digital storage medium or a network in the form of a file or stream. Digital storage medium can include various storage media such as USB, SD, CD, DVD, Blu-ray, HDD, SSD and the like. The transmitter may include an element for generating a media file via a predetermined file format and may include an element for transmission over a transmission/communication network. The receiver can receive/extract the bit stream and transmit the received/extracted bit stream to the decoding apparatus.

[0456] The decoding apparatus may decode a video/image by performing a series of procedures such as dequantization, inverse transform, prediction and the like corresponding to the operation of the encoding apparatus.

[0457] The renderer can render the decoded video/image. The rendered video/image can be displayed across the screen.

[0458] Figure 16 illustrates the structure of a content streaming system to which the present disclosure is applied.

[0459] Furthermore, the content streaming system to which the present disclosure is applied may largely include an encoding server, a streaming server, a web server, a media store, a user equipment, and a device multimedia input.

[0460] The encoding server functions to compress the input content of multimedia input devices, such as the smartphone, camera, camcoder and the like, into digital data to generate a bit stream and transmit it to the streaming server. As another example, in a case where the multimedia input device such as smart phone, camera, video camera or the like directly outputs a bit stream, the encoding server may be omitted. The bitstream may be generated by a coding method or a bitstream generation method to which the present disclosure is applied. And the streaming server can store the bitstream temporarily during a process to transmit or receive the bitstream.

[0461] The streaming server transmits multimedia data to the user's equipment based on a user's request through the web server, which functions as an instrument that informs the user which service exists. When the user requests a service that the user wants, the web server transfers the request to the streaming server, and the streaming server transmits multimedia data to the user. In this sense, the content streaming system may include a separate control server, and in this case, the control server functions to control commands/responses between respective equipment in the content streaming system.

[0462] The streaming server may receive content from the media store and/or the encoding server. For example, if the content is received from the encoding server, the content can be received in real time. In this case, the streaming server can store the bitstream for a predetermined period of time to provide the streaming service smoothly.

[0463] For example, user equipment may include a cell phone, a smartphone, a laptop, a digital streaming terminal, a personal digital assistant (PDA), a portable multimedia player (PMP), a navigation, a PC slate , a tablet PC, an ultrabook, a wearable device (e.g., a watch-type terminal (smart watch), a glass-type terminal (smart glass), a head-mounted display (HMD)), a digital TV, a desktop computer , digital signage or similar. Each of the servers of the content streaming system can be operated as a distributed server, and in this case, the data received by each server can be processed in a distributed manner.

[0464] The claims disclosed herein may be combined in various ways. For example, the technical features of the method claims of the present disclosure may be combined to be implemented or performed in an apparatus and the technical features of the apparatus claims may be combined to be implemented or performed in a method. Furthermore, the technical features of the method claims and apparatus claims may be combined to be implemented or carried out in an apparatus, and the technical features of the method claims and apparatus claims may be combined to be implemented or carried out in a method. .

Claims (9)

1. Decoding device for decoding images, CHARACTERIZED by the fact that it comprises: a memory; and at least one processor connected to the memory, the at least one processor configured to: obtain intraprediction mode information and an LFNST index from a bit stream; deriving an intraprediction mode for a chroma block as a cross-component linear model (CCLM) mode based on the intraprediction mode information; derive prediction samples for the chroma block based on the intraprediction mode; updating the intraprediction mode for the chroma block based on an intraprediction mode for a luma block corresponding to the chroma block; determining an LFNST set comprising LFNST matrices based on the updated intraprediction mode; deriving transform coefficients for the chroma block based on an LFNST matrix derived from the LFNST index and the LFNST set; derive residual samples for the chroma block based on the transform coefficients, and generate a reconstructed engraving based on the prediction samples and the residual samples for the chroma block, wherein the updated intraprediction mode is derived as an intraprediction mode corresponding to a specific position in the luma block, where the specific position is defined based on a color format of the chroma block, where the color format is 4:2:2, where the specific position is defined as ((xTbY + nTbW), (yTbY + nTbH/2)), where xTbY and yTbY denote upper left coordinates of the luma block, and where nTbW and nTbH denote a width and a height of the chroma block. 2. Decoding apparatus according to claim 1, CHARACTERIZED by the fact that the specific position is a central position of the luma block. 3. Decoding apparatus according to claim 1, CHARACTERIZED by the fact that the color format is 4:2:0; where the specific position is defined as ((xTbY + nTbW), (yTbY + nTbH)). 4. Decoding apparatus according to claim 1, CHARACTERIZED by the fact that the luma block is not encoded based on a matrix-based intraprediction (MIP) mode, an intrablock copy (IBC) mode or a palette (PLT). 5. Image coding apparatus for coding images, the coding apparatus CHARACTERIZED by the fact that it comprises: a memory; and at least one processor connected to the memory, the at least one processor configured to: derive an intraprediction mode for a chroma block as a cross-component linear model (CCLM) mode; derive prediction samples for the chroma block based on the CCLM mode; derive residual samples for the chroma block based on the prediction samples; updating the intraprediction mode for the chroma block based on an intraprediction mode for a luma block corresponding to the chroma block determining a set of LFNST comprising LFNST matrices based on the updated intraprediction mode; and deriving modified transform coefficients for the chroma block, and generating a bit stream including intraprediction mode information related to the intraprediction mode for the chroma block and residual information related to the modified transform coefficients for the chroma block, wherein the mode updated intraprediction mode is derived as an intraprediction mode corresponding to a specific position in the luma block, wherein the specific position is defined based on a color format of the chroma block, wherein the color format is 4:2:2; where the specific position is defined as ((xTbY + nTbW), (yTbY + nTbH/2)), where xTbY and yTbY denote upper-left coordinates of the luma block, and where nTbW and nTbH denote a width and a height of the chroma block. 6. Encoding apparatus according to claim 5, CHARACTERIZED by the fact that the specific position is a central position of the luma block. 7. Coding apparatus according to claim 5, CHARACTERIZED by the fact that the color format is 4:2:0; where the specific position is defined as ((xTbY + nTbW), (yTbY + nTbH)). 8. Encoding apparatus according to claim 5, CHARACTERIZED by the fact that the luma block is not encoded based on a matrix-based intraprediction (MIP) mode, an intrablock copy (IBC) mode or an palette (PLT). 9. Apparatus for transmitting data for an image information, the apparatus CHARACTERIZED by the fact that it comprises: at least one processor configured to: obtain a bit stream for the image information, which the bit stream is generated based on in deriving an intraprediction mode for a chroma block as a cross-component linear model (CCLM) mode; derive prediction samples for the chroma block based on the CCLM mode; derive residual samples for the chroma block based on the prediction samples; updating the intraprediction mode for the chroma block based on an intraprediction mode for a luma block corresponding to the chroma block; determining an LFNST set comprising LFNST matrices based on the updated intraprediction mode; derive modified transform coefficients for the chroma block; and encoding intraprediction mode information related to the intraprediction mode for the chroma block and residual information related to the modified transform coefficients to generate the bit stream; and a transmitter configured to transmit the data comprising the bit stream, wherein the updated intraprediction mode is derived as an intraprediction mode corresponding to a specific position in the luma block, wherein the specific position is defined based on a format of chroma block color, where the color format is 4:2:2; where the specific position is defined as ((xTbY + nTbW), (yTbY + nTbH/2)), where xTbY and yTbY denote upper-left coordinates of the luma block, and where nTbW and nTbH denote a width and a height of the chroma block.