Image encoding / decoding method and device using maximum transform size restriction of chroma component encoding block, and method for transmitting bitstream

IN605111BActive Publication Date: 2026-10-06LG ELECTRONICS INC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
IN202218062026
Authority / Receiving Office
IN · IN
Patent Type
Patents
Current Assignee / Owner
Priority Date
2019-06-24
Filing Date
2022-10-31
Publication Date
2026-10-06

AI Technical Summary

Technical Problem

The increasing demand for high-resolution and high-quality images, such as HD and UHD, leads to a significant increase in the amount of transmitted and stored data, resulting in higher transmission and storage costs, necessitating the need for efficient image compression technologies.

Method used

An image encoding/decoding method and apparatus that limits the maximum transform size of a chroma component coding block, allowing for improved encoding/decoding efficiency by determining the prediction mode, generating prediction and residual blocks, and reconstructing blocks based on these components.

Benefits of technology

This approach enhances encoding/decoding efficiency, reduces data transmission and storage costs, and effectively handles high-resolution image data by optimizing the bitstream generation and transmission.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

ABSTRACT IMAGE ENCODING / DECODING METHOD, APPARATUS AND METHOD FOR TRANSMITTING BITSTREAM USING MAXIMUM TRANSFORM SIZE LIMITATION OF CHROMA COMPONENT CODING BLOCK An image encoding / decoding method and apparatus are provided. An image decoding method performed by an image decoding apparatus may include determining a prediction mode of a current block, generating a prediction block of the current block based on intra block copy (IBC) prediction mode information, based on the prediction mode of the current block being an IBC prediction mode, determining a size of a transform block of the current block based on a color component of the current block, generating a residual block of the current block based on the size of the transform block; and reconstructing the current block based on the prediction block and the residual block of the current block.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field[1] The present disclosure relates to an image encoding / decoding method and apparatus,and, more particularly, to a method and apparatus for encoding / decoding an image by limitinga maximum transform size of a chroma component coding block, and a method of transmittinga bitstream generated by the image encoding method / apparatus of the present disclosure.Background Art[2] Recently, demand for high-resolution and high-quality images such as high definition(HD) images and ultra high definition (UHD) images is increasing in various fields. Asresolution and quality of image data are improved, the amount of transmitted information orbits relatively increases as compared to existing image data. An increase in the amount oftransmitted information or bits causes an increase in transmission cost and storage cost.[3] Accordingly, there is a need for high-efficient image compression technology foreffectively transmitting, storing and reproducing information on high-resolution and highquality images.DisclosureTechnical Problem[4] An object of the present disclosure is to provide an image encoding / decoding methodand apparatus with improved encoding / decoding efficiency.[5] An object of the present disclosure is to provide an image encoding / decoding methodand apparatus capable of improving encoding / decoding efficiency by limiting a maximumtransform size of a chroma component coding block.[6] Another object of the present disclosure is to provide a method of transmitting abitstream generated by an image encoding method or apparatus according to the presentdisclosure.[7] Another object of the present disclosure is to provide a recording medium storing abitstream generated by an image encoding method or apparatus according to the presentdisclosure.[8] Another object of the present disclosure is to provide a recording medium storing abitstream received, decoded and used to reconstruct an image by an image decodingapparatus according to the present disclosure.[9] The technical problems solved by the present disclosure are not limited to the abovetechnical problems and other technical problems which are not described herein will becomeapparent to those skilled in the art from the following description.Technical Solution

[10] An image decoding method performed by an image decoding apparatus according toan aspect of the present disclosure may include determining a prediction mode of a currentblock, generating a prediction block of the current block based on intra block copy (IBC)prediction mode information, based on the prediction mode of the current block being an IBCprediction mode, determining a size of a transform block of the current block based on a colorcomponent of the current block, generating a residual block of the current block based on thesize of the transform block; and reconstructing the current block based on the prediction blockand the residual block of the current block.

[11] In addition, an image decoding apparatus according to an aspect of the presentdisclosure may include a memory and at least one processor. The at least one processor maydetermine a prediction mode of a current block, generate a prediction block of the current blockbased on intra block copy (IBC) prediction mode information, based on the prediction mode ofthe current block being an IBC prediction mode, determine a size of a transform block of thecurrent block based on a color component of the current block, and generate a residual blockof the current block based on the size of the transform block. In this case, the decodingapparatus may reconstruct the current block based on the prediction block and the residualblock of the current block.

[12] In addition, an image encoding method performed by an image encoding apparatusaccording to an aspect of the present disclosure may include determining a current block bysplitting an image, generating a prediction block of the current block by performing intra blockcopy (IBC) prediction on the current block, generating a residual block of the current blockbased on the prediction block and encoding prediction mode information of the current block.In this case, the residual block may be encoded based on a size of a transform block of thecurrent block, and the size of the transform block may be determined based on a colorcomponent of the current block.

[13] In addition, a transmission method according to another aspect of the presentdisclosure may transmit a bitstream generated by the image encoding apparatus or the imageencoding method of the present disclosure.

[14] In addition, a computer-readable recording medium according to another aspect ofthe present disclosure may store the bitstream generated by the image encoding apparatus orthe image encoding method of the present disclosure.

[15] The features briefly summarized above with respect to the present disclosure aremerely exemplary aspects of the detailed description below of the present disclosure, and donot limit the scope of the present disclosure.Advantageous Effects

[16] According to the present disclosure, it is possible to provide an imageencoding / decoding method and apparatus with improved encoding / decoding efficiency.

[17] According to the present disclosure, it is possible to provide an imageencoding / decoding method and apparatus capable of improving encoding / decoding efficiencyby limiting a maximum transform size of a chroma component coding block.

[18] Also, according to the present disclosure, it is possible to provide a method oftransmitting a bitstream generated by an image encoding method or apparatus according tothe present disclosure.

[19] Also, according to the present disclosure, it is possible to provide a recording mediumstoring a bitstream generated by an image encoding method or apparatus according to thepresent disclosure.

[20] Also, according to the present disclosure, it is possible to provide a recording mediumstoring a bitstream received, decoded and used to reconstruct an image by an image decodingapparatus according to the present disclosure.

[21] It will be appreciated by persons skilled in the art that that the effects that can beachieved through the present disclosure are not limited to what has been particularly describedhereinabove and other advantages of the present disclosure will be more clearly understoodfrom the detailed description.Description of Drawings

[22] FIG. 1 is a view schematically showing a video coding system, to which anembodiment of the present disclosure is applicable.

[23] FIG. 2 is a view schematically showing an image encoding apparatus, to which anembodiment of the present disclosure is applicable.

[24] FIG. 3 is a view schematically showing an image decoding apparatus, to which anembodiment of the present disclosure is applicable.

[25] FIG. 4 is a view showing a partitioning structure of an image according to anembodiment.

[26] FIG. 5 is a view showing an embodiment of a partitioning type of a block accordingto a multi-type tree structure.

[27] FIG. 6 is a view showing a signaling mechanism of block splitting information in aquadtree with nested multi-type tree structure according to the present disclosure.

[28] FIG. 7 is a view showing an embodiment in which a CTU is partitioned into multipleCUs.

[29] FIG. 8 is a view illustrating an embodiment of a redundant splitting pattern.

[30] FIG. 9 is a flowchart illustrating an inter prediction based video / image encodingmethod.

[31] FIG. 10 is a view illustrating the configuration of an inter prediction unit 180 accordingto the present disclosure.

[32] FIG. 11 is a flowchart illustrating an inter prediction based video / image decodingmethod.

[33] FIG. 12 is a view illustrating the configuration of an inter prediction unit 260 accordingto the present disclosure.

[34] FIG. 13 is a view illustrating neighboring blocks available as a spatial merge candidateaccording to an embodiment.

[35] FIG. 14 is a view schematically illustrating a merge candidate list construction methodaccording to an embodiment.

[36] FIG. 15 is a view schematically illustrating a motion vector predictor candidate listconstruction method according to an embodiment.

[37] FIG. 16 is a view illustrating a syntax structure for transmitting MVD from an imageencoding apparatus to an image decoding apparatus according to an embodiment.

[38] FIG. 17 is a flowchart illustrating an IBC based video / image encoding methodaccording to an embodiment.

[39] FIG. 18 is a view illustrating the configuration of a prediction unit for performing anIBC based video / image encoding method according to an embodiment.

[40] FIG. 19 is a flowchart illustrating an IBC based video / image decoding methodaccording to an embodiment.

[41] FIG. 20 is a view illustrating a configuration of a prediction unit for performing anIBC based video / image decoding method according to an embodiment.

[42] FIG. 21 is a view illustrating syntax for chroma format signaling according to anembodiment.

[43] FIG. 22 is a view illustrating a chroma format classification table according to anembodiment.

[44] FIG. 23 is a view illustrating a splitting limitation example of a CU for virtual pipelineprocessing.

[45] FIGS. 24 to 26 are views illustrating a splitting example of a CU and a TU accordingto an embodiment.

[46] FIGS. 27 and 28 are flowcharts illustrating IBC prediction and intra prediction towhich a maximum transform size according to an embodiment applies.

[47] FIG. 29 is a flowchart illustrating a method of encoding an image by an encodingapparatus according to an embodiment.

[48] FIG. 30 is a flowchart illustrating a method of decoding an image by a decodingapparatus according to an embodiment.

[49] FIG. 31 is a view showing a content streaming system, to which an embodiment ofthe present disclosure is applicable.Mode for Invention

[50] Hereinafter, the embodiments of the present disclosure will be described in detail withreference to the accompanying drawings so as to be easily implemented by those skilled in theart. However, the present disclosure may be implemented in various different forms, and is notlimited to the embodiments described herein.

[51] In describing the present disclosure, if it is determined that the detailed description ofa related known function or construction renders the scope of the present disclosureunnecessarily ambiguous, the detailed description thereof will be omitted. In the drawings,parts not related to the description of the present disclosure are omitted, and similar referencenumerals are attached to similar parts.

[52] In the present disclosure, when a component is "connected", "coupled" or "linked" toanother component, it may include not only a direct connection relationship but also an indirectconnection relationship in which an intervening component is present. In addition, when acomponent "includes" or "has" other components, it means that other components may befurther included, rather than excluding other components unless otherwise stated.

[53] In the present disclosure, the terms first, second, etc. may be used only for the purposeof distinguishing one component from other components, and do not limit the order orimportance of the components unless otherwise stated. Accordingly, within the scope of thepresent disclosure, a first component in one embodiment may be referred to as a secondcomponent in another embodiment, and similarly, a second component in one embodimentmay be referred to as a first component in another embodiment.

[54] In the present disclosure, components that are distinguished from each other areintended to clearly describe each feature, and do not mean that the components are necessarilyseparated. That is, a plurality of components may be integrated and implemented in onehardware or software unit, or one component may be distributed and implemented in a pluralityof hardware or software units. Therefore, even if not stated otherwise, such embodiments inwhich the components are integrated or the component is distributed are also included in thescope of the present disclosure.

[55] In the present disclosure, the components described in various embodiments do notnecessarily mean essential components, and some components may be optional components.Accordingly, an embodiment consisting of a subset of components described in an embodimentis also included in the scope of the present disclosure. In addition, embodiments including othercomponents in addition to components described in the various embodiments are included inthe scope of the present disclosure.

[56] The present disclosure relates to encoding and decoding of an image, and terms usedin the present disclosure may have a general meaning commonly used in the technical field, towhich the present disclosure belongs, unless newly defined in the present disclosure.

[57] In the present disclosure, a "picture" generally refers to a unit representing one imagein a specific time period, and a slice / tile is a coding unit constituting a part of a picture, andone picture may be composed of one or more slices / tiles. In addition, a slice / tile may includeone or more coding tree units (CTUs).

[58] In the present disclosure, a "pixel" or a "pel" may mean a smallest unit constitutingone picture (or image). In addition, "sample" may be used as a term corresponding to a pixel.A sample may generally represent a pixel or a value of a pixel, and may represent only apixel / pixel value of a luma component or only a pixel / pixel value of a chroma component.

[59] In the present disclosure, a "unit" may represent a basic unit of image processing. Theunit may include at least one of a specific region of the picture and information related to theregion. The unit may be used interchangeably with terms such as "sample array", "block" or"area" in some cases. In a general case, an MxN block may include samples (or sample arrays)or a set (or array) of transform coefficients of M columns and N rows.

[60] In the present disclosure, "current block" may mean one of "current coding block","current coding unit", "coding target block", "decoding target block" or "processing targetblock". When prediction is performed, "current block" may mean "current prediction block"or "prediction target block". When transform (inverse transform) / quantization (dequantization)is performed, "current block" may mean "current transform block" or "transform target block".When filtering is performed, "current block" may mean "filtering target block".

[61] In addition, in the present disclosure, a "current block" may mean "a luma block of acurrent block" unless explicitly stated as a chroma block. The "chroma block of the currentblock" may be expressed by including an explicit description of a chroma block, such as"chroma block" or "current chroma block".

[62] In the present disclosure, the term " / " and "," should be interpreted to indicate "and / or."For instance, the expression "A / B" and "A, B" may mean "A and / or B." Further, "A / B / C" and"A / B / C" may mean "at least one of A, B, and / or C."

[63] In the present disclosure, the term "or" should be interpreted to indicate "and / or." Forinstance, the expression "A or B" may comprise 1) only "A", 2) only "B", and / or 3) both "Aand B". In other words, in the present disclosure, the term "or" should be interpreted toindicate "additionally or alternatively."

[64] Overview of video coding system

[65] FIG. 1 is a view showing a video coding system according to the present disclosure.

[66] The video coding system according to an embodiment may include a encodingapparatus 10 and a decoding apparatus 20. The encoding apparatus 10 may deliver encodedvideo and / or image information or data to the decoding apparatus 20 in the form of a file orstreaming via a digital storage medium or network.

[67] The encoding apparatus 10 according to an embodiment may include a video sourcegenerator 11, an encoding unit 12 and a transmitter 13. The decoding apparatus 20 accordingto an embodiment may include a receiver 21, a decoding unit 22 and a renderer 23. Theencoding unit 12 may be called a video / image encoding unit, and the decoding unit 22 may becalled a video / image decoding unit. The transmitter 13 may be included in the encoding unit12. The receiver 21 may be included in the decoding unit 22. The renderer 23 may include adisplay and the display may be configured as a separate device or an external component.

[68] The video source generator 11 may acquire a video / image through a process ofcapturing, synthesizing or generating the video / image. The video source generator 11 mayinclude a video / image capture device and / or a video / image generating device. The video / imagecapture device may include, for example, one or more cameras, video / image archives includingpreviously captured video / images, and the like. The video / image generating device mayinclude, for example, computers, tablets and smartphones, and may (electronically) generatevideo / images. For example, a virtual video / image may be generated through a computer or thelike. In this case, the video / image capturing process may be replaced by a process of generatingrelated data.

[69] The encoding unit 12 may encode an input video / image. The encoding unit 12 mayperform a series of procedures such as prediction, transform, and quantization for compressionand coding efficiency. The encoding unit 12 may output encoded data (encoded video / imageinformation) in the form of a bitstream.

[70] The transmitter 13 may transmit the encoded video / image information or data outputin the form of a bitstream to the receiver 21 of the decoding apparatus 20 through a digitalstorage medium or a network in the form of a file or streaming. The digital storage mediummay include various storage mediums such as USB, SD, CD, DVD, Blu-ray, HDD, SSD, andthe like. The transmitter 13 may include an element for generating a media file through apredetermined file format and may include an element for transmission through abroadcast / communication network. The receiver 21 may extract / receive the bitstream from thestorage medium or network and transmit the bitstream to the decoding unit 22.

[71] The decoding unit 22 may decode the video / image by performing a series ofprocedures such as dequantization, inverse transform, and prediction corresponding to theoperation of the encoding unit 12.

[72] The renderer 23 may render the decoded video / image. The rendered video / image maybe displayed through the display.

[73] Overview of image encoding apparatus

[74] FIG. 2 is a view schematically showing an image encoding apparatus, to which anembodiment of the present disclosure is applicable.

[75] As shown in FIG. 2, the image encoding apparatus 100 may include an imagepartitioner 110, a subtractor 115, a transformer 120, a quantizer 130, a dequantizer 140, aninverse transformer 150, an adder 155, a filter 160, a memory 170, an inter prediction unit180, an intra prediction unit 185 and an entropy encoder 190. The inter prediction unit 180and the intra prediction unit 185 may be collectively referred to as a "prediction unit". Thetransformer 120, the quantizer 130, the dequantizer 140 and the inverse transformer 150 maybe included in a residual processor. The residual processor may further include the subtractor115.

[76] All or at least some of the plurality of components configuring the image encodingapparatus 100 may be configured by one hardware component (e.g., an encoder or a processor)in some embodiments. In addition, the memory 170 may include a decoded picture buffer (DPB)and may be configured by a digital storage medium.

[77] The image partitioner 110 may partition an input image (or a picture or a frame) inputto the image encoding apparatus 100 into one or more processing units. For example, theprocessing unit may be called a coding unit (CU). The coding unit may be acquired byrecursively partitioning a coding tree unit (CTU) or a largest coding unit (LCU) according to aquad-tree binary-tree ternary-tree (QT / BT / TT) structure. For example, one coding unit may bepartitioned into a plurality of coding units of a deeper depth based on a quad tree structure, abinary tree structure, and / or a ternary structure. For partitioning of the coding unit, a quad treestructure may be applied first and the binary tree structure and / or ternary structure may beapplied later. The coding procedure according to the present disclosure may be performedbased on the final coding unit that is no longer partitioned. The largest coding unit may be usedas the final coding unit or the coding unit of deeper depth acquired by partitioning the largestcoding unit may be used as the final coding unit. Here, the coding procedure may include aprocedure of prediction, transform, and reconstruction, which will be described later. Asanother example, the processing unit of the coding procedure may be a prediction unit (PU) ora transform unit (TU). The prediction unit and the transform unit may be split or partitionedfrom the final coding unit. The prediction unit may be a unit of sample prediction, and thetransform unit may be a unit for deriving a transform coefficient and / or a unit for deriving aresidual signal from the transform coefficient.

[78] The prediction unit (the inter prediction unit 180 or the intra prediction unit 185) mayperform prediction on a block to be processed (current block) and generate a predicted blockincluding prediction samples for the current block. The prediction unit may determine whetherintra prediction or inter prediction is applied on a current block or CU basis. The predictionunit may generate various information related to prediction of the current block and transmitthe generated information to the entropy encoder 190. The information on the prediction maybe encoded in the entropy encoder 190 and output in the form of a bitstream.

[79] The intra prediction unit 185 may predict the current block by referring to the samplesin the current picture. The referred samples may be located in the neighborhood of the currentblock or may be located apart according to the intra prediction mode and / or the intra predictiontechnique. The intra prediction modes may include a plurality of non-directional modes and aplurality of directional modes. The non-directional mode may include, for example, a DC modeand a planar mode. The directional mode may include, for example, 33 directional predictionmodes or 65 directional prediction modes according to the degree of detail of the predictiondirection. However, this is merely an example, more or less directional prediction modes maybe used depending on a setting. The intra prediction unit 185 may determine the predictionmode applied to the current block by using a prediction mode applied to a neighboring block.

[80] The inter prediction unit 180 may derive a predicted block for the current block basedon a reference block (reference sample array) specified by a motion vector on a referencepicture. In this case, in order to reduce the amount of motion information transmitted in theinter prediction mode, the motion information may be predicted in units of blocks, subblocks,or samples based on correlation of motion information between the neighboring block and thecurrent block. The motion information may include a motion vector and a reference pictureindex. The motion information may further include inter prediction direction (L0 prediction,L1 prediction, Bi prediction, etc.) information. In the case of inter prediction, the neighboringblock may include a spatial neighboring block present in the current picture and a temporalneighboring block present in the reference picture. The reference picture including thereference block and the reference picture including the temporal neighboring block may be thesame or different. The temporal neighboring block may be called a collocated reference block,a co-located CU (colCU), and the like. The reference picture including the temporalneighboring block may be called a collocated picture (colPic). For example, the inter predictionunit 180 may configure a motion information candidate list based on neighboring blocks andgenerate information specifying which candidate is used to derive a motion vector and / or areference picture index of the current block. Inter prediction may be performed based onvarious prediction modes. For example, in the case of a skip mode and a merge mode, the interprediction unit 180 may use motion information of the neighboring block as motioninformation of the current block. In the case of the skip mode, unlike the merge mode, theresidual signal may not be transmitted. In the case of the motion vector prediction (MVP) mode,the motion vector of the neighboring block may be used as a motion vector predictor, and themotion vector of the current block may be signaled by encoding a motion vector difference andan indicator for a motion vector predictor. The motion vector difference may mean a differencebetween the motion vector of the current block and the motion vector predictor.

[81] The prediction unit may generate a prediction signal based on various predictionmethods and prediction techniques described below. For example, the prediction unit may notonly apply intra prediction or inter prediction but also simultaneously apply both intraprediction and inter prediction, in order to predict the current block. A prediction method ofsimultaneously applying both intra prediction and inter prediction for prediction of the currentblock may be called combined inter and intra prediction (CIIP). In addition, the prediction unitmay perform intra block copy (IBC) for prediction of the current block. Intra block copy maybe used for content image / video coding of a game or the like, for example, screen contentcoding (SCC). IBC is a method of predicting a current picture using a previously reconstructedreference block in the current picture at a location apart from the current block by apredetermined distance. When IBC is applied, the location of the reference block in the currentpicture may be encoded as a vector (block vector) corresponding to the predetermined distance.IBC basically performs prediction in the current picture, but may be performed similarly tointer prediction in that a reference block is derived within the current picture. That is, IBC mayuse at least one of the inter prediction techniques described in the present disclosure.

[82] The prediction signal generated by the prediction unit may be used to generate areconstructed signal or to generate a residual signal. The subtractor 115 may generate a residualsignal (residual block or residual sample array) by subtracting the prediction signal (predictedblock or prediction sample array) output from the prediction unit from the input image signal(original block or original sample array). The generated residual signal may be transmitted tothe transformer 120.

[83] The transformer 120 may generate transform coefficients by applying a transformtechnique to the residual signal. For example, the transform technique may include at least oneof a discrete cosine transform (DCT), a discrete sine transform (DST), a karhunen-loevetransform (KLT), a graph-based transform (GBT), or a conditionally non-linear transform(CNT). Here, the GBT means transform obtained from a graph when relationship informationbetween pixels is represented by the graph. The CNT refers to transform acquired based on aprediction signal generated using all previously reconstructed pixels. In addition, the transformprocess may be applied to square pixel blocks having the same size or may be applied to blockshaving a variable size rather than square.

[84] The quantizer 130 may quantize the transform coefficients and transmit them to theentropy encoder 190. The entropy encoder 190 may encode the quantized signal (informationon the quantized transform coefficients) and output a bitstream. The information on thequantized transform coefficients may be referred to as residual information. The quantizer 130may rearrange quantized transform coefficients in a block form into a one-dimensional vectorform based on a coefficient scanning order and generate information on the quantized transformcoefficients based on the quantized transform coefficients in the one-dimensional vector form.

[85] The entropy encoder 190 may perform various encoding methods such as, for example,exponential Golomb, context-adaptive variable length coding (CAVLC), context-adaptivebinary arithmetic coding (CABAC), and the like. The entropy encoder 190 may encodeinformation necessary for video / image reconstruction other than quantized transformcoefficients (e.g., values of syntax elements, etc.) together or separately. Encoded information(e.g., encoded video / image information) may be transmitted or stored in units of networkabstraction layers (NALs) in the form of a bitstream. The video / image information may furtherinclude information on various parameter sets such as an adaptation parameter set (APS), apicture parameter set (PPS), a sequence parameter set (SPS), or a video parameter set (VPS).In addition, the video / image information may further include general constraint information.The signaled information, transmitted information and / or syntax elements described in thepresent disclosure may be encoded through the above-described encoding procedure andincluded in the bitstream.

[86] The bitstream may be transmitted over a network or may be stored in a digital storagemedium. The network may include a broadcasting network and / or a communication network,and the digital storage medium may include various storage media such as USB, SD, CD, DVD,Blu-ray, HDD, SSD, and the like. A transmitter (not shown) transmitting a signal output fromthe entropy encoder 190 and / or a storage unit (not shown) storing the signal may be includedas internal / external element of the image encoding apparatus 100. Alternatively, the transmittermay be provided as the component of the entropy encoder 190.

[87] The quantized transform coefficients output from the quantizer 130 may be used togenerate a residual signal. For example, the residual signal (residual block or residual samples)may be reconstructed by applying dequantization and inverse transform to the quantizedtransform coefficients through the dequantizer 140 and the inverse transformer 150.

[88] The adder 155 adds the reconstructed residual signal to the prediction signal outputfrom the inter prediction unit 180 or the intra prediction unit 185 to generate a reconstructedsignal (reconstructed picture, reconstructed block, reconstructed sample array). If there is noresidual for the block to be processed, such as a case where the skip mode is applied, thepredicted block may be used as the reconstructed block. The adder 155 may be called areconstructor or a reconstructed block generator. The generated reconstructed signal may beused for intra prediction of a next block to be processed in the current picture and may be usedfor inter prediction of a next picture through filtering as described below.

[89] The filter 160 may improve subjective / objective image quality by applying filtering tothe reconstructed signal. For example, the filter 160 may generate a modified reconstructedpicture by applying various filtering methods to the reconstructed picture and store themodified reconstructed picture in the memory 170, specifically, a DPB of the memory 170.The various filtering methods may include, for example, deblocking filtering, a sampleadaptive offset, an adaptive loop filter, a bilateral filter, and the like. The filter 160 maygenerate various information related to filtering and transmit the generated information to theentropy encoder 190 as described later in the description of each filtering method. Theinformation related to filtering may be encoded by the entropy encoder 190 and output in theform of a bitstream.

[90] The modified reconstructed picture transmitted to the memory 170 may be used as thereference picture in the inter prediction unit 180. When inter prediction is applied through theimage encoding apparatus 100, prediction mismatch between the image encoding apparatus100 and the image decoding apparatus may be avoided and encoding efficiency may beimproved.

[91] The DPB of the memory 170 may store the modified reconstructed picture for use asa reference picture in the inter prediction unit 180. The memory 170 may store the motioninformation of the block from which the motion information in the current picture is derived(or encoded) and / or the motion information of the blocks in the picture that have already beenreconstructed. The stored motion information may be transmitted to the inter prediction unit180 and used as the motion information of the spatial neighboring block or the motioninformation of the temporal neighboring block. The memory 170 may store reconstructedsamples of reconstructed blocks in the current picture and may transfer the reconstructedsamples to the intra prediction unit 185.

[92] Overview of image decoding apparatus

[93] FIG. 3 is a view schematically showing an image decoding apparatus, to which anembodiment of the present disclosure is applicable.

[94] As shown in FIG. 3, the image decoding apparatus 200 may include an entropy decoder210, a dequantizer 220, an inverse transformer 230, an adder 235, a filter 240, a memory 250,an inter prediction unit 260 and an intra prediction unit 265. The inter prediction unit 260 andthe intra prediction unit 265 may be collectively referred to as a "prediction unit". Thedequantizer 220 and the inverse transformer 230 may be included in a residual processor.

[95] All or at least some of a plurality of components configuring the image decodingapparatus 200 may be configured by a hardware component (e.g., a decoder or a processor)according to an embodiment. In addition, the memory 250 may include a decoded picturebuffer (DPB) or may be configured by a digital storage medium.

[96] The image decoding apparatus 200, which has received a bitstream includingvideo / image information, may reconstruct an image by performing a process corresponding toa process performed by the image encoding apparatus 100 of FIG. 2. For example, the imagedecoding apparatus 200 may perform decoding using a processing unit applied in the imageencoding apparatus. Thus, the processing unit of decoding may be a coding unit, for example.The coding unit may be acquired by partitioning a coding tree unit or a largest coding unit. Thereconstructed image signal decoded and output through the image decoding apparatus 200 maybe reproduced through a reproducing apparatus (not shown).

[97] The image decoding apparatus 200 may receive a signal output from the imageencoding apparatus of FIG. 2 in the form of a bitstream. The received signal may be decodedthrough the entropy decoder 210. For example, the entropy decoder 210 may parse thebitstream to derive information (e.g., video / image information) necessary for imagereconstruction (or picture reconstruction). The video / image information may further includeinformation on various parameter sets such as an adaptation parameter set (APS), a pictureparameter set (PPS), a sequence parameter set (SPS), or a video parameter set (VPS). Inaddition, the video / image information may further include general constraint information. Theimage decoding apparatus may further decode picture based on the information on theparameter set and / or the general constraint information. Signaled / received information and / orsyntax elements described in the present disclosure may be decoded through the decodingprocedure and obtained from the bitstream. For example, the entropy decoder 210 decodes theinformation in the bitstream based on a coding method such as exponential Golomb coding,CAVLC, or CABAC, and output values of syntax elements required for image reconstructionand quantized values of transform coefficients for residual. More specifically, the CABACentropy decoding method may receive a bin corresponding to each syntax element in thebitstream, determine a context model using a decoding target syntax element information,decoding information of a neighboring block and a decoding target block or information of asymbol / bin decoded in a previous stage, and perform arithmetic decoding on the bin bypredicting a probability of occurrence of a bin according to the determined context model, andgenerate a symbol corresponding to the value of each syntax element. In this case, the CABACentropy decoding method may update the context model by using the information of thedecoded symbol / bin for a context model of a next symbol / bin after determining the contextmodel. The information related to the prediction among the information decoded by the entropydecoder 210 may be provided to the prediction unit (the inter prediction unit 260 and the intraprediction unit 265), and the residual value on which the entropy decoding was performed inthe entropy decoder 210, that is, the quantized transform coefficients and related parameterinformation, may be input to the dequantizer 220. In addition, information on filtering amonginformation decoded by the entropy decoder 210 may be provided to the filter 240. Meanwhile,a receiver (not shown) for receiving a signal output from the image encoding apparatus may befurther configured as an internal / external element of the image decoding apparatus 200, or thereceiver may be a component of the entropy decoder 210.

[98] Meanwhile, the image decoding apparatus according to the present disclosure may bereferred to as a video / image / picture decoding apparatus. The image decoding apparatus maybe classified into an information decoder (video / image / picture information decoder) and asample decoder (video / image / picture sample decoder). The information decoder may includethe entropy decoder 210. The sample decoder may include at least one of the dequantizer 220,the inverse transformer 230, the adder 235, the filter 240, the memory 250, the inter predictionunit 260 or the intra prediction unit 265.

[99] The dequantizer 220 may dequantize the quantized transform coefficients and outputthe transform coefficients. The dequantizer 220 may rearrange the quantized transformcoefficients in the form of a two-dimensional block. In this case, the rearrangement may beperformed based on the coefficient scanning order performed in the image encoding apparatus.The dequantizer 220 may perform dequantization on the quantized transform coefficients byusing a quantization parameter (e.g., quantization step size information) and obtain transformcoefficients.

[100] The inverse transformer 230 may inversely transform the transform coefficients toobtain a residual signal (residual block, residual sample array).

[101] The prediction unit may perform prediction on the current block and generate apredicted block including prediction samples for the current block. The prediction unit maydetermine whether intra prediction or inter prediction is applied to the current block based onthe information on the prediction output from the entropy decoder 210 and may determine aspecific intra / inter prediction mode (prediction technique).

[102] It is the same as described in the prediction unit of the image encoding apparatus 100that the prediction unit may generate the prediction signal based on various prediction methods(techniques) which will be described later.

[103] The intra prediction unit 265 may predict the current block by referring to the samplesin the current picture. The description of the intra prediction unit 185 is equally applied to theintra prediction unit 265.

[104] The inter prediction unit 260 may derive a predicted block for the current block basedon a reference block (reference sample array) specified by a motion vector on a referencepicture. In this case, in order to reduce the amount of motion information transmitted in theinter prediction mode, motion information may be predicted in units of blocks, subblocks, orsamples based on correlation of motion information between the neighboring block and thecurrent block. The motion information may include a motion vector and a reference pictureindex. The motion information may further include inter prediction direction (L0 prediction,L1 prediction, Bi prediction, etc.) information. In the case of inter prediction, the neighboringblock may include a spatial neighboring block present in the current picture and a temporalneighboring block present in the reference picture. For example, the inter prediction unit 260may configure a motion information candidate list based on neighboring blocks and derive amotion vector of the current block and / or a reference picture index based on the receivedcandidate selection information. Inter prediction may be performed based on various predictionmodes, and the information on the prediction may include information specifying a mode ofinter prediction for the current block.

[105] The adder 235 may generate a reconstructed signal (reconstructed picture,reconstructed block, reconstructed sample array) by adding the obtained residual signal to theprediction signal (predicted block, predicted sample array) output from the prediction unit(including the inter prediction unit 260 and / or the intra prediction unit 265). If there is noresidual for the block to be processed, such as when the skip mode is applied, the predictedblock may be used as the reconstructed block. The description of the adder 155 is equallyapplicable to the adder 235. The adder 235 may be called a reconstructor or a reconstructedblock generator. The generated reconstructed signal may be used for intra prediction of a nextblock to be processed in the current picture and may be used for inter prediction of a nextpicture through filtering as described below.

[106] The filter 240 may improve subjective / objective image quality by applying filtering tothe reconstructed signal. For example, the filter 240 may generate a modified reconstructedpicture by applying various filtering methods to the reconstructed picture and store themodified reconstructed picture in the memory 250, specifically, a DPB of the memory 250.The various filtering methods may include, for example, deblocking filtering, a sampleadaptive offset, an adaptive loop filter, a bilateral filter, and the like.

[107] The (modified) reconstructed picture stored in the DPB of the memory 250 may beused as a reference picture in the inter prediction unit 260. The memory 250 may store themotion information of the block from which the motion information in the current picture isderived (or decoded) and / or the motion information of the blocks in the picture that havealready been reconstructed. The stored motion information may be transmitted to the interprediction unit 260 so as to be utilized as the motion information of the spatial neighboringblock or the motion information of the temporal neighboring block. The memory 250 may storereconstructed samples of reconstructed blocks in the current picture and transfer thereconstructed samples to the intra prediction unit 265.

[108] In the present disclosure, the embodiments described in the filter 160, the interprediction unit 180, and the intra prediction unit 185 of the image encoding apparatus 100 maybe equally or correspondingly applied to the filter 240, the inter prediction unit 260, and theintra prediction unit 265 of the image decoding apparatus 200.

[109] Overview of image partitioning

[110] The video / image coding method according to the present disclosure may be performedbased on an image partitioning structure as follows. Specifically, the procedures of prediction,residual processing ((inverse) transform, (de)quantization, etc.), syntax element coding, andfiltering, which will be described later, may be performed based on a CTU, CU (and / or TU,PU) derived based on the image partitioning structure. The image may be partitioned in blockunits and the block partitioning procedure may be performed in the image partitioner 110 ofthe encoding apparatus. The partitioning related information may be encoded by the entropyencoder 190 and transmitted to the decoding apparatus in the form of a bitstream. The entropydecoder 210 of the decoding apparatus may derive a block partitioning structure of the currentpicture based on the partitioning related information obtained from the bitstream, and based onthis, may perform a series of procedures (e.g., prediction, residual processing, block / picturereconstruction, in-loop filtering, etc.) for image decoding.

[111] Pictures may be partitioned into a sequence of coding tree units (CTUs). FIG. 4 showsan example in which a picture is partitioned into CTUs. The CTU may correspond to a codingtree block (CTB). Alternatively, the CTU may include a coding tree block of luma samples andtwo coding tree blocks of corresponding chroma samples. For example, for a picture thatcontains three sample arrays, the CTU may include an NxN block of luma samples and twocorresponding blocks of chroma samples.

[112] Overview of partitioning of CTU

[113] As described above, the coding unit may be acquired by recursively partitioning thecoding tree unit (CTU) or the largest coding unit (LCU) according to a quad-tree / binarytree / ternary-tree (QT / BT / TT) structure. For example, the CTU may be first partitioned intoquadtree structures. Thereafter, leaf nodes of the quadtree structure may be further partitionedby a multi-type tree structure.

[114] Partitioning according to quadtree means that a current CU (or CTU) is partitioned intoequally four. By partitioning according to quadtree, the current CU may be partitioned into fourCUs having the same width and the same height. When the current CU is no longer partitionedinto the quadtree structure, the current CU corresponds to the leaf node of the quad-treestructure. The CU corresponding to the leaf node of the quadtree structure may be no longerpartitioned and may be used as the above-described final coding unit. Alternatively, the CUcorresponding to the leaf node of the quadtree structure may be further partitioned by a multitype tree structure.

[115] FIG. 5 is a view showing an embodiment of a partitioning type of a block accordingto a multi-type tree structure. Partitioning according to the multi-type tree structure mayinclude two types of splitting according to a binary tree structure and two types of splittingaccording to a ternary tree structure.

[116] The two types of splitting according to the binary tree structure may include verticalbinary splitting (SPLIT_BT_VER) and horizontal binary splitting (SPLIT_BT_HOR). Verticalbinary splitting (SPLIT_BT_VER) means that the current CU is split into equally two in thevertical direction. As shown in FIG. 4, by vertical binary splitting, two CUs having the sameheight as the current CU and having a width which is half the width of the current CU may begenerated. Horizontal binary splitting (SPLIT_BT_HOR) means that the current CU is splitinto equally two in the horizontal direction. As shown in FIG. 5, by horizontal binary splitting,two CUs having a height which is half the height of the current CU and having the same widthas the current CU may be generated.

[117] Two types of splitting according to the ternary tree structure may include verticalternary splitting (SPLIT_TT_VER) and horizontal ternary splitting (SPLIT_TT_HOR). Invertical ternary splitting (SPLIT_TT_VER), the current CU is split in the vertical direction ata ratio of 1:2:1. As shown in FIG. 5, by vertical ternary splitting, two CUs having the sameheight as the current CU and having a width which is 1 / 4 of the width of the current CU and aCU having the same height as the current CU and having a width which is half the width of thecurrent CU may be generated. In horizontal ternary splitting (SPLIT_TT_HOR), the currentCU is split in the horizontal direction at a ratio of 1:2:1. As shown in FIG. 5, by horizontalternary splitting, two CUs having a height which is 1 / 4 of the height of the current CU andhaving the same width as the current CU and a CU having a height which is half the height ofthe current CU and having the same width as the current CU may be generated.

[118] FIG. 6 is a view showing a signaling mechanism of block splitting information in aquadtree with nested multi-type tree structure according to the present disclosure.

[119] Here, the CTU is treated as the root node of the quadtree, and is partitioned for the firsttime into a quadtree structure. Information (e.g., qt_split_flag) specifying whether quadtreesplitting is performed on the current CU (CTU or node (QT_node) of the quadtree) is signaled.For example, when qt_split_flag has a first value (e.g., "1"), the current CU may be quadtreepartitioned. In addition, when qt_split_flag has a second value (e.g., "0"), the current CU is notquadtree-partitioned, but becomes the leaf node (QT_leaf_node) of the quadtree. Each quadtreeleaf node may then be further partitioned into multitype tree structures. That is, the leaf nodeof the quadtree may become the node (MTT_node) of the multi-type tree. In the multitype treestructure, a first flag (e.g., Mtt_split_cu_flag) is signaled to specify whether the current nodeis additionally partitioned. If the corresponding node is additionally partitioned (e.g., if the firstflag is 1), a second flag (e.g., Mtt_split_cu_vertical_flag) may be signaled to specify thesplitting direction. For example, the splitting direction may be a vertical direction if the secondflag is 1 and may be a horizontal direction if the second flag is 0. Then, a third flag (e.g.,Mtt_split_cu_binary_flag) may be signaled to specify whether the split type is a binary splittype or a ternary split type. For example, the split type may be a binary split type when thethird flag is 1 and may be a ternary split type when the third flag is 0. The node of the multitype tree acquired by binary splitting or ternary splitting may be further partitioned into multitype tree structures. However, the node of the multi-type tree may not be partitioned intoquadtree structures. If the first flag is 0, the corresponding node of the multi-type tree is nolonger split but becomes the leaf node (MTT_leaf_node) of the multi-type tree. The CUcorresponding to the leaf node of the multi-type tree may be used as the above-described finalcoding unit.

[120] Based on the mtt_split_cu_vertical_flag and the mtt_split_cu_binary_flag, a multitype tree splitting mode (MttSplitMode) of a CU may be derived as shown in Table 1 below.In the following description, the multi-type tree splitting mode may be referred to as a multitree splitting type or splitting type.

[121] [Table 1]

[122] FIG. 7 is a view showing an example in which a CTU is partitioned into multiple CUsby applying a multi-type tree after applying a quadtree. In FIG. 7, bold block edges 710represent quadtree partitioning and the remaining edges 720 represent multitype treepartitioning. The CU may correspond to a coding block (CB). In an embodiment, the CU mayinclude a coding block of luma samples and two coding blocks of chroma samplescorresponding to the luma samples. A chroma component (sample) CB or TB size may bederived based on a luma component (sample) CB or TB size according to the component ratioaccording to the color format (chroma format, e.g., 4:4:4, 4:2:2, 4:2:0 or the like) of thepicture / image. In case of 4:4:4 color format, the chroma component CB / TB size may be setequal to be luma component CB / TB size. In case of 4:2:2 color format, the width of the chromacomponent CB / TB may be set to half the width of the luma component CB / TB and the heightof the chroma component CB / TB may be set to the height of the luma component CB / TB. Incase of 4:2:0 color format, the width of the chroma component CB / TB may be set to half thewidth of the luma component CB / TB and the height of the chroma component CB / TB may beset to half the height of the luma component CB / TB.

[123] In an embodiment, when the size of the CTU is 128 based on the luma sample unit,the size of the CU may have a size from 128x128 to 4x4 which is the same size as the CTU. Inone embodiment, in case of 4:2:0 color format (or chroma format), a chroma CB size may havea size from 64x64 to 2x2.

[124] Meanwhile, in an embodiment, the CU size and the TU size may be the same.Alternatively, there may be a plurality of TUs in a CU region. The TU size generally representsa luma component (sample) transform block (TB) size.

[125] The TU size may be derived based a largest allowable TB size maxTbSize which is apredetermined value. For example, when the CU size is greater than maxTbSize, a plurality ofTUs (TBs) having maxTbSize may be derived from the CU and transform / inverse transformmay be performed in units of TU (TB). For example, the largest allowable luma TB size maybe 64x64 and the largest allowable chroma TB size may be 32x32. If the width or height of theCB partitioned according to the tree structure is larger than the largest transform width or height,the CB may be automatically (or implicitly) partitioned until the TB size limit in the horizontaland vertical directions is satisfied.

[126] In addition, for example, when intra prediction is applied, an intra predictionmode / type may be derived in units of CU (or CB) and a neighboring reference samplederivation and prediction sample generation procedure may be performed in units of TU (orTB). In this case, there may be one or a plurality of TUs (or TBs) in one CU (or CB) regionand, in this case, the plurality of TUs or (TBs) may share the same intra prediction mode / type.

[127] Meanwhile, for a quadtree coding tree scheme with nested multitype tree, thefollowing parameters may be signaled as SPS syntax elements from the encoding apparatus tothe decoding apparatus. For example, at least one of a CTU size which is a parameterrepresenting the root node size of a quadtree, MinQTSize which is a parameter representingthe minimum allowed quadtree leaf node size, MaxBtSize which is a parameter representingthe maximum allowed binary tree root node size, MaxTtSize which is a parameter representingthe maximum allowed ternary tree root node size, MaxMttDepth which is a parameterrepresenting the maximum allowed hierarchy depth of multi-type tree splitting from a quadtreeleaf node, MinBtSize which is a parameter representing the minimum allowed binary tree leafnode size, or MinTtSize which is a parameter representing the minimum allowed ternary treeleaf node size is signaled.

[128] As an embodiment of using 4:2:0 chroma format, the CTU size may be set to 128x128luma blocks and two 64x64 chroma blocks corresponding to the luma blocks. In this case,MinOTSize may be set to 16x16, MaxBtSize may be set to 128x128, MaxTtSzie may be set to64x64, MinBtSize and MinTtSize may be set to 4x4, and MaxMttDepth may be set to 4.Quadtree partitioning may be applied to the CTU to generate quadtree leaf nodes. The quadtreeleaf node may be called a leaf QT node. Quadtree leaf nodes may have a size from a 16x16size (e.g., the MinOTSize) to a 128x128 size (e.g., the CTU size). If the leaf QT node is128x128, it may not be additionally partitioned into a binary tree / ternary tree. This is because,in this case, even if partitioned, it exceeds MaxBtsize and MaxTtszie (e.g., 64x64). In othercases, leaf QT nodes may be further partitioned into a multitype tree. Therefore, the leaf QTnode is the root node for the multitype tree, and the leaf QT node may have a multitype treedepth (mttDepth) 0 value. If the multitype tree depth reaches MaxMttdepth (e.g., 4), furtherpartitioning may not be considered further. If the width of the multitype tree node is equal toMinBtSize and less than or equal to 2xMinTtSize, then no further horizontal partitioning maybe considered. If the height of the multitype tree node is equal to MinBtSize and less than orequal to 2xMinTtSize, no further vertical partitioning may be considered. When partitioning isnot considered, the encoding apparatus may skip signaling of partitioning information. In thiscase, the decoding apparatus may derive partitioning information with a predetermined value.

[129] Meanwhile, one CTU may include a coding block of luma samples (hereinafterreferred to as a "luma block") and two coding blocks of chroma samples corresponding thereto(hereinafter referred to as "chroma blocks"). The above-described coding tree scheme may beequally or separately applied to the luma block and chroma block of the current CU.Specifically, the luma and chroma blocks in one CTU may be partitioned into the same blocktree structure and, in this case, the tree structure is represented as SINGLE_TREE.Alternatively, the luma and chroma blocks in one CTU may be partitioned into separate blocktree structures, and, in this case, the tree structure may be represented as DUAL_TREE. Thatis, when the CTU is partitioned into dual trees, the block tree structure for the luma block andthe block tree structure for the chroma block may be separately present. In this case, the blocktree structure for the luma block may be called DUAL_TREE_LUMA, and the block treestructure for the chroma component may be called DUAL_TREE_CHROMA. For P and Bslice / tile groups, luma and chroma blocks in one CTU may be limited to have the same codingtree structure. However, for I slice / tile groups, luma and chroma blocks may have a separateblock tree structure from each other. If the separate block tree structure is applied, the lumaCTB may be partitioned into CUs based on a particular coding tree structure, and the chromaCTB may be partitioned into chroma CUs based on another coding tree structure. That is, thismeans that a CU in an I slice / tile group, to which the separate block tree structure is applied,may include a coding block of luma components or coding blocks of two chroma componentsand a CU of a P or B slice / tile group may include blocks of three color components (a lumacomponent and two chroma components).

[130] Although a quadtree coding tree structure with a nested multitype tree has beendescribed, a structure in which a CU is partitioned is not limited thereto. For example, the BTstructure and the TT structure may be interpreted as a concept included in a multiplepartitioning tree (MPT) structure, and the CU may be interpreted as being partitioned throughthe QT structure and the MPT structure. In an example where the CU is partitioned through aQT structure and an MPT structure, a syntax element (e.g., MPT_split_type) includinginformation on how many blocks the leaf node of the QT structure is partitioned into and asyntax element (ex. MPT_split_mode) including information on which of vertical andhorizontal directions the leaf node of the QT structure is partitioned into may be signaled todetermine a partitioning structure.

[131] In another example, the CU may be partitioned in a different way than the QT structure,BT structure or TT structure. That is, unlike that the CU of the lower depth is partitioned into1 / 4 of the CU of the higher depth according to the QT structure, the CU of the lower depth ispartitioned into 1 / 2 of the CU of the higher depth according to the BT structure, or the CU ofthe lower depth is partitioned into 1 / 4 or 1 / 2 of the CU of the higher depth according to the TTstructure, the CU of the lower depth may be partitioned into 1 / 5, 1 / 3, 3 / 8, 3 / 5, 2 / 3, or 5 / 8 of theCU of the higher depth in some cases, and the method of partitioning the CU is not limitedthereto.

[132] The quadtree coding block structure with the multi-type tree may provide a veryflexible block partitioning structure. Because of the partition types supported in a multi-typetree, different partition patterns may potentially result in the same coding block structure insome cases. In the encoding apparatus and the decoding apparatus, by limiting the occurrenceof such redundant partition patterns, a data amount of partitioning information may be reduced.

[133] For example, FIG. 8 shows redundant splitting patterns which may occur in binary treesplitting and ternary tree splitting. As shown in FIG. 8, continuous binary splitting 810 and 820for one direction of two-step levels have the same coding block structure as binary splitting fora center partition after ternary splitting. In this case, binary tree splitting for center blocks 830and 840 of ternary tree splitting may be prohibited. this prohibition is applicable to CUs of allpictures. When such specific splitting is prohibited, signaling of corresponding syntax elementsmay be modified by reflecting this prohibited case, thereby reducing the number of bitssignaled for splitting. For example, as shown in the example shown in FIG. 8, when binary treesplitting for the center block of the CU is prohibited, a syntax element mtt_split_cu_binary_flagspecifying whether splitting is binary splitting or ternary splitting is not signaled and the valuethereof may be derived as 0 by a decoding apparatus.

[134] Overview of inter prediction

[135] Hereinafter, inter prediction according to the present disclosure will be described.

[136] The prediction unit of an image encoding apparatus / image decoding apparatusaccording to the present disclosure may perform inter prediction in units of blocks to derive aprediction sample. Inter prediction may represent prediction derived in a manner that isdependent on data elements (e.g., sample values, motion information, etc.) of picture(s) otherthan a current picture. When inter prediction applies to the current block, a predicted block(prediction block or a prediction sample array) for the current block may be derived based ona reference block (reference sample array) specified by a motion vector on a reference pictureindiated by a reference picture index. In this case, in order to reduce the amount of motioninformation transmitted in an inter prediction mode, motion information of the current blockmay be predicted in units of blocks, subblocks or samples, based on correlation of motioninformation between a neighboring block and the current block. The motion information mayinclude a motion vector and a reference picture index. The motion information may furtherinclude inter prediction type (L0 prediction, L1 prediction, Bi prediction, etc.) information.When applying inter prediction, the neighboring block may include a spatial neighboring blockpresent in the current picture and a temporal neighboring block present in the reference picture.A reference picture including the reference block and a reference picture including the temporalneighboring block may be the same or different. The temporal neighboring block may bereferred to as a collocated reference block, collocated CU (ColCU) or colBlock, and thereference picture including the temporal neighboring block may be referred to as a collocatedpicture (colPic) or colPicture. For example, a motion information candidate list may beconstructed based on the neighboring blocks of the current block, and flag or index informationspecifying which candidate is selected (used) may be signaled in order to derive the motionvector of the current block and / or the reference picture index.

[137] Inter prediction may be performed based on various prediction modes. For example,in the case of a skip mode and a merge mode, the motion information of the current block maybe equal to the motion information of the selected neighboring block. In the case of the skipmode, a residual signal may not be transmitted unlike the merge mode. In the case of a motioninformation prediction (MVP) mode, the motion vector of the selected neighboring block maybe used as a motion vector predictor and a motion vector difference may be signaled. In thiscase, the motion vector of the current block may be derived using a sum of the motion vectorpredictor and the motion vector difference. In the present disclosure, the MVP mode may havethe same meaning as advanced motion vector prediction (AMVP).

[138] The motion information may include L0 motion information and / or L1 motioninformation according to the inter prediction type (L0 prediction, L1 prediction, Bi prediction,etc.). The motion vector in an L0 direction may be referred to as an L0 motion vector or MVL0,and the motion vector in an L1 direction may be referred to as an L1 motion vector or MVL1.Prediction based on the L0 motion vector may be referred to as L0 prediction, prediction basedon the L1 motion vector may be referred to as L1 prediction, and prediction based both the L0motion vector and the L1 motion vector may be referred to as Bi prediction. Here, the L0motion vector may specify a motion vector associated with a reference picture list L0 (L0) andthe L1 motion vector may specify a motion vector associated with a reference picture list L1(L1). The reference picture list L0 may include pictures before the current picture in outputorder as reference pictures, and the reference picture list L1 may include pictures after thecurrent picture in output order. The previous pictures may be referred to as forward (reference)pictures and the subsequent pictures may be referred to as reverse (reference) pictures. Thereference picture list L0 may further include pictures after the current picture in output orderas reference pictures. In this case, within the reference picture list L0, the previous picturesmay be first indexed and the subsequent pictures may then be indexed. The reference picturelist L1 may further include pictures before the current picture in output order as referencepictures. In this case, within the reference picture list L1, the subsequent pictures may be firstindexed and the previous pictures may then be indexed. Here, the output order maycorrespond to picture order count (POC) order.

[139] FIG. 9 is a flowchart illustrating an inter prediction based video / image encodingmethod.

[140] FIG. 10 is a view illustrating the configuration of an inter prediction unit 180 accordingto the present disclosure.

[141] The encoding method of FIG. 9 may be performed by the image encoding apparatusof FIG. 2. Specifically, step S610 may be performed by the inter prediction unit 180, and stepS620 may be performed by the residual processor. Specifically, step S620 may be performedby the subtractor 115. Step S630 may be performed by the entropy encoder 190. The predictioninformation of step S630 may be derived by the inter prediction unit 180, and the residualinformation of step S630 may be derived by the residual processor. The residual informationis information on the residual samples. The residual information may include information onquantized transform coefficients for the residual samples. As described above, the residualsamples may be derived as transform coefficients through the transformer 120 of the imageencoding apparatus, and the transform coefficients may be derived as quantized transformcoefficients through the quantizer 130. Information on the quantized transform coefficientsmay be encoded by the entropy encoder 190 through a residual coding procedure.

[142] The image encoding apparatus may perform inter prediction on a current block (S610).The image encoding apparatus may derive an inter prediction mode and motion information ofthe current block and generate prediction samples of the current block. Here, inter predictionmode determination, motion information derivation and prediction samples generationprocedures may be simultaneously performed or any one thereof may be performed before theother procedures. For example, as shown in FIG. 10, the inter prediction unit 180 of the imageencoding apparatus may include a prediction mode determination unit 181, a motioninformation derivation unit 182 and a prediction sample derivation unit 183. The predictionmode determination unit 181 may determine the prediction mode of the current block, themotion information derivation unit 182 may derive the motion information of the current block,and the prediction sample derivation unit 183 may derive the prediction samples of the currentblock. For example, the inter prediction unit 180 of the image encoding apparatus may searchfor a block similar to the current block within a predetermined area (search area) of referencepictures through motion estimation, and derive a reference block whose a difference from thecurrent block is equal to or less than a predetermined criterion or a minimum. Based on this, areference picture index specifying a reference picture in which the reference block is locatedmay be derived, and a motion vector may be derived based on a position difference betweenthe reference block and the current block. The image encoding apparatus may determine amode applying to the current block among various prediction modes. The image encodingapparatus may compare rate-distortion (RD) costs for the various prediction modes anddetermine an optimal prediction mode of the current block. However, the method ofdetermining the prediction mode of the current block by the image encoding apparatus is notlimited to the above example, and various methods may be used.

[143] For example, when a skip mode or a merge mode applies to the current block, theimage encoding apparatus may derive merge candidates from neighboring blocks of the currentblock and construct a merge candidate list using the derived merge candidates. In addition, theimage encoding apparatus may derive a reference block whose a difference from the currentblock is equal to or less than a predetermined criterion or a minimum, among reference blocksspecified by merge candidates included in the merge candidate list. In this case, a mergecandidate associated with the derived reference block may be selected, and merge indexinformation specifying the selected merge candidate may be generated and signaled to an imagedecoding apparatus. The motion information of the current block may be derived using themotion information of the selected merge candidate.

[144] As another example, when an MVP mode applies to the current block, the imageencoding apparatus may derive motion vector predictor (mvp) candidates from the neighboringblocks of the current block and construct an mvp candidate list using the derived mvpcandidates. In addition, the image encoding apparatus may use the motion vector of the mvpcandidate selected from among the mvp candidates included in the mvp candidate list as themvp of the current block. In this case, for example, the motion vector indicating the referenceblock derived by the above-described motion estimation may be used as the motion vector ofthe current block, an mvp candidate with a motion vector having a smallest difference from themotion vector of the current block among the mvp candidates may be the selected mvpcandidate. A motion vector difference (MVD) which is a difference obtained by subtractingthe mvp from the motion vector of the current block may be derived. In this case, indexinformation specifying the selected mvp candidate and information on the MVD may besignaled to the image decoding apparatus. In addition, when applying the MVP mode, the valueof the reference picture index may be constructed as reference picture index information andseparately signaled to the image decoding apparatus.

[145] The image encoding apparatus may derive residual samples based on the predictionsamples (S620). The image encoding apparatus may derive the residual samples throughcomparison between original samples of the current block and the prediction samples. Forexample, the residual sample may be derived by subtracting a corresponding prediction samplefrom an original sample.

[146] The image encoding apparatus may encode image information including predictioninformation and residual information (S630). The image encoding apparatus may output theencoded image information in the form of a bitstream. The prediction information may includeprediction mode information (e.g., skip flag, merge flag or mode index, etc.) and informationon motion information as information related to the prediction procedure. Among theprediction mode information, the skip flag specifies whether a skip mode applies to the currentblock, and the merge flag specifies whether the merge mode applies to the current block.Alternatively, the prediction mode information may specify one of a plurality of predictionmodes, such as a mode index. When the skip flag and the merge flag are 0, it may be determinedthat the MVP mode applies to the current block. The information on the motion informationmay include candidate selection information (e.g., merge index, mvp flag or mvp index) whichis information for deriving a motion vector. Among the candidate selection information, themerge index may be signaled when the merge mode applies to the current block and may beinformation for selecting one of merge candidates included in a merge candidate list. Amongthe candidate selection information, the mvp flag or the mvp index may be signaled when theMVP mode applies to the current block and may be information for selecting one of mvpcandidates in an mvp candidate list. In addition, the information on the motion informationmay include information on the above-described MVD and / or reference picture indexinformation. In addition, the information on the motion information may include informationspecifying whether to apply L0 prediction, L1 prediction or Bi prediction. The residualinformation is information on the residual samples. The residual information may includeinformation on quantized transform coefficients for the residual samples.

[147] The output bitstream may be stored in a (digital) storage medium and transmitted tothe image decoding apparatus or may be transmitted to the image decoding apparatus via anetwork.

[148] As described above, the image encoding apparatus may generate a reconstructedpicture (a picture including reconstructed samples and a reconstructed block) based on thereference samples and the residual samples. This is for the image encoding apparatus to derivethe same prediction result as that performed by the image decoding apparatus, therebyincreasing coding efficiency. Accordingly, the image encoding apparatus may store thereconstructed picture (or the reconstructed samples and the reconstructed block) in a memoryand use the same as a reference picture for inter prediction. As described above, an in-loopfiltering procedure is further applicable to the reconstructed picture.

[149] FIG. 11 is a flowchart illustrating an inter prediction based video / image decodingmethod.

[150] FIG. 12 is a view illustrating the configuration of an inter prediction unit 260 accordingto the present disclosure.

[151] The image decoding apparatus may perform operation corresponding to operationperformed by the image encoding apparatus. The image decoding apparatus may performprediction on a current block based on received prediction information and derive predictionsamples.

[152] The decoding method of FIG. 11 may be performed by the image decoding apparatusof FIG. 3. Steps S810 to S830 may be performed by the inter prediction unit 260, and theprediction information of step S810 and the residual information of step S840 may be obtainedfrom a bitstream by the entropy decoder 210. The residual processor of the image decodingapparatus may derive residual samples for a current block based on the residual information(S840). Specifically, the dequantizer 220 of the residual processor may perform dequantizationbased on dequantized transform coefficients derived based on the residual information to derivetransform coefficients, and the inverse transformer 230 of the residual processor may performinverse transform on the transform coefficients to derive the residual samples for the currentblock. Step S850 may be performed by the adder 235 or the reconstructor.

[153] Specifically, the image decoding apparatus may determine the prediction mode of thecurrent block based on the received prediction information (S810). The image decodingapparatus may determine which inter prediction mode applies to the current block based on theprediction mode information in the prediction information.

[154] For example, it may be determined whether the skip mode applies to the current blockbased on the skip flag. In addition, it may be determined whether the merge mode or the MVPmode applies to the current block based on the merge flag. Alternatively, one of various interprediction mode candidates may be selected based on the mode index. The inter predictionmode candidates may include a skip mode, a merge mode and / or an MVP mode or may includevarious inter prediction modes which will be described below.

[155] The image decoding apparatus may derive the motion information of the current blockbased on the determined inter prediction mode (S820). For example, when the skip mode orthe merge mode applies to the current block, the image decoding apparatus may construct amerge candidate list, which will be described below, and select one of merge candidatesincluded in the merge candidate list. The selection may be performed based on the abovedescribed candidate selection information (merge index). The motion information of thecurrent block may be derived using the motion information of the selected merge candidate.For example, the motion information of the selected merge candidate may be used as the motioninformation of the current block.

[156] As another example, when the MVP mode applies to the current block, the imagedecoding apparatus may construct an mvp candidate list and use the motion vector of an mvpcandidate selected from among mvp candidates included in the mvp candidate list as an mvpof the current block. The selection may be performed based on the above-described candidateselection information (mvp flag or mvp index). In this case, the MVD of the current block maybe derived based on information on the MVD, and the motion vector of the current block maybe derived based on mvp and MVD of the current block. In addition, the reference picture indexof the current block may be derived based on the reference picture index information. A pictureindicated by the reference picture index in the reference picture list of the current block maybe derived as a reference picture referenced for inter prediction of the current block.

[157] The image decoding apparatus may generate prediction samples of the current blockbased on motion information of the current block (S830). In this case, the reference picturemay be derived based on the reference picture index of the current block, and the predictionsamples of the current block may be derived using the samples of the reference block indicatedby the motion vector of the current block on the reference picture. In some cases, a predictionsample filtering procedure may be further performed on all or some of the prediction samplesof the current block.

[158] For example, as shown in FIG. 12, the inter prediction unit 260 of the image decodingapparatus may include a prediction mode determination unit 261, a motion informationderivation unit 262 and a prediction sample derivation unit 263. In the inter prediction unit 260of the image decoding apparatus, the prediction mode determination unit 261 may determinethe prediction mode of the current block based on the received prediction mode information,the motion information derivation unit 262 may derive the motion information (a motion vectorand / or a reference picture index, etc.) of the current block based on the received motioninformation, and the prediction sample derivation unit 263 may derive the prediction samplesof the current block.

[159] The image decoding apparatus may generate residual samples of the current blockbased the received residual information (S840). The image decoding apparatus may generatethe reconstructed samples of the current block based on the prediction samples and the residualsamples and generate a reconstructed picture based on this (S850). Thereafter, an in-loopfiltering procedure is applicable to the reconstructed picture as described above.

[160] As described above, the inter prediction procedure may include step of determining aninter prediction mode, step of deriving motion information according to the determinedprediction mode, and step of performing prediction (generating prediction samples) based onthe derived motion information. The inter prediction procedure may be performed by the imageencoding apparatus and the image decoding apparatus, as described above.

[161] Hereinafter, the step of deriving the motion information according to the predictionmode will be described in greater detail.

[162] As described above, inter prediction may be performed using motion information of acurrent block. An image encoding apparatus may derive optimal motion information of acurrent block through a motion estimation procedure. For example, the image encodingapparatus may search for a similar reference block with high correlation within a predeterminedsearch range in the reference picture using an original block in an original picture for the currentblock in fractional pixel unit, and derive motion information using the same. Similarity of theblock may be calculated based on a sum of absolute differences (SAD) between the currentblock and the reference block. In this case, motion information may be derived based on areference block with a smallest SAD in the search area. The derived motion information maybe signaled to an image decoding apparatus according to various methods based on an interprediction mode.

[163] When a merge mode applies to a current block, motion information of the current blockis not directly transmitted and motion information of the current block is derived using motioninformation of a neighboring block. Accordingly, motion information of a current predictionblock may be indicated by transmitting flag information specifying that the merge mode is usedand candidate selection information (e.g., a merge index) specifying which neighboring blockis used as a merge candidate. In the present disclosure, since the current block is a unit ofprediction performance, the current block may be used as the same meaning as the currentprediction block, and the neighboring block may be used as the same meaning as a neighboringprediction block.

[164] The image encoding apparatus may search for merge candidate blocks used to derivethe motion information of the current block to perform the merge mode. For example, up tofive merge candidate blocks may be used, without being limited thereto. The maximum numberof merge candidate blocks may be transmitted in a slice header or a tile group header, withoutbeing limited thereto. After finding the merge candidate blocks, the image encoding apparatusmay generate a merge candidate list and select a merge candidate block with smallest RD costas a final merge candidate block.

[165] The present disclosure provides various embodiments for the merge candidate blocksconfiguring the merge candidate list. The merge candidate list may use, for example, five mergecandidate blocks. For example, four spatial merge candidates and one temporal mergecandidate may be used.

[166] FIG. 13 is a view illustrating neighboring blocks available as a spatial merge candidate.

[167] FIG. 14 is a view schematically illustrating a merge candidate list construction methodaccording to an example of the present disclosure.

[168] An image encoding / decoding apparatus may insert, into a merge candidate list, spatialmerge candidates derived by searching for spatial neighboring blocks of a current block(S1110). For example, as shown in FIG. 13, the spatial neighboring blocks may include abottom-left corner neighboring block A0, a left neighboring block A1, a top-right cornerneighboring block B0, a top neighboring block B1, and a top-left corner neighboring block B2of the current block. However, this is an example and, in addition to the above-described spatialneighboring blocks, additional neighboring blocks such as a right neighboring block, a bottomneighboring block and a bottom-right neighboring block may be further used as the spatialneighboring blocks. The image encoding / decoding apparatus may detect available blocks bysearching for the spatial neighboring blocks based on priority and derive motion informationof the detected blocks as the spatial merge candidates. For example, the imageencoding / decoding apparatus may construct a merge candidate list by searching for the fiveblocks shown in FIG. 13 in order of A1, B1, B0, A0 and B2 and sequentially indexing availablecandidates.

[169] The image encoding / decoding apparatus may insert, into the merge candidate list, atemporal merge candidate derived by searching for temporal neighboring blocks of the currentblock (S1120). The temporal neighboring blocks may be located on a reference picture whichis different from a current picture in which the current block is located. A reference picture inwhich the temporal neighboring block is located may be referred to as a collocated picture ora col picture. The temporal neighboring block may be searched for in order of a bottom-rightcorner neighboring block and a bottom-right center block of the co-located block for the currentblock on the col picture. Meanwhile, when applying motion data compression in order toreduce memory load, specific motion information may be stored as representative motioninformation for each predetermined storage unit for the col picture. In this case, motioninformation of all blocks in the predetermined storage unit does not need to be stored, therebyobtaining motion data compression effect. In this case, the predetermined storage unit may bepredetermined as, for example, 16x16 sample unit or 8x8 sample unit or size information ofthe predetermined storage unit may be signaled from the image encoding apparatus to theimage decoding apparatus. When applying the motion data compression, the motioninformation of the temporal neighboring block may be replaced with the representative motioninformation of the predetermined storage unit in which the temporal neighboring block islocated. That is, in this case, from the viewpoint of implementation, the temporal mergecandidate may be derived based on the motion information of a prediction block covering anarithmetic left-shifted position after an arithmetic right shift by a predetermined value basedon coordinates (top-left sample position) of the temporal neighboring block, not a predictionblock located on the coordinates of the temporal neighboring block. For example, when thepredetermined storage unit is a 2nx2nsample unit and the coordinates of the temporalneighboring block are (xTnb, yTnb), the motion information of a prediction block located at amodified position ((xTnb>>n)<<n), (yTnb>>n)<<n)) may be used for the temporal mergecandidate. Specifically, for example, when the predetermined storage unit is a 16x16 sampleunit and the coordinates of the temporal neighboring block are (xTnb, yTnb), the motioninformation of a prediction block located at a modified position ((xTnb>>4)<<4),(yTnb>>4)<<4)) may be used for the temporal merge candidate. Alternatively, for example,when the predetermined storage unit is an 8x8 sample unit and the coordinates of the temporalneighboring block are (xTnb, yTnb), the motion information of a prediction block located at amodified position ((xTnb>>3)<<3), (yTnb>>3)<<3)) may be used for the temporal mergecandidate.

[170] Referring to FIG. 14 again, the image encoding / decoding apparatus may checkwhether the number of current merge candidates is less than a maximum number of mergecandidates (S1130). The maximum number of merge candidates may be predefined or signaledfrom the image encoding apparatus to the image decoding apparatus. For example, the imageencoding apparatus may generate and encode information on the maximum number of mergecandidates and transmit the encoded information to the image decoding apparatus in the formof a bitstream. When the maximum number of merge candidates is satisfied, a subsequentcandidate addition process S1140 may not be performed.

[171] When the number of current merge candidates is less than the maximum number ofmerge candidates as a checked result of step S1130, the image encoding / decoding apparatusmay derive an additional merge candidate according to a predetermined method and then insertthe additional merge candidate to the merge candidate list (S1140).

[172] When the number of current merge candidates is not less than the maximum numberof merge candidates as a checked result of step S1130, the image encoding / decoding apparatusmay end the construction of the merge candidate list. In this case, the image encoding apparatusmay select an optimal merge candidate from among the merge candidates configuring themerge candidate list, and signal candidate selection information (e.g., merge index) specifyingthe selected merge candidate to the image decoding apparatus. The image decoding apparatusmay select the optimal merge candidate based on the merge candidate list and the candidateselection information.

[173] The motion information of the selected merge candidate may be used as the motioninformation of the current block, and the prediction samples of the current block may be derivedbased on the motion information of the current block, as described above. The image encodingapparatus may derive the residual samples of the current block based on the prediction samplesand signal residual information of the residual samples to the image decoding apparatus. Theimage decoding apparatus may generate reconstructed samples based on the residual samplesderived based on the residual information and the prediction samples and generate thereconstructed picture based on the same, as described above.

[174] When applying a skip mode to the current block, the motion information of the currentblock may be derived using the same method as the case of applying the merge mode. However,when applying the skip mode, a residual signal for a corresponding block is omitted and thusthe prediction samples may be directly used as the reconstructed samples.

[175] When applying an MVP mode to the current block, a motion vector predictor (mvp)candidate list may be generated using a motion vector of reconstructed spatial neighboringblocks (e.g., the neighboring blocks shown in FIG. 13) and / or a motion vector correspondingto the temporal neighboring blocks (or Col blocks). That is, the motion vector of thereconstructed spatial neighboring blocks and the motion vector corresponding to the temporalneighboring blocks may be used as motion vector predictor candidates of the current block.When applying bi-prediction, an mvp candidate list for L0 motion information derivation andan mvp candidate list for L1 motion information derivation are individually generated and used.Prediction information (or information on prediction) of the current block may includecandidate selection information (e.g., an MVP flag or an MVP index) specifying an optimalmotion vector predictor candidate selected from among the motion vector predictor candidatesincluded in the mvp candidate list. In this case, a prediction unit may select a motion vectorpredictor of a current block from among the motion vector predictor candidates included in themvp candidate list using the candidate selection information. The prediction unit of the imageencoding apparatus may obtain and encode a motion vector difference (MVD) between themotion vector of the current block and the motion vector predictor and output the encodedMVD in the form of a bitstream. That is, the MVD may be obtained by subtracting the motionvector predictor from the motion vector of the current block. The prediction unit of the imagedecoding apparatus may obtain a motion vector difference included in the information onprediction and derive the motion vector of the current block through addition of the motionvector difference and the motion vector predictor. The prediction unit of the image encodingapparatus may obtain or derive a reference picture index specifying a reference picture fromthe information on prediction.

[176] FIG. 15 is a view schematically illustrating a motion vector predictor candidate listconstruction method according to an example of the present disclosure.

[177] First, a spatial candidate block of a current block may be searched for and availablecandidate blocks may be inserted into an mvp candidate list (S1210). Thereafter, it isdetermined whether the number of mvp candidates included in the mvp candidate list is lessthan 2 (S1220) and, when the number of mvp candidates is two, construction of the mvpcandidate list may be completed.

[178] In step S1220, when the number of available spatial candidate blocks is less than 2, atemporal candidate block of the current block may be searched for and available candidateblocks may be inserted into the mvp candidate list (S1230). When the temporal candidateblocks are not available, a zero motion vector may be inserted into the mvp candidate list,thereby completing construction of the mvp candidate list.

[179] Meanwhile, when applying an mvp mode, a reference picture index may be explicitlysignaled. In this case, a reference picture index refidxL0 for L0 prediction and a referencepicture index refidxL1 for L1 prediction may be distinguishably signaled. For example, whenapplying the MVP mode and applying Bi prediction, both information on refidxL0 andinformation on refidxL1 may be signaled.

[180] As described above, when applying the MVP mode, information on MVP derived bythe image encoding apparatus may be signaled to the image decoding apparatus. Informationon the MVD may include, for example, information specifying x and y components for anabsolute value (MVD absolute value) and a sign of the MVD. In this case, when the MVDabsolute value is greater than 0, whether the MVD absolute value is greater than 1 andinformation specifying an MVD remainder may be signaled stepwise. For example,information specifying whether the MVD absolute value is greater than 1 may be signaled onlywhen a value of flag information specifying whether the MVD absolute value is greater than 0is 1.

[181] FIG. 16 is a view illustrating a syntax structure for transmitting MVD from an imageencoding apparatus to an image decoding apparatus according to an embodiment of the presentdisclosure.

[182] In FIG. 16, abs_mvd_greater0_flag [0] specifies whether the absolute value of the x component of MVD is greater than 0, and abs_mvd_greater0_flag [1] specifies the absolute value of the y component of MVD is greater than 0. Similarly, abs_mvd_greater1_flag [0] specifies whether the absolute value of the x component of MVD is greater than 1, and abs_mvd_greater1_flag [1] specifies whether the absolute value of the y component of MVD is greater than 1. As shown in FIG. 16, abs_mvd_greater1_flag may be transmitted only when abs_mvd_greater0_flag is 1. In FIG. 16, abs_mvd_minus2 may specify a value obtained by subtracting 2 from the absolute value of MVD, and mvd_sign_flag specify whether the sign of MVD is positive or negative. Using the syntax structure shown in FIG. 16, MVD may be derived as shown in Equation 1 below.

[183] [Equation 1]

[184] MVD [compIdx] = abs_mvd_greater0_flag [compIdx] *(abs_mvd_minus2 [compIdx] + 2) *(1 - 2 * mvd_sign_flag [compIdx])

[185] Meanwhile, MVD (MVDL0) for L0 prediction and MVD (MVDL1) for L1 predictionmay be distinguishably signaled, and the information on MVD may include information onMVDL0 and / or information on MVDL1. For example, when applying the MVP mode andapplying BI prediction to the current block, both the information on MVDL0 and theinformation on MVDL1 may be signaled.

[186] Overview of intra block copy (IBC) prediction

[187] Hereinafter, IBC prediction according to the present disclosure will be described.

[188] IBC prediction may be performed by a prediction unit of an image encoding / decodingapparatus. IBC prediction may be simply referred to as IBC. The IBC may be used for contentimage / moving image coding such as screen content coding (SCC). The IBC prediction may bebasically performed in the current picture, but may be performed similarly to inter predictionin that a reference block is derived within the current picture. That is, IBC may use at least oneof inter prediction techniques described in the present disclosure. For example, IBC may useat least one of the above-described motion information (motion vector) derivation methods. Atleast one of the inter prediction techniques may be partially modified and used in considerationof the IBC prediction. The IBC may refer to a current picture and thus may be referred to ascurrent picture referencing (CPR).

[189] For IBC, the image encoding apparatus may perform block matching (BM) and derivean optimal block vector (or motion vector) for a current block (e.g., a CU). The derived blockvector (or motion vector) may be signaled to the image decoding apparatus through a bitstreamusing a method similar to signaling of motion information (motion vector) in the abovedescribed inter prediction. The image decoding apparatus may derive a reference block for thecurrent block in the current picture through the signaled block vector (motion vector), andderive a prediction signal (predicted block or prediction samples) for the current block throughthis. Here, the block vector (or motion vector) may specify displacement from the current blockto a reference block located in an already reconstructed area in the current picture. Accordingly,the block vector (or the motion vector) may be referred to a displacement vector. Hereinafter,in IBC, the motion vector may correspond to the block vector or the displacement vector. Themotion vector of the current block may include a motion vector (luma motion vector) for aluma component or a motion vector (chroma motion vector) for a chroma component. Forexample, the luma motion vector for an IBC-coded CU may be an integer sample unit (that is,integer precision). The chroma motion vector may be clipped in integer sample units. Asdescribed above, IBC may use at least one of inter prediction techniques, and, for example, theluma motion vector may be encoded / decoded using the above-described merge mode or MVPmode.

[190] When applying a merge mode to the luma IBC block, a merge candidate list for theluma IBC block may be constructed similarly to a merge candidate list in the inter modedescribed with reference to FIG. 14. However, in the case of the luma IBC block, a temporalneighboring block may not be used as a merge candidate.

[191] When applying the MVP mode to the luma IBC block, an mvp candidate list for theluma IBC block may be constructed similarly to the mvp candidate list in the inter modedescribed with reference to FIG. 15. However, in the case of the luma IBC block, a temporalcandidate block may not be used as the mvp candidate.

[192] In IBC, a reference block is derived from the already reconstructed area in the currentpicture. In this case, in order to reduce memory consumption and complexity of the imagedecoding apparatus, only a predefined area among already reconstructed areas in the currentpicture may be referenced. The predefined area may include a current CTU in which the currentblock is included. By restricting referenceable reconstructed area to the predefined area, theIBC mode may be implemented in hardware using a local on-chip memory.

[193] The image encoding apparatus for performing IBC may search the predefined area todetermine a reference block with smallest RD cost and derive a motion vector (block vector)based on the positions of the reference block and the current block.

[194] Whether to apply IBC to the current block may be signaled as IBC performanceinformation at a CU level. Information on a signaling method (IBC MVP mode or IBCskip / merger mode) of the motion vector of the current block may be signaled. IBC performanceinformation may be used to determine the prediction mode of the current block. Accordingly,the IBC performance information may be included in information on the prediction mode ofthe current block.

[195] In the case of the IBC skip / merge mode, a merge candidate index may be signaled tospecify a block vector to be used for prediction of the current luma block among block vectorsincluded in the merge candidate list. In this case, the merge candidate list may include IBCencoded neighboring blocks. The merge candidate list may be configured to include spatialmerge candidates and not to include temporal merge candidates. In addition, the mergecandidate list may further include history-based motion vector predictor (HMVP) candidatesand / or pairwise candidates.

[196] In the case of the IBC MVP mode, a block vector difference value may be encodedusing the same method as a motion vector difference value of the above-described inter mode.The block vector prediction method may construct and use an mvp candidate list including twocandidates as predictors similarly to the MVP mode of the inter mode. One of the twocandidates may be derived from a left neighboring block and the other candidate may bederived from a top neighboring block. In this case, only when the left or top neighboring blockis IBC-encoded, candidates may be derived from the corresponding neighboring block. If theleft or top neighboring block is not available, for example, is not IBC-encoded, a default blockvector may be included in the mvp candidate list as a predictor. In addition, information (e.g.,flag) specifying one of two block vector predictors is signaled and used as candidate selectioninformation similarly to the MVP mode of the inter mode. The mvp candidate list may includean HMVP candidate and / or a zero motion vector as the default block vector.

[197] The HMVP candidate may be referred to as a history-based MVP candidate, and anMVP candidate used before encoding / decoding of the current block, a merge candidate or ablock vector candidate may be stored in an HMVP list as HMVP candidates. Thereafter, whenthe merge candidate list of the current block or the mvp candidate list does not include amaximum number of candidates, candidates stored in the HMVP list may be added to the mergecandidate list or mvp candidate list of the current block as HMVP candidates.

[198] The pairwise candidate means a candidate derived by selecting two candidatesaccording to a predetermined order from among candidates already included in the mergecandidate list of the current block and averaging the selected two candidates.

[199] FIG. 17 is a flowchart illustrating an IBC based video / image encoding method.

[200] FIG. 18 is a view illustrating the configuration of a prediction unit for performing anIBC based video / image encoding method according to the present disclosure.

[201] The encoding method of FIG. 17 may be performed by the image encoding apparatusof FIG. 2. Specifically, step S1410 may be performed by the prediction unit and step S1420may be performed by the residual processor. Specifically, step S1420 may be performed by thesubtractor 115. Step S1430 may be performed by the entropy encoder 190. The predictioninformation of step S1430 may be derived by the prediction unit and the residual informationof step S1430 may be derived by the residual processor. The residual information is informationon the residual samples. The residual information may include information on quantizedtransform coefficients for the residual samples. As described above, the residual samples maybe derived by the transform coefficients through the transformer 120 of the image encodingapparatus, and the transform coefficients may be derived by transform coefficients quantizedthrough the quantizer 130. Information on the quantized transform coefficients may be encodedby the entropy encoder 190 through a residual coding procedure.

[202] The image encoding apparatus may perform IBC prediction (IBC based prediction) forthe current block (S1410). The image encoding apparatus may derive a prediction mode andmotion vector (block vector) of the current block and generate prediction samples of the currentblock. The prediction mode may include at least one of the above-described inter predictionmodes. Here, prediction mode determination, motion vector derivation and prediction samplesgeneration procedures may be simultaneously performed or any one procedure may beperformed before the other procedures. For example, as shown in FIG. 18, the prediction unitof the image encoding apparatus for performing an IBC-based video / image encoding methodmay include a prediction mode determination unit, a motion vector derivation unit and aprediction sample derivation unit. The prediction mode determination unit may determine theprediction mode of the current block, the motion vector derivation unit may derives the motionvector of the current block, and the prediction sample derivation unit may derive the predictionsamples of the current block. For example, the prediction unit of the image encoding apparatusmay search for a block similar to the current block in a reconstructed area (or a certain area(search area) of the reconstructed area) of a current picture and derive a reference block whosea difference from the current block is equal to or less than a certain criterion or a minimum.The image encoding apparatus may derive a motion vector based on a displacement differencebetween the reference block and the current block. The image encoding apparatus maydetermine a mode applying to the current block among various prediction modes. The imageencoding apparatus may compare RD costs for the various prediction modes and determine anoptimal prediction mode for the current block. However, a method of determining theprediction mode for the current block by the image encoding apparatus is not limited to theabove example and various methods may be used.

[203] For example, when applying a skip mode or a merge mode to the current block, theimage encoding apparatus may derive merge candidates from neighboring blocks of the currentblock and construct a merge candidate list using the derived merge candidates. In addition, theimage encoding apparatus may derive a reference block whose a difference from the currentblock is equal to or less than a certain criterion or a minimum among reference blocks indicatedby the merge candidates included in the merge candidate list. In this case, a merge candidateassociated with the derived reference block may be selected and merge index informationspecifying the selected merge candidate may be generated and signaled to the image decodingapparatus. Using the motion vector of the selected merge candidate, the motion vector of thecurrent block may be derived.

[204] As another example, when applying an MVP mode to the current block, the imageencoding apparatus may derive motion vector predictor (mvp) candidates from the neighboringblocks of the current block and construct an mvp candidate list using the derived mvpcandidates. In addition, the image encoding apparatus may use the motion vector of the mvpcandidate selected from among the mvp candidates included in the mvp candidate list as themvp of the current block. In this case, for example, a motion vector indicating a reference blockderived by the above-described motion estimation may be used as the motion vector of thecurrent block, and an mvp candidate having a smallest difference from the motion vector of thecurrent block among the mvp candidates may become the selected mvp candidate. A motionvector difference (MVD) which is obtained by subtracting the mvp from the motion vector ofthe current block may be derived. In this case, index information specifying the selected mvpcandidate and information on the MVD may be signaled to the image decoding apparatus.

[205] The image encoding apparatus may derive residual samples based on the predictionsamples (S1420). The image encoding apparatus may derive the residual samples throughcomparison between the original samples of the current block and the prediction samples. Forexample, the residual sample may be derived by subtracting the corresponding predictionsample from the original sample.

[206] The image encoding apparatus may encode image information including predictioninformation and residual information (S1430). The image encoding apparatus may output theencoded image information in the form of a bitstream. The prediction information may includeprediction mode information (e.g., skip flag, merge flag or mode index) and information on amotion vector as information related to the prediction procedure. Among the prediction modeinformation, the skip flag specifies whether to apply the skip mode to the current block and themerge flag specifies whether to apply the merge mode to the current block. Alternatively, theprediction mode information may specify one of a plurality of prediction modes, such as amode index. When the skip flag and the merge flag are 0, it may be determined that the MVPmode applies to the current block. The information on the motion vector may include candidateselection information (e.g., merge index, mvp flag or mvp index) which is information forderiving the motion vector. Among the candidate selection information, the merge index maybe signaled when applying the merge mode to the current block and may be information forselecting one of the merge candidates included in the merge candidate list. Among thecandidate selection information, the mvp flag or mvp index may be signaled when applyingthe MVP mode to the current block and may be information for selecting one of the mvpcandidates included in the mvp candidate list. In addition, the information on the motion vectormay include information on the above-described MVD. In addition, the information on themotion vector may include information specifying whether to apply L0 prediction, L1prediction or bi prediction. The residual information is information on the residual samples.The residual information may include information on the quantized transform coefficients forthe residual samples.

[207] The output bitstream may be stored in a (digital) storage medium and transmitted tothe image decoding apparatus or may be transmitted to the image decoding apparatus througha network.

[208] Meanwhile, as described above, the image encoding apparatus may generate areconstructed picture (picture including reconstructed samples and a reconstructed block)based on the reference samples and the residual samples. This is for the image encodingapparatus to derive the same prediction result as that performed by the image decodingapparatus, thereby increasing coding efficiency. Accordingly, the image encoding apparatusmay store the reconstructed picture (or reconstructed samples and reconstructed block) in amemory and use the same as a reference picture for inter prediction. An in-loop filteringprocedure is further applicable to the reconstructed picture, as described above.

[209] FIG. 19 is a flowchart illustrating an IBC based video / image decoding method.

[210] FIG. 20 is a view illustrating a configuration of a prediction unit for performing anIBC based video / image decoding method according to the present disclosure.

[211] The image decoding apparatus may perform operation corresponding to operationperformed by the image encoding apparatus. The image decoding apparatus may perform IBCprediction for a current block based on received prediction information to derive predictionsamples.

[212] The decoding method of FIG. 19 may be performed by the image decoding apparatusof FIG. 3. Steps S1610 to S1630 may be performed by the prediction unit and the predictioninformation of step S1610 and the residual information of step S1640 may be obtained from abitstream by the entropy decoder 210. The residual processor of the image decoding apparatusmay derive residual samples for the current block based on the residual information (S1640).Specifically, the dequantizer 220 of the residual processor may perform dequantization basedon the quantized transform coefficient derived based on the residual information to derivetransform coefficients, and the inverse transformer 230 of the residual processor may performinverse transform on the transform coefficients to derive residual samples for the current block .Step S1650 may be performed by the adder 235 or the reconstructor.

[213] Specifically, the image decoding apparatus may determine the prediction mode of thecurrent block based on the received prediction information (S1610). The image decodingapparatus may determine which prediction mode applies to the current block based on theprediction mode information in the prediction information.

[214] For example, it may be determined whether to apply the skip mode to the current blockbased on the skip flag. In addition, it may be determined whether to apply the merge node orMVP mode to the current block based on the merge flag. Alternatively, one of variousprediction mode candidates may be selected based on the mode index. The prediction modecandidates may include a skip mode, a merge mode and / or an MVP mode or may include theabove-described various inter prediction modes.

[215] The image encoding apparatus may derive the motion vector of the current block basedon the determined prediction mode (S1620). For example, when applying the skip mode or themerge mode to the current block, the image decoding apparatus may construct the abovedescribed merge candidate list and select one of the merge modes included in the mergecandidate list. The selection may be performed based on the above-described candidateselection information (merge index). The motion vector of the current block may be derivedusing the motion vector of the selected merge candidate. For example, the motion vector of theselected merge candidate may be used as the motion vector of the current block.

[216] As another example, when applying the MVP mode to the current block, the imagedecoding apparatus may construct an mvp candidate list and use the motion vector of the mvpcandidate selected from among the mvp candidates included in the mvp candidate list as themvp of the current block. The selection may be performed based on the above-describedcandidate selection information (mvp flag or mvp index). In this case, the MVD of the currentblock may be derived based on information on the MVD, and the motion vector of the currentblock may be derived based on the mvp and MVD of the current block.

[217] The image decoding apparatus may generate prediction samples of the current blockbased on the motion vector of the current block (S1630). The prediction samples of the currentblock may be derived using the samples of the reference block indicated by the motion vectorof the current block on the current picture. In some cases, a prediction sample filteringprocedure for all or some of the prediction samples of the current block may be furtherperformed.

[218] For example, as shown in FIG. 20, the prediction unit of the image decoding apparatusfor performing an IBC based video / image decoding method may include a prediction modedetermination unit, a motion vector derivation unit and a prediction sample derivation unit. Thecurrent block based on the received prediction mode information in the prediction modedetermination unit, derive the motion vector of the current block based on the receivedinformation on the motion vector in the motion vector derivation unit , and derive the predictionsamples of the current block in the prediction sample derivation unit.

[219] The image decoding apparatus may generate residual samples of the current blockbased on the received residual information (S1640). The image decoding apparatus maygenerate reconstructed samples for the current block based on the prediction samples and theresidual samples, and generate a reconstructed picture based on this (S1650). Thereafter, an inloop filtering procedure is further applicable to the reconstructed picture, as described above.

[220] As described above, one unit (e.g., a coding unit (CU)) may include a luma block (lumacoding block (CB)) and a chroma block (chroma CB). In this case, the luma block and thechroma block corresponding thereto may have the same motion information (e.g., motionvector) or different motion information. For example, the motion information of the chromablock may be derived based on the motion information of the luma block, such that the lumablock and the chroma block corresponding thereto have the same motion information.

[221] Overview of chroma format

[222] Hereinafter, a chroma format will be described. An image may be encoded intoencoded data including a luma component (e.g., Y) array and two chroma component (e.g., Cband Cr) arrays. For example, one pixel of the encoded image may include a luma sample and achroma sample. A chroma format may be used to represent a configuration format of the lumasample and the chroma sample, and the chroma format may be referred to as a color format.

[223] In an embodiment, an image may be encoded into various chroma formats such asmonochrome, 4:2:0, 4:2:2 or 4:4:4. In monochrome sampling, there may be one sample arrayand the sample array may be a luma array. In 4:2:0 sampling , there may be one luma samplearray and two chroma sample arrays, each of the two chroma arrays may have a height equalto half that of the luma array and a width equal to half that of the luma array. In 4:2:2 sampling ,there may be one luma sample array and two chroma sample arrays, each of the two chromaarrays may have a height equal to that of the luma array and a width equal to half that of theluma array. In 4:4:4 sampling, there may be one luma sample array and two chroma samplearrays, and each of the two chroma arrays may have a height and width equal to those of theluma array.

[224] For example, in 4:2:0 sampling, a chroma sample may be located below a luma samplecorresponding thereto. In 4:2:2 sampling, a chroma sample may be located to overlap a lumasample corresponding thereto. In 4:4:4 sampling, both a luma sample and a chroma samplemay be located at an overlapping position.

[225] A chroma format used in an encoding apparatus and a decoding apparatus may bepredetermined. Alternatively, a chroma format may be signaled from an encoding apparatus toa decoding apparatus to be adaptively used in the encoding apparatus and the decodingapparatus. In an embodiment, the chroma format may be signaled based on at least one ofchroma_format_idc or separate_colour_plane_flag. At least one of chroma_format_idc orseparate_colour_plane_flag may be signaled through higher level syntax such as DPS, VPS,SPS or PPS. For example, chroma_format_idc and separate_colour_plane_flag may beincluded in SPS syntax shown in FIG. 21.

[226] Meanwhile, FIG. 22 shows an embodiment of chroma format classification usingsignaling of chroma_format_idc and separate_colour_plane_flag. chroma_format_idc may beinformation specifying a chroma format applying to an encoded image.separate_colour_plane_flag may specify whether a color array is separately processed in aspecific chroma format. For example, a first value (e.g., 0) of chroma_format_idc may specifymonochrome sampling. A second value (e.g., 1) of chroma_format_idc may specify 4:2:0sampling. A third value (e.g., 2) of chroma_format_idc may specify 4:2:2 sampling. A fourthvalue (e.g., 3) of chroma_format_idc may specify 4:4:4 sampling.

[227] In 4:4:4, the following may apply based on the value of separate_colour_plane_flag.If the value of separate_colour_plane_flag is a first value (e.g., 0), each of two chroma arraysmay have the same height and width as a luma array. In this case, a value of ChromaArrayTypespecifying a type of a chroma sample array may be set equal to chroma_format_idc. If the valueof separate_colour_plane_flag is a second value (e.g., 1), luma, Cb and Cr sample arrays maybe separately processed and processed along with monochrome-sampled pictures. In this case,ChromaArrayType may be set to 0.

[228] Intra prediction on chroma block

[229] When intra prediction is performed on a current block, prediction on a luma componentblock (luma block) of the current block and prediction on a chroma component block (chromablock) may be performed. In this case, the intra prediction mode for the chroma block may beset separately from the intra prediction mode for the luma block.

[230] For example, the intra prediction mode for the chroma block may be specified basedon intra chroma prediction mode information, and the intra chroma prediction modeinformation may be signaled in the form of an intra_chroma_pred_mode syntax element. Forexample, the intra chroma prediction mode information may represent one of a planar mode, aDC mode, a vertical mode, a horizontal mode, a derived mode (DM), and a cross-componentlinear model (CCLM) mode. Here, the planar mode may specify intra prediction mode #0, theDC mode may specify intra prediction mode #1, the vertical mode may specify intra predictionmode #26, and the horizontal mode may specify intra prediction mode #10. DM may also bereferred to as a direct mode. The CCLM may also be referred to as a linear model (LM).

[231] Meanwhile, the DM and the CCLM are dependent intra prediction modes forpredicting the chroma block using information on the luma block. The DM may represent amode in which the same intra prediction mode as the intra prediction mode for the lumacomponent applies as the intra prediction mode for the chroma component. In addition, theCCLM may represent an intra prediction mode using, as the prediction samples of the chromablock, samples derived by subsampling reconstructed samples of the luma block and thenapplying α and β which are CCLM parameters to subsampled samples in a process ofgenerating the prediction block for the chroma block.

[232] [Equation 2]

[234] where, predc(i,j) may denote the prediction sample of (i, j) coordinates of the currentchroma block in the current CU. recL'(i,j) may denote the reconstructed sample of (i, j)coordinates of the current luma block in the CU. For example, recL'(i,j) may denote the downsampled reconstructed sample of the current luma block. Linear model coefficients α and βmay be signaled or derived from neighboring samples.

[235] Virtual pipeline data unit

[236] Virtual pipeline data units (VPDUs) may be defined for pipeline processing within apicture. The VPDUs may be defined as non-overlapping units within one picture. In a hardwaredecoding apparatus, successive VPDUs may be simultaneously processed by multiple pipelinestages. In most pipeline stages, a VPDU size may be roughly proportional to a buffer size.Accordingly, keeping the VPDU size small is important when considering the buffer size froma point of view of hardware. In most hardware decoding apparatuses, the VPDU size may beset equal to a maximum transform block (TB) size. For example, the VPDU size may be 64x64(64x64 luma samples) size. Alternatively, in VVC, the VPDU size may be changed (increasedor decreased) in consideration of the above-described ternary tree (TT) and / or binary tree (BT)partition.

[237] Meanwhile, to keep the VPDU size at the sample of 64x64 luma samples, splitting ofa CU shown in FIG. 23 may be restricted. More specifically, at least one of the followingrestrictions may be applied.

[238] Restriction 1 : Ternary tree (TT) splitting is not allowed for a CU having a width orheight of 128 or a width and height of 128.

[239] Restriction 2 : Horizontal binary tree (BT) splitting is not allowed for a CU having128xN (where, N is an integer equal to or less than 64 and greater than 0) (e.g., horizontalbinary tree splitting is not allowed for a CU having a width of 128 and a height less than 128).

[240] Restriction 3 : Vertical binary tree (BT) splitting is not allowed for a CU having Nx128(where, N is an integer equal to or less than 64 and greater than 0) (e.g., vertical binary treesplitting is not allowed for a CU having a height of 128 and a width less than 128).

[241] Maximum size limitation problem of chroma block for pipeline processing

[242] As described above with respect to a partitioning structure and a transform process, aCU may be split to generate a plurality of TUs. When a size of the CU is greater than amaximum TU size, the CU may be split into a plurality of TUs. Therefore, transform and / orinverse transform may be performed on each TU. In general, a maximum TU size for a lumablock may be set to a maximum available transform size which may be performed by anencoding apparatus and / or a decoding apparatus. Examples of splitting a CU and a TUaccording to an embodiment are shown in FIGS. 24 to 26.

[243] FIG. 24 shows an example of a TU generated by splitting a luma CU and a chroma CUaccording to an embodiment. In an embodiment, a maximum size of the luma CU may be64x64, a maximum available transform size may be 32x32, and a non-square TU may not beallowed. Therefore, a maximum size of a luma component transform block may be 32x32. Inthis embodiment, the maximum TU size may be set as shown in the following equation.

[244] [Equation 3]

[245] maxTbSize = ( cIdx = = 0 ) ? MaxTbSizeY : MaxTbSizeY / max (SubWidthC,SubHeightC)

[246] In the above equation, maxTbSize may be a maximum size of a transform block (TB),and cIdx may be a color component of a corresponding block. cIdx 0 may denote a lumacomponent, 1 may denote a Cb chroma component, and 2 may denote a Cr chroma component.MaxTbSizeY may denote a maximum size of a luma component transform block, SubWidthCmay denote a ratio of the width of the luma block to the width of the chroma block, SubHeightCmay denote a ratio of the height of the luma block to the height of the chroma block, and max(A,B) may denote a function for returning the larger value of A and B as a result value.

[247] According to the above equation, in the above-described embodiment, in the case ofthe luma block, a maximum size of a transform block may be set as a maximum size of a lumacomponent transform block. Here, the maximum size of the luma component transform blockis a value set during encoding and may be signaled from the encoding apparatus to the decodingapparatus through a bitstream.

[248] In addition, in the above embodiment, a maximum size of a transform block of thechroma block may be set to a value obtained by dividing the maximum size of the lumacomponent transform block by the larger value of SubWidthC and SubHeightC. Here,SubWidthC and SubHeightC may be determined based on chroma_format_idc andseparate_colour_plane_flag signaled from the encoding apparatus to the decoding apparatusthrough a bitstream as shown in FIG. 23.

[249] According to the above Equation, in the above embodiment, the maximum size of thetransform block may be determined to be any one of a minimum width and a minimum heightof the transform block. Therefore, TU splitting of the luma block and the chroma block in theabove embodiment may be performed as shown in FIG. 24. For example, as shown in FIG. 24,in the case of a chroma block having 4:2:2 format, as the maximum size of the transform blockis determined to be 16, a chroma CU may be split into a plurality of transform blocks in theform different from the form of splitting of a luma CU into transform blocks.

[250] FIG. 25 shows an example of a TU generated by splitting a luma CU and a chroma CUaccording to another embodiment. In an embodiment, a maximum size of the luma CU may be128x128, a maximum available transform size may be 64x64, and a non-square TU may notbe allowed. Therefore, the maximum size of a luma component transform block may be 64x64.In this embodiment, the maximum size of the transform block may be set as shown in thefollowing equation. In the following equation, min(A, B) may be a function for returning thesmaller value of A and B.

[251] [Equation 4]

[252] maxTbSize = ( cIdx = = 0 ) ? MaxTbSizeY : MaxTbSizeY / min (SubWidthC,SubHeightC)

[253] Meanwhile, according to the above equation, as the larger value of the width and heightof the corresponding block applies as the maximum size of the transform block, the luma CUand the chroma CU may be split into a plurality of TUs as in the example of FIG. 25.

[254] In the examples of FIGS. 24 and 25, in splitting a chroma CU having a 4:2:2 formatinto TUs, it is split in the form different from the form of splitting the corresponding luma CUinto TUs. However, when encoding / decoding of the chroma block is performed with referenceto the luma block as in a DM mode or CCLM mode for prediction of the chroma blockdescribed above, performing encoding (or decoding) of the corresponding chroma blockimmediately after encoding (or decoding) of the luma block corresponding to the chroma blockis efficient in order to reduce delay in pipeline processing and to save a memory.

[255] However, in the example of FIG. 24, encoding of two chroma blocks 2421 and 2423shall be performed after encoding of one luma transform block 2411, which requires separateprocessing in a relationship with different color formats (4:4:4 or 4:2:0). In addition, in theexample of FIG. 25, encoding of one chroma transform block 2521 shall be performed afterencoding of two luma transform blocks 2511 and 2512. In this way, in the above-described TUsplitting method, when a 4:2:2 format is used, since a luma block and a chroma blockcorresponding thereto do not match, a separate process may be added to perform pipelineprocessing or pipeline processing may not be performed.

[256] Maximum size limitation of chroma transform block for pipeline processing

[257] Hereinafter, a method of setting a size of a maximum transform block for a chromaCU to satisfy a condition for performing the above-described VPDU will be described.

[258] FIG. 26 shows an example of a TU generated by splitting a luma CU and a chroma CUaccording to another embodiment. In an embodiment, a maximum size of a luma CU may be128x128, a maximum available transform size may be 64x64, and splitting into non-square TUmay be allowed. Therefore, a maximum size of a luma component transform block may be64x64.

[259] As shown in FIG. 26, for splitting into non-square TUs, the maximum size of thetransform block may be defined for the width and the height. For example, the maximum width(maxTbWidth) of the transform block and the maximum height (maxTbHeight) of thetransform block may be defined as shown in the following equations, thereby defining themaximum size of the transform block.

[260] [Equation 5]

[261] maxTbWidth = ( cIdx = = 0 ) ? MaxTbSizeY : MaxTbSizeY / SubWidthC

[262] [Equation 6]

[263] maxTbHeight = ( cIdx = = 0 ) ? MaxTbSizeY : MaxTbSizeY / SubHeightC

[264] As in the above embodiment, by defining the maximum size of the transform block asthe width and the height, as shown in the example of FIG. 26, even in the case of a chroma CUhaving a 4:2:2 format, as in the form of splitting of the corresponding luma CU into TUs, thechroma CU may be split into TUs. Therefore, by splitting the chroma CU into TUs tocorrespond to the TUs of the luma CU, immediately after encoding (or decoding) of the lumablock, encoding (or decoding) of the chroma block corresponding thereto may be performed,thereby reducing delay in pipeline processing.

[265] Maximum size limitation of chroma transform block in IBC prediction mode andinter prediction mode

[266] Hereinafter, performing of an intra block copy (IBC) prediction mode and an interprediction mode, to which maximum size limitation of a chroma transform block for the abovedescribed chroma pipeline processing applies, will be described. An encoding apparatus and adecoding apparatus may perform IBC prediction and inter prediction by limiting the maximumsize of the chroma transform block according to the following description and operationsthereof may correspond to each other. In addition, the following description of IBC predictionmay apply to the inter prediction mode with change. Therefore, hereinafter, IBC predictionoperation of a decoding apparatus according to an embodiment will be described.

[267] The decoding apparatus according to an embodiment may generate a luma predictionblock predSamplesL having a size of (cbWidth)x(cbHeight) and chroma prediction blockspredSamplesCb and predSamplesCr having a size of(cbWidth / SubWidthC)x(cbHeight / SubHeightC), by performing IBC prediction. Here,cbWidth may be the width of a current CU and cbHeight may be the height of the current CU.

[268] In addition, the decoding apparatus may generate a luma residual block resSamplesLhaving a size of (cbWidth)x(cbHeight) and chroma residual blocks resSamplesCr andresSamplesCb having a size of (cbWidth / SubWidthC)x(cbHeight / SubHeightC). Finally, thedecoding apparatus may generate a reconstructed block using the prediction blocks and theresidual blocks.

[269] Hereinafter, a method of limiting a maximum size of a chroma transform block togenerate a residual block of a CU encoded in an IBC prediction mode by a decoding apparatusaccording to an embodiment will be described. The decoding apparatus may generate areconstructed block using the residual block generated in this step.

[270] The decoding apparatus according to an embodiment may directly obtain the followinginformation from a bitstream or derive the following information from other informationobtained from the bitstream to generate a residual block having a size of (nTbW)x(nTbH) of aCU encoded in an IBC prediction mode. Here, nTbW and nTbH may be set to the widthcbWidth of the current CU and the height cbHeight of the current CU.

[271] - Sample position (xTb0, yTb0) specifying the position of the top-left sample of acurrent transform block relative to the position of the top-left sample of a current picture

[272] - Parameter nTbW specifying the width of the current transform block

[273] - Parameter nTbH specifying the height of the current transform block

[274] - Parameter cIdx specifying a color component of the current CU

[275] The decoding apparatus may derive a maximum width maxTbWidth of a transformblock and a maximum height maxTbHeight of the transform block from the receivedinformation as follows.

[276] [Equation 7]

[277] maxTbWidth = ( cIdx = = 0 ) ? MaxTbSizeY : MaxTbSizeY / SubWidthC

[278] [Equation 8]

[279] maxTbHeight = ( cIdx = = 0 ) ? MaxTbSizeY : MaxTbSizeY / SubHeightC

[280] Furthermore, the decoding apparatus may derive the top-left sample position ( xTbY,yTbY ) of the current transform block based on whether the current CU is a luma componentor a chroma component as follows.

[281] [Equation 9]

[282] ( xTbY, yTbY ) = ( cIdx = = 0 ) ? ( xTb0, yTb0 ) : ( xTb0 * SubWidthC, yTb0 *SubHeightC )

[283] As in the above equation, when a current transform block is a chroma block, in orderto reflect the size of a chroma block determined according to the chroma format of the currenttransform block, the maximum width and height of the transform block and the top-left sampleposition of the current transform block may be determined based on the chroma format.

[284] Hereinafter, the decoding apparatus may generate a residual block by performing thefollowing procedure. This will be described with reference to FIG. 27. First, the decodingapparatus may determine whether the current transform block is split (S2710). For example,the decoding apparatus may determine whether the current transform block is split based onwhether the width and height of the current transform block are greater than the width andheight of the maximum transform block. For example, when nTbW is greater thanmaxTbWidth or nTbH is greater than maxTbHeight, the decoding apparatus may determinethat low-layer transform blocks are generated by splitting the current transform block.

[285] When the current transform block is split into lower-layer transform blocks, asdescribed above, the decoding apparatus may derive the width newTbW of the lower-layertransform block and the height newTbH of the lower-layer transform block as shown in thefollowing equation (S2720).

[286] [Equation 10]

[287] newTbW = ( nTbW > maxTbWidth ) ? ( nTbW / 2 ) : nTbW

[288] [Equation 11]

[289] newTbH = ( nTbH > maxTbHeight) ? ( nTbH / 2 ) : nTbH

[290] Next, the decoding apparatus may generate a residual block using the lower-layertransform blocks split from the current transform block (S2730). In an embodiment, as shownin FIG. 26, the current transform block may be a transform block having the width and heightof a chroma CU having a 4:2:2 format, and the lower-layer transform blocks may be a firstlower-layer transform block 2621 to a fourth lower-layer block 2624 obtained by splitting thecurrent transform block in four in the non-square form.

[291] First, the decoding apparatus may generate a residual block for the first lower-layertransform block. Referring to FIG. 26, the first lower-layer transform block 2621 may bespecified by the sample position (xTb0, yTb0), the width newTbW of the lower-layer transformblock and the height newTbH of the lower-layer transform block. The decoding apparatus maygenerate the residual block of the first lower-layer transform block 2621 using a colorcomponent cIdx of a current CU. Based on this, the decoding apparatus may generate amodified reconstructed picture. Thereafter, in-loop filtering may be performed for the modifiedreconstructed picture.

[292] Next, when nTbW is greater than maxTbWidth, the decoding apparatus may generatea residual block for the second lower-layer transform block. The second lower-layer transformblock 2622 may be specified by a sample position (xTb0 + newTbW, yTb0), the widthnewTbW of the lower-layer transform block and the height newTbH of the lower-layertransform block. The decoding apparatus may generate the residual block of the second lowerlayer transform block 2622 using the color component cIdx of the current CU. Based on this,the decoding apparatus may generate a modified reconstructed picture. Thereafter, in-loopfiltering may be performed for the modified reconstructed picture.

[293] Next, when nTbH is greater than maxTbHeight, the decoding apparatus may generatea residual block for the third lower-layer transform block. The third lower-layer transformblock 2623 may be specified by a sample position (xTb0, yTb0 + newTbH), the width newTbWof the lower-layer transform block and the height newTbH of the lower-layer transform block.Similarly to the above description, the decoding apparatus may generate the residual blockusing the color component cIdx of the current CU.

[294] Next, when nTbW is greater than maxTbWidth and nTbH is greater than maxTbHeight,the decoding apparatus may generate a residual block for the fourth lower-layer transformblock. The fourth lower-layer transform block 2624 may be specified by a sample position(xTb0 + newTbW, yTb0 + newTbH), the width newTbW of the lower-layer transform blockand the height newTbH of the lower-layer transform block. Similarly to the above description,the decoding apparatus may generate the residual block using the color component cIdx of thecurrent CU.

[295] Meanwhile, when splitting of the current transform block is not performed, thedecoding apparatus may perform IBC prediction as follows. For example, when nTbW is lessthan maxTbWidth and nTbH is less than maxTbHeight, the current transform block may notbe split. In this case, the decoding apparatus may generate a residual block for an IBCprediction mode, by performing a scaling and transform process using a sample position (xTbY,xTbY), a color component cIdx of the current CU, a transform block width nTbW, and atransform block height nTbH as input. Based on this, the decoding apparatus may generate amodified reconstructed picture. Thereafter, in-loop filtering may be performed for the modifiedreconstructed picture.

[296]

[297] Maximum size limitation of chroma transform block in intra prediction mode

[298] Hereinafter, performing of an intra prediction mode, to which maximum size limitationof a chroma transform block for the above-described chroma pipeline processing applies, willbe described. An encoding apparatus and a decoding apparatus may perform intra predictionby limiting the maximum size of the chroma transform block according to the followingdescription and operations thereof may correspond to each other. Therefore, hereinafter,operation of the decoding apparatus will be described.

[299] The decoding apparatus according to an embodiment may generate a reconstructedpicture, by performing intra prediction. In-loop filtering may be performed on the reconstructedpicture. The decoding apparatus according to an embodiment may directly obtain the followinginformation from a bitstream or derive the following information from other informationobtained from the bitstream to perform intra prediction.

[300] - Sample position (xTb0, yTb0) specifying the position of the top-left sample of acurrent transform block relative to the position of the top-left sample of a current picture

[301] - Parameter nTbW specifying the width of the current transform block

[302] - Parameter nTbH specifying the height of the current transform block

[303] - Parameter predModeIntra specifying the intra prediction mode of the current CU

[304] - Parameter cIdx specifying a color component of the current CU

[305] The decoding apparatus may derive a maximum width maxTbWidth of a transformblock and a maximum height maxTbHeight of the transform block from the receivedinformation as follows.

[306] [Equation 12]

[307] maxTbWidth = ( cIdx = = 0 ) ? MaxTbSizeY : MaxTbSizeY / SubWidthC

[308] [Equation 13]

[309] maxTbHeight = ( cIdx = = 0 ) ? MaxTbSizeY : MaxTbSizeY / SubHeightC

[310] Furthermore, the decoding apparatus may derive the top-left sample position ( xTbY,yTbY ) of the current transform block based on whether the current CU is a luma componentor a chroma component as follows.

[311] [Equation 14]

[312] ( xTbY, yTbY ) = ( cIdx = = 0 ) ? ( xTb0, yTb0 ) : ( xTb0 * SubWidthC, yTb0 *SubHeightC )

[313] Hereinafter, the decoding apparatus may perform intra prediction by performing thefollowing procedure. This will be described with reference to FIG. 28. First, the decodingapparatus may determine whether the current transform block is split (S2810). For example,the decoding apparatus may determine whether the current transform block is split based onwhether the width and height of the current transform block are greater than the width andheight of the maximum transform block. In addition, the decoding apparatus may determinewhether splitting is performed by further considering whether intra sub-partition (ISP) appliesto the current CU. For example, when nTbW is greater than maxTbWidth or nTbH is greaterthan maxTbHeight, the decoding apparatus may determine that intra prediction is performedby splitting the current transform block. In addition, even in this case, the decoding apparatusmay determine that intra prediction is performed by splitting the current transform block, onlywhen ISP does not apply to the current CU (e.g., the value of IntraSubpartitonSplitType isNO_ISP_SPLIT, that is, ISP does not apply to the current CU).

[314] When the current transform block is split into lower-layer transform blocks, thedecoding apparatus may derive the width newTbW of the lower-layer transform block and theheight newTbH of the lower-layer transform block as shown in the following equation (S2820).

[315] [Equation 15]

[316] newTbW = ( nTbW > maxTbWidth ) ? ( nTbW / 2 ) : nTbW

[317] [Equation 16]

[318] newTbH = ( nTbH > maxTbHeight) ? ( nTbH / 2 ) : nTbH

[319] It will be described with reference to FIG. 26. In an embodiment, the width nTbW ofthe current transform block may be the width of the chroma CU, and the height nTbH of thecurrent transform block may be the height of the chroma CU. In this embodiment, the widthnewTbW of the lower-layer transform block and the height newTbH of the lower-layertransform block may be determined to be the width and height of the transform block 2621split from the chroma CU. That is, in this embodiment, the current transform block may be atransform block having the width and height of the chroma CU having a 4:2:2 format, and thelower-layer transform blocks may be a first lower-layer transform block 2621 to a fourth lowerlayer block 2624 obtained by splitting the current transform block in four in the non-squareform.

[320] Next, the decoding apparatus may perform intra prediction using the lower-layertransform blocks split from the current transform block (S2830). First, the decoding apparatusmay perform intra prediction on the first lower-layer transform block. Referring to FIG. 26, thefirst lower-layer transform block 2621 may be specified by the sample position (xTb0, yTb0),the width newTbW of the lower-layer transform block and the height newTbH of the lowerlayer transform block. The decoding apparatus may perform intra prediction of the first lowerlayer transform block 2621 using the intra prediction mode predModeIntra of the current CUand the color component cIdx of the CU. Therefore, a modified reconstructed picture of thefirst lower-layer transform block 2621 may be generated.

[321] For example, the decoding apparatus may generate a prediction sample matrixpredSamples having a size of (newTbW)x(newTbH), by performing an intra sample predictionprocess. For example, the decoding apparatus may perform an intra sample prediction processusing the sample position (xTb0, yTb0), the intra prediction mode predModeIntra, thetransform block width (nTbW) newTbW, the transform block height (nTbH) newTbH, thecoding block width (nCbW) nTbW and the coding block height (nCbH) nTbH and the value ofthe parameter cIdx.

[322] In addition, the decoding apparatus may generate a residual sample matrix reSampleshaving a size of (newTbW)x(newTbH), by performing a scaling and transform process. Forexample, the decoding apparatus may perform the scaling and transform process based on thesample position (xTb0, yTb0), the value of the parameter cIdx, the transform block width(nTbW) newTbW, and the transform block height (nTbH) newTbH.

[323] In addition, the decoding apparatus may generate a reconstructed picture, byperforming a picture reconstruction process on a color component. For example, the decodingapparatus may set a transform block position to (xTb0, yTb0), a transform block width (nTbW)to newTbW, set a transform block height (nTbH) to newTbH, use the value of the parametercIdx, use the prediction sample matrix predSamples having a size of (newTbW)x(newTbH),and use a residual sample matrix reSamples having a size of (newTbW)x(newTbH), therebyperforming a picture reconstruction process on the color component.

[324] Next, when nTbW is greater than maxTbWidth, the decoding apparatus may performintra prediction on the second lower-layer transform block. The second lower-layer transformblock 2622 may be specified by the sample position (xTb0 + newTbW, yTb0), the widthnewTbW of the lower-layer transform block and the height newTbH of the lower-layertransform block. The decoding apparatus may perform intra prediction of the second lowerlayer transform block 2622 using the intra prediction mode predModeIntra of the current CUand the color component cIdx of the current CU. Intra prediction of the second lower-layertransform block 2622 may be performed on the sample position thereof similarly to intraprediction of the first lower-layer transform block 2621. Therefore, a modified reconstructedpicture of the second lower-layer transform block 2622 may be generated.

[325] Next, when nTbH is greater than maxTbHeight, the decoding apparatus may performintra prediction on the third lower-layer transform block. The third lower-layer transform block2623 may be specified by the sample position (xTb0, yTb0 + newTbH), the width newTbW ofthe lower-layer transform block and the height newTbH of the lower-layer transform block.Similarly to the above description, the decoding apparatus may perform intra prediction usingthe intra prediction mode predModeIntra of the current CU and the color component cIdx ofthe current CU.

[326] Next, when nTbW is greater than maxTbWidth and nTbH is greater than maxTbHeight,the decoding apparatus may perform intra prediction on the fourth lower-layer transform block.The fourth lower-layer transform block 2624 may be specified by the sample position (xTb0 +newTbW, yTb0 + newTbH), the width newTbW of the lower-layer transform block and theheight newTbH of the lower-layer transform block. Similarly to the above description, thedecoding apparatus may perform intra prediction using the intra prediction modepredModeIntra of the current CU and the color component cIdx of the current CU.

[327] Meanwhile, the decoding apparatus may perform intra prediction as follows, whensplitting of the current transform block is not performed. For example, when nTbW is less thanmaxTbWidth and nTbH is less than maxTbHeight or ISP applies to the current CU (e.g., thevalue of IntraSubpartitonSplitType is not NO_ISP_SPLIT), the current transform block maynot be split.

[328] First, the decoding apparatus may derive parameters nW, nH, numPartsX andnumPartsY as shown in the following equations.

[329] [Equation 17]

[330] nW = IntraSubPartitionsSplitType = = ISP_VER_SPLIT ? nTbW / NumIntraSubPartitions : nTbW

[331] nH = IntraSubPartitionsSplitType = = ISP_HOR_SPLIT ? nTbH / NumIntraSubPartitions : nTbH

[332] numPartsX = IntraSubPartitionsSplitType = = ISP_VER_SPLIT ?NumIntraSubPartitions : 1

[333] numPartsY = IntraSubPartitionsSplitType = = ISP_HOR_SPLIT ?NumIntraSubPartitions : 1

[334] In the above equation, IntraSubPartitionsSplitType specifies an ISP splitting type ofthe current CU, and ISP_VER_SPLIT specifies vertical ISP splitting, and ISP_HOR_SPLITspecifies horizontal ISP splitting. NumIntraSubPartitions specifies the number of ISP subpartitions.

[335] Next, the decoding apparatus may generate a prediction sample matrix predSampleshaving a size of (nTbW)x(nTbH), by performing the intra sample prediction process. Forexample, the decoding apparatus may perform the intra sample prediction process using thesample position ( xTb0 + nW * xPartIdx, yTb0 + nH * yPartIdx ), the intra prediction modepredModeIntra, the transform block width (nTbW) nW, the transform block height (nTbH) nH,the coding block width (nCbW) nTbW and the coding block height (nCbH) nTbH and the valueof the parameter cIdx. Here, the value of the partition index xPartIdx may have a value from 0to numPartX-1, and yPartIdx may have a value from 0 to numPartsY-1.

[336] Next, the decoding apparatus may generate a residual sample matrix reSamples havinga size of (nTbW)x(nTbH), by performing a scaling and transform process. For example, thedecoding apparatus may perform the scaling and transform process based on the sampleposition ( xTbY + nW * xPartIdx, yTbY + nH * yPartIdx ), the value of the parameter cIdx,the transform block width (nTbW) nW, and the transform block height (nTbH) nH.

[337] Next, the decoding apparatus may generate a reconstructed picture, by performing apicture reconstruction process on a color component. For example, the decoding apparatus mayset a transform block position to ( xTb0 + nW * xPartIdx, yTb0 + nH * yPartIdx), set atransform block width (nTbW) to nW, set a transform block height (nTbH) to nH, use the valueof the preset cIdx and use the prediction sample matrix predSamples having a size of(nTbW)x(nTbH) and the residual sample matrix reSamples having a size of (nTbW)x(nTbH),thereby performing the picture reconstruction process on the color component.

[338] Encoding method

[339] Hereinafter, a method of performing encoding by an encoding apparatus according toan embodiment using the above-described method will be described with reference to FIG. 29.An encoding apparatus according to an embodiment may include a memory and at least oneprocessor, and the at least one processor may perform the following encoding method.

[340] First, the encoding apparatus may determine a current block by splitting an image(S2910). Next, the encoding apparatus may generate a prediction block of the current block byperforming IBC prediction on the current block (S2920). Next, the encoding apparatus maygenerate a residual block of the current block based on the prediction block (S2930). Next, theencoding apparatus may encode prediction mode information of the current block (S2940). Inthis case, the residual block may be encoded based on a size of a transform block of the currentblock, and the size of the transform block may be determined based on a color component ofthe current block.

[341] More specifically, when the color component of the current block is a chromacomponent, the size of the transform block may be determined based on a color format. Inaddition, a width of the transform block may be determined based on a maximum width of thetransform block, and the maximum width of the transform block may be determined based ona maximum size and color format of a transform block of a luma block corresponding to thecurrent block.

[342] In addition, a height of the transform block may be determined based on a maximumheight of the transform block, and the maximum height of the transform block may bedetermined based on a maximum size and color format of a transform block of a luma blockcorresponding to the current block.

[343] In addition, when the color format of the current block is a format specifying that thewidth and height of the chroma block is half the width of the corresponding luma block, themaximum width of the transform block may be determined to be half the maximum width ofthe transform block of the luma block corresponding to the current block, and the maximumheight of the transform block may be determined to be half the maximum height of thetransform block of the luma block corresponding to the current block.

[344] In addition, when the current block is a chroma block, a position of a top-left sampleof the transform block may be determined based on a position and color format of a top-leftsample of the luma block corresponding to the current block.

[345] In addition, when the current block is a chroma block and the transform block is splitinto a plurality of lower-layer transform blocks, the top-left position of the lower-layertransform blocks may be determined based on the maximum width of the transform block andthe maximum height of the transform block. In this case, the maximum width of the transformblock may be determined based on the maximum size and color format of the transform blockof the luma block corresponding to the current block, and the maximum height of the transformblock may be determined based on the maximum size and color format of the transform blockof the luma block corresponding to the current block.

[346] In addition, when the current block is a chroma block and the width of the transformblock is greater than the maximum width of the transform block, a plurality of lower-layertransform blocks may be generated by vertically splitting the current block. The maximumwidth of the transform block may be determined based on the maximum size and color formatof the transform block of the luma block corresponding to the current block. In this case, theplurality of lower-layer transform blocks may include a first lower-layer transform block anda second lower-layer transform block, and the width of the first lower-layer transform blockand the width of the second lower-layer transform block may be determined to be the maximumwidth of the transform block.

[347] In addition, when the current block is a chroma block and the height of the transformblock is greater than the maximum height of the transform block, a plurality of lower-layertransform blocks may be generated by horizontally splitting the current block, and themaximum height of the transform block may be determined based on the maximum height andcolor format of the transform block of the luma block corresponding to the current block. Inthis case, the plurality of lower-layer transform blocks may include a third lower-layertransform block and a fourth lower-layer transform block, and the height of the third lowerlayer transform block and the height of the fourth lower-layer transform block may bedetermined to be the maximum height of the transform block.

[348] Decoding method

[349] Hereinafter, a method of performing decoding by a decoding apparatus according toan embodiment using the above-described method will be described with reference to FIG. 30.A decoding apparatus according to an embodiment may include a memory and at least oneprocessor, and the at least one processor may perform the following decoding method.

[350] First, the decoding apparatus may determine a prediction mode of a current block(S3010). Next, when the prediction mode of the current block is an intra block copy (IBC)prediction mode, the decoding apparatus may generate a prediction block of the current blockbased on IBC prediction mode information (S3020). Next, the decoding apparatus maydetermine a size of a transform block of the current block based on a color component of thecurrent block (S3030). Next, the decoding apparatus may generate a residual block of thecurrent block based on the size of the transform block (S3040). Next, the decoding apparatusmay reconstruct the current block based on the prediction block and the residual block of thecurrent block (S3050).

[351] More specifically, when the color component of the current block is a chromacomponent, the size of the transform block may be determined based on the color format. Inaddition, the width of the transform block may be determined based on a maximum width ofthe transform block, and the maximum width of the transform block may be determined basedon a maximum size and color format of a transform block of a luma block corresponding to thecurrent block.

[352] In addition, the height of the transform block may be determined based on a maximumheight of the transform block, and the maximum height of the transform block may bedetermined based on the maximum size and color format of the transform block of the lumablock corresponding to the current block.

[353] In addition, when the color format of the current block is a format specifying that thewidth and height of the chroma block is half the width of the luma block corresponding thereto,the maximum width of the transform block may be determined to be half the maximum widthof the transform block of the luma block corresponding to the current block, and the maximumheight of the transform block may be determined to be half the maximum height of thetransform block of the luma block corresponding to the current block.

[354] In addition, when the current block is a chroma block, a position of a top-left sampleof the transform block may be determined based on a position and color format of a top-leftsample of the luma block corresponding to the current block.

[355] In addition, when the current block is a chroma block and the transform block is splitinto a plurality of lower-layer transform blocks, the top-left position of the lower-layertransform blocks may be determined based on the maximum width of the transform block andthe maximum height of the transform block. In this case, the maximum width of the transformblock may be determined based on the maximum size and color format of the transform blockof the luma block corresponding to the current block, and the maximum height of the transformblock may be determined based on the maximum size and color format of the transform blockof the luma block corresponding to the current block.

[356] In addition, when the current block is a chroma block and the width of the transformblock is greater than the maximum width of the transform block, a plurality of lower-layertransform blocks may be generated by vertically splitting the current block. The maximumwidth of the transform block may be determined based on the maximum size and color formatof the transform block of the luma block corresponding to the current block. In this case, theplurality of lower-layer transform blocks may include a first lower-layer transform block anda second lower-layer transform block, and the width of the first lower-layer transform blockand the width of the second lower-layer transform block may be determined to be the maximumwidth of the transform block.

[357] In addition, when the current block is a chroma block and the height of the transformblock is greater than the maximum height of the transform block, a plurality of lower-layertransform blocks may be generated by horizontally splitting the current block, and themaximum height of the transform block may be determined based on the maximum height andcolor format of the transform block of the luma block corresponding to the current block. Inthis case, the plurality of lower-layer transform blocks may include a third lower-layertransform block and a fourth lower-layer transform block, and the height of the third lowerlayer transform block and the height of the fourth lower-layer transform block may bedetermined to be the maximum height of the transform block.

[358] Application embodiment

[359] While the exemplary methods of the present disclosure described above arerepresented as a series of operations for clarity of description, it is not intended to limit theorder in which the steps are performed, and the steps may be performed simultaneously or indifferent order as necessary. In order to implement the method according to the presentdisclosure, the described steps may further include other steps, may include remaining stepsexcept for some of the steps, or may include other additional steps except for some steps.

[360] In the present disclosure, the image encoding apparatus or the image decodingapparatus that performs a predetermined operation (step) may perform an operation (step) ofconfirming an execution condition or situation of the corresponding operation (step). Forexample, if it is described that predetermined operation is performed when a predeterminedcondition is satisfied, the image encoding apparatus or the image decoding apparatus mayperform the predetermined operation after determining whether the predetermined condition issatisfied.

[361] The various embodiments of the present disclosure are not a list of all possiblecombinations and are intended to describe representative aspects of the present disclosure, andthe matters described in the various embodiments may be applied independently or incombination of two or more.

[362] Various embodiments of the present disclosure may be implemented in hardware,firmware, software, or a combination thereof. In the case of implementing the presentdisclosure by hardware, the present disclosure can be implemented with application specificintegrated circuits (ASICs), Digital signal processors (DSPs), digital signal processing devices(DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs),general processors, controllers, microcontrollers, microprocessors, etc.

[363] In addition, the image decoding apparatus and the image encoding apparatus, to whichthe embodiments of the present disclosure are applied, may be included in a multimediabroadcasting transmission and reception device, a mobile communication terminal, a homecinema video device, a digital cinema video device, a surveillance camera, a video chat device,a real time communication device such as video communication, a mobile streaming device, astorage medium, a camcorder, a video on demand (VoD) service providing device, an OTTvideo (over the top video) device, an Internet streaming service providing device, a threedimensional (3D) video device, a video telephony video device, a medical video device, andthe like, and may be used to process video signals or data signals. For example, the OTT videodevices may include a game console, a blu-ray player, an Internet access TV, a home theatersystem, a smartphone, a tablet PC, a digital video recorder (DVR), or the like.

[364] FIG. 31 is a view showing a contents streaming system, to which an embodiment ofthe present disclosure is applicable.

[365] As shown in FIG. 31, the contents streaming system, to which the embodiment of thepresent disclosure is applied, may largely include an encoding server, a streaming server, a webserver, a media storage, a user device, and a multimedia input device.

[366] The encoding server compresses contents input from multimedia input devices such asa smartphone, a camera, a camcorder, etc. into digital data to generate a bitstream and transmitsthe bitstream to the streaming server. As another example, when the multimedia input devicessuch as smartphones, cameras, camcorders, etc. directly generate a bitstream, the encodingserver may be omitted.

[367] The bitstream may be generated by an image encoding method or an image encodingapparatus, to which the embodiment of the present disclosure is applied, and the streamingserver may temporarily store the bitstream in the process of transmitting or receiving thebitstream.

[368] The streaming server transmits the multimedia data to the user device based on a user'srequest through the web server, and the web server serves as a medium for informing the userof a service. When the user requests a desired service from the web server, the web server maydeliver it to a streaming server, and the streaming server may transmit multimedia data to theuser. In this case, the contents streaming system may include a separate control server. In thiscase, the control server serves to control a command / response between devices in the contentsstreaming system.

[369] The streaming server may receive contents from a media storage and / or an encodingserver. For example, when the contents are received from the encoding server, the contentsmay be received in real time. In this case, in order to provide a smooth streaming service, thestreaming server may store the bitstream for a predetermined time.

[370] Examples of the user device may include a mobile phone, a smartphone, a laptopcomputer, a digital broadcasting terminal, a personal digital assistant (PDA), a portablemultimedia player (PMP), navigation, a slate PC, tablet PCs, ultrabooks, wearable devices (e.g.,smartwatches, smart glasses, head mounted displays), digital TVs, desktops computer, digitalsignage, and the like.

[371] Each server in the contents streaming system may be operated as a distributed server,in which case data received from each server may be distributed.

[372] The scope of the disclosure includes software or machine-executable commands (e.g.,an operating system, an application, firmware, a program, etc.) for enabling operationsaccording to the methods of various embodiments to be executed on an apparatus or a computer,a non-transitory computer-readable medium having such software or commands stored thereonand executable on the apparatus or the computer.Industrial Applicability

[373] The embodiments of the present disclosure may be used to encode or decode animage.

Claims

1. An image decoding apparatus (200) for image decoding, the image decoding apparatus comprising: a memory; and at least one processor connected to the memory, the at least one processor configured to: (a) determine a prediction mode of a current block; (b) generate a prediction block of the current block based on intra block copy (IBC) prediction mode information, based on the prediction mode of the current block being an IBC prediction mode; (c) determine a size of a transform block of the current block based on a color component of the current block; (d) generate a residual block of the current block based on the size of the transform block; and (e) reconstruct the current block based on the prediction block and the residual block of the current block, characterized in that the size of the transform block is determined based on a color format of the current block, based on the color component of the current block being a chroma component, the current block is a chroma block, and a position of a top-left sample of the transform block is determined based on a position of a top-left sample of a luma transform block corresponding to the current block and the color format of the current block.

2. The image decoding apparatus of claim 1, wherein a width of the transform block is determined based on a maximum width of the transform block, and wherein the maximum width of the transform block is determined based on a maximum size of the luma transform block and the color format of the current block.

3. The image decoding apparatus of claim 1, wherein a height of the transform block is determined based on a maximum height of the transform block, and wherein the maximum height of the transform block is determined based on a maximum size of the luma transform block and the color format of the current block.

4. The image decoding apparatus of claim 1, wherein, based on the color format of the current block being a format specifying that a width and height of a chroma block is half a width of a luma block corresponding thereto, a maximum width of the transform block is determined to be half a maximum width of the luma transform block, and a maximum height of the transform block is determined to be half a maximum height of the luma transform block.

5. The image decoding apparatus of claim 1, wherein, based on the transform block being split into a plurality of lower-layer transform blocks, a top-left position of the lower-layer transform blocks is determined based on a maximum width of the transform block and a maximum height of the transform block.

6. The image decoding apparatus of claim 5, wherein the maximum width of the transform block is determined based on a maximum size of the luma transform block and the color format of the current block, and wherein the maximum height of the transform block is determined based on a maximum size of the luma transform block and the color format of the current block.

7. The image decoding apparatus of claim 1, wherein a plurality of lower-layer transform blocks are generated by vertically splitting the current block, based on a width of the transform block being greater than a maximum width of the transform block, and wherein the maximum width of the transform block is determined based on a maximum size of the luma transform block and the color format of the current block.

8. The image decoding apparatus of claim 7, wherein the plurality of lower-layer transform blocks include a first lower-layer block and a second lower-layer transform block, and wherein a width of the first lower-layer transform block and a width of the second lower-layer transform block are determined to be the maximum width of the transform block.

9. The image decoding apparatus of claim 1, wherein a plurality of lower-layer transform blocks are generated by horizontally splitting the current block, based on a height of the transform block being greater than a maximum height of the transform block, and wherein the maximum height of the transform block is determined based on a maximum height of the luma transform block and the color format of the current block.

10. The image decoding apparatus of claim 9, wherein the plurality of lower-layer transform blocks include a third lower-layer block and a fourth lower-layer transform block, and wherein a height of the third lower-layer transform block and a height of the fourth lower-layer transform block are determined to be the maximum height of the transform block.

11. An image encoding apparatus (100) for image encoding, the image encoding apparatus comprising: a memory; and at least one processor connected to the memory, the at least one processor configured to: (a) determine a current block by splitting an image; (b) generate a prediction block of the current block by performing intra block copy (IBC) prediction on the current block; (c) generate a residual block of the current block based on the prediction block; and (d) encode prediction mode information of the current block, characterized in that the residual block is encoded based on a size of a transform block of the current block, the size of the transform block is determined based on a color component of the current block, the size of the transform block is determined based on a color format of the current block, based on the color component of the current block being a chroma component, the current block is a chroma block, and a position of a top-left sample of the transform block is determined based on a position of a top-left sample of a luma transform block corresponding to the current block and the color format of the current block.

12. An apparatus for transmitting data for an image, the apparatus comprising: at least one processor configured to obtain a bitstream generated by an image encoding method; and a transmitter configured to transmit the bitstream, the image encoding method comprising: (a) determining a current block by splitting an image; (b) generating a prediction block of the current block by performing intra block copy (IBC) prediction on the current block; (c) generating a residual block of the current block based on the prediction block; and (d) encoding prediction mode information of the current block, characterized in that the residual block is encoded based on a size of a transform block of the current block, the size of the transform block is determined based on a color component of the current block, the size of the transform block is determined based on a color format of the current block, based on the color component of the current block being a chroma component, the current block is a chroma block, and a position of a top-left sample of the transform block is determined based on a position of a top-left sample of a luma transform block corresponding to the current block and the color format of the current block.