RU2805056C2 - Method and device for encoding/decoding image signals - Google Patents

Abstract

FIELD: video data processing.

SUBSTANCE: encoding or decoding video signals. A method for decoding an image used in a video decoder is proposed, which includes the following steps: determining whether to split a coding block into a first prediction element and a second prediction element; determining the section type of the coding block when determining that the coding block is broken; determining an index of the second merger candidate based on a comparison between the value of the second index information and the value of the first index information; obtaining first motion information with respect to the first prediction element and second motion information with respect to the second prediction element in the coding block; and fetching a prediction sample in the coding block based on the first motion information and the second motion information.

EFFECT: increasing the efficiency of decoding and obtaining a candidate for merging using a table of information about the movement between sections when encoding or decoding video signals.

15 cl, 48 dwg, 10 tbl

Description

FIELD OF TECHNOLOGY TO WHICH THE INVENTION RELATES

[0001] The present invention relates to a method and apparatus for encoding or decoding video signals.

BACKGROUND OF THE INVENTION

[0002] With the trend toward larger display panels, demand for high-definition video services is growing. The biggest problem of high-definition video services is that the data volume increases significantly, and to solve this problem, research has been actively carried out to improve the video compression ratio. As a representative example, the Joint Video Coding Group (JCT-VC) within the Video Coding Experts Group (VCEG) within the Telecommunication Standardization Sector of the International Telecommunication Union (ITU-T) and the Moving Picture Experts Group (MPEG) was established in 2009. . JCT-VC proposed High Efficiency Video Coding (HEVC), which is a video compression standard whose compression efficiency is approximately twice that of H.264/AVC, and was approved as a standard on January 25, 2013. With the rapid development of high-definition video services, HEVC performance is gradually revealing its limitations.

SUMMARY OF THE INVENTION

TECHNICAL PROBLEM

[0003] The present invention is directed to providing a method for dividing a coding block into a plurality of prediction blocks when encoding or decoding video signals, and an apparatus for performing the method.

[0004] The present invention is directed to providing a method for obtaining motion information for each of a plurality of prediction blocks when encoding or decoding video signals and an apparatus for performing the method.

[0005] The present invention is directed to providing a method for obtaining a merge candidate using a table of inter-section motion information when encoding or decoding video signals and an apparatus for executing the method.

[0006] The technical problems to be solved by the present invention are not limited to the above problems, and other technical problems that are not described herein will be clearly understood by those skilled in the art from the following description.

TECHNICAL SOLUTION

[0007] A method for decoding or encoding a video signal according to the present invention includes determining whether an encoding block should be divided into a first prediction element and a second prediction element, determining a partition type of the encoding block when determining that the encoding block should be divided, determining an index of a second candidate merging based on the comparison between the value of the second index information and the value of the first index information, obtaining first motion information with respect to the first prediction element in the encoding block and second motion information with respect to the second prediction element in the encoding block, and sampling the prediction sample in the encoding block based on the first motion information and the second motion information. In this case, the first motion information with respect to the first prediction element may be obtained from the first merger candidate indicated by the first index information from the plurality of merger candidates included in the merger candidate list, and the second motion information with respect to the second the prediction element may be obtained from a second merge candidate indicated by the second index information from the plurality of merge candidates included in the merge candidate list.

[0008] When the value of the second index information is greater than or equal to the value of the first index information, the index of the second merge candidate is equal to the value of the second index information plus one.

[0009] When the value of the second index information is less than the value of the first index information, the index of the second merge candidate is equal to the value of the second index information.

[0010] When a prediction sample is included in the boundary portion between the first prediction element and the second prediction element, the prediction sample may be obtained based on a weighted sum operation of the first prediction sample obtained based on the first motion information and the second prediction sample obtained based on the second traffic information.

[0011] The first weight applied to the first prediction sample may be determined based on the x coordinate and y coordinate of the prediction sample.

[0012] A second weight applied to the second prediction sample may be obtained by subtracting the first weight from a constant value.

[0013] The size of the boundary portion may be determined based on at least one of a coding block size or a coding block shape.

[0014] The features summarized above with respect to the present invention are merely illustrative aspects of the detailed description of the present invention described below and do not limit the scope of the present invention.

POSITIVE EFFECTS

[0015] According to the present invention, by dividing a coding block into a plurality of prediction blocks and obtaining motion information for each of the prediction blocks, the performance of inter-picture prediction can be improved.

[0016] According to the present invention, by providing a method for obtaining a fusion candidate using an inter-section motion information table, the performance of prediction performed between images can be improved.

[0017] The beneficial effects of the present invention are not limited to the above effects, and other beneficial effects that are not described herein will be clearly understood by those skilled in the art from the following description.

BRIEF DESCRIPTION OF GRAPHIC MATERIALS

[0018] In FIG. 1 is a block diagram of an encoder according to an embodiment of the present invention.

[0019] In FIG. 2 is a block diagram of a decoder according to an embodiment of the present invention.

[0020] In FIG. 3 is a schematic diagram showing a basic element of a code tree according to an embodiment of the present invention.

[0021] In FIG. 4 is a schematic diagram showing various types of encoding block section.

[0022] In FIG. 5 is a schematic diagram illustrating an aspect of a code tree element section.

[0023] In FIG. 6 is a flowchart of an inter-image prediction method according to an embodiment of the present invention.

[0024] In FIG. 7 is a schematic diagram illustrating the nonlinear motion of an object.

[0025] In FIG. 8 is a flowchart of an inter-image prediction method based on affine motion according to an embodiment of the present invention.

[0026] In FIG. 9 is a schematic diagram illustrating the affine source vector for each affine motion model.

[0027] In FIG. 10 is a schematic diagram illustrating the affine vectors of subblocks under a 4-parameter motion model.

[0028] In FIG. 11 is a schematic diagram showing neighboring blocks that can be used to obtain a merge candidate.

[0029] In FIG. 12 is a schematic diagram showing an example of obtaining an affine source vector of a current block based on an affine source vector of an affine neighboring block.

[0030] In FIG. 13 is a schematic diagram showing an example of setting a motion vector of a sub-block as an affine source vector of an affine neighboring block.

[0031] In FIG. 14-16 are schematic diagrams showing the location of the reference count.

[0032] In FIG. 17 is a schematic diagram showing an example in which a method for obtaining a modified affine fusion vector is applied.

[0033] In FIG. 18 is a schematic diagram showing an example of obtaining an affine source vector of an affine merge candidate based on the motion vectors of subblocks belonging to a neighboring block.

[0034] In FIG. 19 is a schematic diagram showing an example of obtaining affine source vectors of an affine merge candidate based on the motion vectors of sub-blocks located to the left of the current block.

[0035] In FIG. 20 is a schematic diagram showing an example of obtaining affine source vectors of an affine merge candidate based on motion information of an adjacent block or a non-adjacent block located to the left of the current block.

[0036] In FIG. 21 is a schematic diagram showing the location of a block for obtaining an affine source vector of an affine merge candidate.

[0037] In FIG. 22 is a schematic diagram for describing an example of combining motion vectors of multiple neighboring blocks to obtain a combined fusion candidate.

[0038] In FIG. 23 is a schematic diagram showing an inaccessible neighbor block.

[0039] In FIG. 24 is a flowchart of the process of obtaining information about the movement of the current block in the merge mode.

[0040] In FIG. 25 is a schematic diagram for describing an aspect of updating an inter-section traffic information table.

[0041] In FIG. 26 is a schematic diagram showing an aspect of updating an inter-site merge candidate table.

[0042] In FIG. 27 is a schematic diagram showing an example in which the index of a previously stored inter-site merge candidate is updated.

[0043] In FIG. 28 is a schematic diagram showing the location of a representative subblock.

[0044] In FIG. 29 shows an example in which a table of inter-section motion information is generated for each inter-picture prediction mode.

[0045] In FIG. 30 is a schematic diagram showing an example in which a cross-section merging candidate included in the long-term traffic information table is added to the merging candidate list.

[0046] In FIG. 31 is a schematic diagram showing an example in which redundancy checking is performed on only some merge candidates.

[0047] In FIG. 32 is a schematic diagram showing an example in which redundancy control is omitted with respect to a particular merger candidate.

[0048] In FIG. 33 is a schematic diagram showing an example of dividing a coding block into a plurality of prediction blocks using a diagonal line.

[0049] In FIG. 34 is a schematic diagram showing an example of dividing a coding block into two prediction elements.

[0050] In FIG. 35 shows examples in which a coding block is divided into multiple prediction blocks of different sizes.

[0051] In FIG. 36 is a schematic diagram showing adjacent blocks used to obtain a triangular merge candidate.

[0052] In FIG. 37 is a schematic diagram for describing an example of determining the availability of a neighboring block for each triangular prediction element.

[0053] In FIG. 38 and 39 are schematic diagrams showing an example of obtaining a prediction sample based on a weighted sum operation of the first prediction sample and the second prediction sample.

[0054] In FIG. 40 is a flowchart of an intra prediction method according to an embodiment of the present invention.

[0055] In FIG. 41 is a schematic diagram showing intra prediction modes.

[0056] In FIG. 42 and 43 are schematic diagrams showing an example of a one-dimensional array in which reference samples are arranged in a line.

[0057] In FIG. 44 is a schematic view illustrating angles formed between a straight line parallel to the x-axis and directional intra prediction modes.

[0058] In FIG. 45 is a schematic view showing an aspect in which a prediction count is obtained when the current block has a non-square shape.

[0059] In FIG. 46 is a schematic diagram showing wide-angle intra prediction modes.

[0060] In FIG. 47 is a flowchart showing the process of determining the quality indicator of a block.

[0061] In FIG. 48 shows predefined candidate filters.

DETAILED DESCRIPTION

[0062] Embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0063] The image is encoded and decoded in block elements. As an example, encoding and decoding processing such as transform, quantization, prediction, in-loop filtering, or reconstruction may be performed on an encoding block, a transform block, or a prediction block.

[0064] Hereinafter, the block to be encoded or decoded will be referred to as the “current block”. As an example, the current block may indicate an encoding block, a transform block, or a prediction block depending on the current stage of encoding or decoding processing.

[0065] In addition, the term "element" as used herein may be understood to indicate a base element for performing a particular encoding and decoding process, and the term "block" may be understood to indicate an array of samples of a predetermined size. Unless otherwise noted, the terms "block" and "element" may be used interchangeably. As an example, in the following embodiments, a coding block and a coding element may be understood to have equivalent meanings.

[0066] In FIG. 1 is a block diagram of an encoder according to an embodiment of the present invention.

[0067] As shown in FIG. 1, the video encoding apparatus 100 may include an image splitter 110, predictors 120 and 125, an inverter 130, a quantizer 135, a reorderer 160, an entropy encoder 165, an inverse quantizer 140, an inverter 145, a filter 150, and a memory 155.

[0068] Elements in FIG. 1 are shown independently to represent various characteristic functions in a video encoding apparatus, and each of the elements is not intended to be implemented in a separate hardware unit or as a single software unit. In other words, for ease of description, the elements are arranged independently of each other. To perform functions, at least two elements may be combined into one element, or one element may be divided into multiple elements. In this case, the embodiment for combining elements and the embodiment for separating an element are within the scope of the present invention without deviating from the spirit of the present invention.

[0069] In addition, some of the elements may not be essential elements for performing essential functions in the present invention, but may simply be optional elements to enhance performance. The present invention can be implemented by including only the elements necessary to implement the essence of the present invention, and not elements used only to improve performance. Even a structure including only essential elements and not optional elements used only to improve performance is within the scope of the present invention.

[0070] The image splitter 110 may split the input image into at least one processing element. In this case, the processing element may be a prediction element (PU), a transformation element (TU), or a coding element (CU). The image partitioning apparatus 110 may partition one image into a plurality of combinations of encoding elements, prediction elements, and transform elements, and may select one combination of encoding elements, prediction elements, and transform elements according to a predetermined criterion (eg, a cost function) to encode the image.

[0071] For example, one image may be divided into multiple encoding elements. A recursive tree structure, such as a quaternary tree structure, can be used to partition the image into encoding elements. An encoding element split into other encoding elements using a single image or the largest encoding element as the root may have a number of child nodes corresponding to the number of encoding elements split. An encoding element that is no longer split due to a predefined constraint serves as a leaf node. That is, when it is assumed that only a square division is possible for one encoding element, one encoding element can be divided into at most four other encoding elements.

[0072] In further embodiments of the present invention, an encoding element may refer to an element configured to perform encoding or an element configured to perform decoding.

[0073] One coding element may be divided into at least one or more equally sized square or rectangular shaped prediction elements, and may be divided into prediction elements such that one of the prediction elements differs from another prediction element in shape and/or size .

[0074] When a prediction element subject to intra prediction is generated based on a coding element, and the coding element is not a minimum coding element, intra prediction can be performed without dividing the coding element into a plurality of N×N prediction elements.

[0075] Predictors 120 and 125 may include an inter-picture prediction predictor 120 configured to perform inter-picture prediction, and an internal predictor 125 configured to perform intra-prediction. Predictors 120 and 125 can determine whether to perform intra prediction or use inter-picture prediction with respect to prediction elements, and can determine detailed information (eg, intra prediction mode, motion vector, reference picture, etc.) corresponding each method of prediction. In this case, the processing element in which the prediction is performed may be different from the processing element in which the prediction method and specific details are determined. For example, prediction method, prediction mode, etc. may be determined by a prediction element, and the prediction may be performed by a transformation element. A residual value (residual block) between the generated prediction block and the original block may be input to the converter 130. Moreover, motion vector information, prediction mode information, etc., which are used for prediction, in addition to the residual value, may be encoded by entropy encoder 165 and transferred to the decoder. When a particular encoding mode is used, the original block can be encoded unchanged and passed to the decoder without generating a prediction block.

[0076] The inter-picture predictor 120 may predict a prediction element based on information about at least one of the previous image or the next image relative to the current image, and in some cases, the prediction element may be predicted based on information about a partial portion of the current image , in which the encoding is carried out. The inter-image predictor 120 may include a reference image interpolator, a motion predictor, and a motion compensator.

[0077] The reference image interpolator can obtain information about the reference image from the memory 155 and can generate information about pixels whose number is less than or equal to an integer number of pixels from the reference image. In the case of luminance pixels, a DCT-based 8-point interpolation filter having different filter coefficients can be used to generate information about pixels whose number is less than or equal to the integer number of pixels in 1/4-pixel elements. For chrominance signals, a DCT-based 4-point interpolation filter having different filter coefficients can be used to generate information about pixels whose number is less than or equal to the integer number of pixels in 1/8-pixel elements.

[0078] The motion predictor may perform motion prediction based on a reference image interpolated by a reference image interpolator. As a method for calculating the motion vector, various methods may be used, such as the Full Block Matching Algorithm (FBMA), the Three-Stage Search (TSS) algorithm, and the New Three-Stage Search (NTS) algorithm. The motion vector can have a motion vector value in 1/2 or 1/4 pixel elements based on the interpolated pixels. The motion prediction element may predict the current prediction element by using another motion prediction method. As the motion prediction method, various methods such as a skip method, a merging method, an advanced motion vector prediction (AMVP) method, and an intra-block copy method can be used.

[0079] The internal predictor 125 may generate a prediction element based on information about a reference pixel adjacent to the current block, which is information about a pixel in the current image. When the nearest block of the current prediction element is a block subject to inter-picture prediction, and thus the reference pixel is a pixel subject to inter-picture prediction, the reference pixel information of the nearest block subject to intra-prediction may be used instead a reference pixel included in a block subject to inter-image prediction. That is, when a reference pixel is not available, at least one available reference pixel can be used instead of the unavailable reference pixel information.

[0080] Intra prediction prediction modes may include a directional prediction mode that uses reference pixel information depending on the prediction direction, and an omnidirectional mode that does not use directionality information when making prediction. The mode for predicting luminance information and the mode for predicting chrominance information may be different, and information about the intra prediction mode used to predict luminance information or information about a predicted luminance signal may be used to predict chrominance information.

[0081] When intra prediction is performed and the size of the prediction element is equal to the size of the transform element, intra prediction may be performed on the prediction element based on pixels located to the left of the prediction element, pixels located in the upper left corner of the prediction element, and pixels located in the top of the prediction element. However, when intra prediction is performed and the prediction element is different in size from the transform element, intra prediction can be performed using a reference pixel based on the transform element. In addition, intra prediction using N×N partitioning only for the minimum coding element can be used.

[0082] In the intra prediction method, a prediction block may be generated after applying an adaptive intra smoothing (AIS) filter to a reference pixel depending on the prediction mode. The type of AIS filter applied to the reference pixel may vary. For the purpose of executing the intra prediction method, the intra prediction mode of the current prediction element can be predicted from the intra prediction mode of the prediction element adjacent to the current prediction element. When predicting the prediction mode of the current prediction element using mode information predicted from the nearest prediction element, information indicating that the current prediction element and the nearest prediction element have the same prediction mode can be transmitted using the predetermined flag information when the intra prediction mode the current prediction element is the same as the intra prediction mode of the nearest prediction element, and entropy encoding may be performed to encode prediction mode information of the current block when the prediction mode of the current prediction element is different from the prediction mode of the nearest prediction element.

[0083] In addition, a residual block containing information about a residual value that represents the difference between a prediction element being predicted and an original prediction element block may be generated based on the prediction elements generated by predictors 120 and 125. The generated residual block may be introduced into the converter 130.

[0084] Transformer 130 may transform a residual block containing residual value information between the original block and the prediction elements generated by predictors 120 and 125 by using a transformation method such as discrete cosine transform (DCT) or discrete sine transform (DST). In this case, the DCT transform base contains at least one of DCT2 or DCT8, and the DST transform base contains DST7. Whether to apply DCT or DST to transform the residual block can be determined based on information about the intra prediction mode of the prediction element used to generate the residual block. Residual block conversion can be skipped. A flag indicating whether conversion of the remainder block should be skipped may be encoded. Skip conversion may be allowed for residual blocks whose size is less than or equal to the threshold value of the luma component or the chrominance component in a 4:4:4 format.

[0085] Quantizer 135 may quantize values that are converted to the frequency domain by converter 130. Quantization factors may vary depending on the block or importance of the image. The values calculated by quantizer 135 may be passed to inverse quantizer 140 and reorderer 160.

[0086] Reorderer 160 may perform reordering of coefficient values with respect to the quantized residual values.

[0087] The rearranger 160 can change the coefficients from a two-dimensional block form to a one-dimensional vector form using a coefficient scanning method. For example, reorderer 160 may scan DC coefficients and even high frequency coefficients using a zigzag scanning technique to convert the coefficients into one-dimensional vector form. Depending on the intra prediction mode and the size of the transform element, instead of zigzag scanning, vertical scanning may be used, in which 2D block type coefficients are scanned in the column direction, or horizontal scanning, in which 2D block type coefficients are scanned in the row direction. That is, the reordering apparatus can determine the scanning method to be used from zigzag scanning, vertical scanning, and horizontal scanning, depending on the transform element size and the intra prediction mode.

[0088] Entropy encoder 165 may perform entropy encoding based on values calculated by reordering apparatus 160. For entropy coding, various encoding methods can be used, for example, such as exponential Golomb, context-adaptive variable-length coding (CAVLC), and context-adaptive binary arithmetic coding (CABAC).

[0089] The entropy encoder 165 can encode various information such as residual coefficient information and coding element block type information, prediction mode information, division information, prediction element information, transmission element information, motion vector information, reference frame information, block interpolation information, and filtering information from reorderer 160 and predictors 120 and 125.

[0090] The entropy encoder 165 may perform entropy encoding on the values of the coding element coefficients input from the reordering device 160.

[0091] Inverse quantizer 140 may inversely quantize values quantized by quantizer 135, and inverse transformer 145 may inversely transform values transformed by transformer 130. The residual value generated by inverse quantizer 140 and inverse transformer 145 may be combined with a prediction element predicted by the device motion estimation, motion compensator, and internal predictor from predictors 120 and 125 to generate a reconstructed block.

[0092] Filter 150 may employ at least one of a deblocking filter, an offset corrector, and an adaptive loop filter.

[0093] The deblocking filter can remove block distortion that occurred due to boundaries between blocks of the reconstructed image. For the purpose of determining whether to perform deblocking, you can determine whether to apply a deblocking filter to the current block based on the pixels included in multiple rows or columns in the block. When a deblocking filter is applied to a block, a strong filter or a weak filter may be applied depending on the required deblocking filtering strength. In addition, when a deblock filter is applied, vertical filtering and horizontal filtering can be performed such that horizontal filtering and vertical filtering are processed in parallel.

[0094] The offset corrector can correct the offset from the original image in the deblocked image in pixel elements. A method of dividing pixels included in an image into a certain number of areas, determining an area to be shifted, and applying the offset to the specific area, or a method of applying the offset taking into account the edge information of each pixel can be used to perform offset correction on a particular image. .

[0095] Adaptive contour filtering (ALF) can be performed based on values obtained by comparing the filtered reconstructed image and the original image. By dividing the pixels included in the image into predefined groups and defining a filter to be applied to each group, differential filtering can be performed on each group. Information about whether to apply ALF may be transmitted on a per coding unit (CU) basis, and the shape and filter coefficients of the ALF filter to be applied may vary from block to block. Moreover, the ALF filter in the same form (fixed form) can be applied regardless of the characteristic of the block to be filtered.

[0096] Memory 155 may store the reconstructed block or image calculated by filter 150. The stored reconstructed block or image may be provided to predictors 120 and 125 when performing inter-image prediction.

[0097] In FIG. 2 is a block diagram of a decoder according to an embodiment of the present invention.

[0098] As shown in FIG. 2, video decoder 200 may include an entropy decoder 210, a reorderer 215, an inverse quantizer 220, an inverse transformer 225, predictors 230 and 235, a filter 240, and a memory 245.

[0099] When a video bitstream is input from an encoder, the input bitstream can be decoded according to a procedure opposite to that of the encoder.

[0100] The entropy decoder 210 may perform entropy decoding according to a procedure opposite to the procedure in which the encoder entropy encoder performs entropy encoding. For example, various methods such as exponential Golomb method, context adaptive variable length coding (CAVLC), and context adaptive binary arithmetic coding (CABAC) may be used to match the method performed by the encoder.

[0101] The entropy decoder 210 can decode information related to intra and inter-picture prediction performed by the encoder.

[0102] The reordering apparatus 215 may perform reordering on the bitstream subjected to entropy decoding by the entropy decoder 210 based on the reordering method used by the encoder. Reorderer 215 may recover coefficients expressed in one-dimensional vector form into two-dimensional block type coefficients to reorder the two-dimensional block type coefficients. The reordering apparatus 215 may receive information related to coefficient scanning performed by the encoder and perform reordering by an inverse scanning method based on the scanning order of the encoder.

[0103] Inverse quantizer 220 may perform inverse quantization based on the values of the reordered block coefficients and quantization parameters provided by the encoder.

[0104] The inverse transformer 225 may perform an inverse DCT or an inverse DST on a quantization result performed by the encoder, wherein the inverse DCT or inverse DST is a transform inverse to the transform that was performed by the transform element, that is, the DCT or DST. In this case, the DCT transform base may comprise at least one of DCT2 or DCT8, and the DST transform base may comprise DST7. Alternatively, when the transform is skipped in the encoder, the inverter 225 may not perform the inverse transform. The inverse transform may be performed based on the transmission element determined by the encoder. In the decoder inverter 225, transform methods (eg, DCT and DST) may be selectively performed depending on multiple pieces of information such as prediction method, current block size, and prediction direction.

[0105] Predictors 230 and 235 may generate a prediction block based on information related to prediction block generation that is provided by entropy decoder 210 and information about a previously decoded block or image that is provided by storage device 245.

[0106] As described above, when intra prediction is performed in the same manner as in the encoder, and the prediction element is equal in size to the transform element, intra prediction may be performed on the prediction element based on pixels located to the left of the prediction element, pixels located in the upper left corner of the prediction element, and pixels located at the top of the prediction element. On the other hand, when intra prediction is performed and the prediction element is different in size from the transform element, intra prediction can be performed using a reference pixel based on the transform element. In addition, intra prediction using N×N partitioning only for the minimum coding element can be used.

[0107] Predictors 230 and 235 may include a prediction element qualifier, an inter-image prediction predictor, and an internal predictor. The prediction element determiner may receive various information such as information about the prediction element, information about the prediction mode for the intra prediction method, and information related to motion prediction for the inter-picture prediction method from the entropy decoder 210, classify the prediction element in the current element encoding and determine whether the prediction element performs inter-picture prediction or intra prediction. Using information necessary for inter-picture prediction of the current prediction element provided by the encoder, inter-picture predictor 230 can perform inter-picture prediction regarding the current prediction element based on information included in at least one of an image preceding the current image including the current prediction element, or an image following the current image. Alternatively, inter-image prediction may be performed based on information about some previously reconstructed regions in the current image including the current prediction element.

[0108] The inter-picture prediction predictor 230 can determine whether a motion prediction method for a prediction element that is included in a corresponding encoding element is a skip mode, a merge mode, an AMVP mode, or an intra-block copy mode based on the encoding element to perform prediction carried out between images.

[0109] The internal predictor 235 may generate a prediction block based on information about the pixels in the current image. When the prediction element is a prediction element subject to intra prediction, the intra predictor 235 may perform intra prediction based on the intra prediction mode information of the prediction element provided by the encoder. The internal predictor 235 may include an AIS filter, a reference pixel interpolator, and a DC filter. The AIS filter, which is a part that performs filtering on a reference pixel of the current block, can determine whether to apply a filter depending on the prediction mode of the current prediction element. The AIS filter may perform AIS filtering on a reference pixel of the current block using the AIS filter information and prediction mode of the prediction element provided by the encoder. When the prediction mode of the current block is a mode in which AIS filtering is not performed, the AIS filter may not be applied.

[0110] When the prediction mode of the prediction block is a prediction mode in which intra prediction is performed based on a pixel value obtained by interpolating a reference pixel, a reference pixel interpolator may interpolate the reference pixel to generate a reference pixel at a pixel element less than or equal to an integer. When the prediction mode of the current prediction element is a prediction mode in which a prediction block is generated without interpolating a reference pixel, the reference pixel cannot be interpolated. When the prediction mode of the current block is the DC mode, the DC filter can generate the prediction block by filtering.

[0111] The reconstructed block or image may be provided to filter 240. Filter 240 may include a deblocking filter, an offset corrector, and an ALF filter.

[0112] Filter 240 may receive from the encoder information about whether a deblocking filter is applied to a corresponding block or image, or information about whether a strong filter is applied or a weak filter is applied when applying the deblocking filter. The deblocking filter of the decoder may receive information related to the deblocking filter that is provided by the encoder, and the decoder may perform deblocking filtering on the corresponding block.

[0113] The offset corrector can perform offset correction on the reconstructed image based on the type of offset correction, offset value information, and the like that is applied to the image during encoding.

[0114] ALF may be applied to an encoding element based on information about whether to apply ALF, ALF coefficient information, or the like, which is provided from the encoder. ALF information may be provided by inclusion in a specific parameter set.

[0115] Memory 245 may store the reconstructed image or block such that the image or block can be used as a reference image or reference block, and may also provide the reconstructed image to an output element.

[0116] In FIG. 3 is a schematic diagram showing a basic element of a code tree according to an embodiment of the present invention.

[0117] The largest encoding element may be defined as a code tree block. A single image is divided into many code tree units (CTUs). The code tree element that represents the largest coding element may be called the largest coding unit (LCU). In fig. Figure 3 shows an example in which one image is divided into many code tree elements.

[0118] The code tree element size may be specified at the image level or the sequence level. To this end, information indicating the size of a code tree element may be conveyed by an image parameter set or a sequence parameter set.

[0119] As an example, the code tree element size for the entire image in the sequence may be set to 128×128. Alternatively, one of the sizes 128×128 or 256×256 may be defined as the size of the image-level code tree element. As an example, the code tree element size in the first image may be set to 128×128, and the code tree element size in the second image may be set to 256×256.

[0120] A code tree element may be split to generate a coding block. The encoding block specifies a base element for encoding or decoding processing. As an example, prediction or transformation may be performed for each encoding block, or a predictive encoding mode may be determined for each encoding block. In this case, the predictive coding mode specifies a method for generating a prediction image. As an example, the predictive coding mode may include intra prediction, inter-picture prediction, current picture reference (CPR) (or intra block copy (IBC)), or combined prediction. A prediction block for a coding block may be generated using at least one predictive coding mode of intra prediction, inter-picture prediction, anchor to current picture, or combined prediction for the coding block.

[0121] Information indicating the predictive encoding mode of the current block may be transmitted in the bit stream. As an example, the information may be a 1-bit flag indicating whether the predictive encoding mode is an intra-image mode or an inter-picture mode. Lock to the current picture or combined prediction may be available only when it is determined that the predictive encoding mode of the current block is an inter-picture mode.

[0122] Linking to the current picture is for setting the current picture as a reference picture and obtaining a current block prediction block from a portion of the current picture where encoding or decoding has been performed. In this case, the current image refers to the image containing the current block. Information indicating that the current image anchor applies to the current block may be carried in the bitstream. As an example, the information may be a 1-bit flag. The predictive encoding mode of the current block may be determined to be an anchor to the current image when the flag is true, and the prediction mode of the current block may be determined to be an inter-image prediction when the flag is false ( lie).

[0123] Alternatively, the predictive coding mode of the current block may be determined based on the index of the reference picture. As an example, when the reference picture index indicates the current picture, it may be determined that the predictive encoding mode of the current block is a reference to the current picture. When the reference picture index points to a picture other than the current picture, it can be determined that the predictive coding mode of the current block is inter-picture prediction. That is, anchoring to the current image is a prediction method that uses information about a portion of the current image where encoding or decoding is performed, and inter-image prediction is a prediction method that uses information about another image where encoding or decoding is performed.

[0124] Combined prediction indicates a coding mode obtained by combining two or more of intra prediction, inter-picture prediction, and reference to the current picture. As an example, when combined prediction is applied, the first prediction block may be generated based on one of intra-prediction, inter-image prediction, or current image anchoring, and the second prediction block may be generated based on the other. When the first prediction block and the second prediction block are generated, the final prediction block may be generated by an averaging operation or a weighted sum operation between the first prediction block and the second prediction block. Information indicating whether combined prediction is applied may be carried in the bit stream. The information may be a 1-bit flag.

[0125] In FIG. 4 is a schematic diagram showing various types of encoding block section.

[0126] A coding block may be divided into a plurality of coding blocks based on a quaternary tree division, a binary tree division, or a ternary tree division. The encoding block obtained by dividing may be re-partitioned into a plurality of encoding blocks based on a quaternary tree division, a binary tree division, or a ternary tree division.

[0127] Quaternary tree-based partitioning indicates a partitioning method for dividing the current block into four blocks. As a result of quaternary tree partitioning, the current block can be split into four square partitions (see "SPLIT QT" in Fig. 4A).

[0128] Binary tree-based partitioning indicates a partitioning method for splitting the current block into two blocks. Splitting the current block into two blocks in the vertical direction (i.e., using a vertical line across the current block) can be called a vertical binary tree-based split, and dividing the current block into two blocks in the horizontal direction (i.e., using a horizontal line across the current block) can be called a horizontal binary tree partition. As a result of binary tree partitioning, the current block can be split into two non-square partitions. In fig. 4B "SPLIT_BT_VER" represents the result of vertical splitting based on a binary tree. In fig. 4C "SPLIT_ВТ_HOR" represents the result of a horizontal split based on a binary tree.

[0129] Ternary tree-based partitioning indicates a partitioning method for dividing the current block into three blocks. Dividing the current block into three blocks in the vertical direction (i.e., using two vertical lines across the current block) can be called a ternary tree-based vertical division, and dividing the current block into three blocks in the horizontal direction (i.e., using two horizontal lines across the current block) can be called horizontal division based on a ternary tree. As a result of ternary tree partitioning, the current block can be split into three non-square partitions. In this case, the width or height of the section located in the center of the current block can be twice as large as the other sections. In fig. 4D “SPLIT_TT_VER” represents the result of a vertical split based on a ternary tree. In fig. 4E "SPLIT_TT_HOR" represents the result of horizontal splitting based on a ternary tree.

[0130] The number of times that a code tree element is split can be set as a section depth (split depth). The maximum section depth of a code tree element can be defined at the sequence level or the image level. Thus, the maximum section depth of a code tree element may vary depending on the sequence or image.

[0131] Alternatively, the maximum section depth may be determined individually for each separation method. As an example, the maximum partition depth allowed for a quaternary tree partition may be different from the maximum partition depth allowed for a binary tree partition and/or ternary tree partition.

[0132] The encoder may transmit information indicating at least one of the section types or section depths of the current block in the bitstream. The decoder may determine the section type and section depth of the code tree element based on the information parsed from the bitstream.

[0133] In FIG. 5 is a schematic diagram illustrating the division aspect of a code tree element.

[0134] Dividing a coding block using a division method such as a quaternary tree division, a binary tree division, and/or a ternary tree division may be referred to as a multi-tree division (multi-tree division).

[0135] Coding blocks generated by applying multi-tree partitioning to a coding block may be referred to as lower coding blocks. When the section depth of the coding block is k, the section depth of the lower coding blocks is set to k+1.

[0136] Conversely, with respect to coding blocks with section depth k+1, a coding block with section depth k may be referred to as an upper coding block.

[0137] The section type of the current encoding block may be determined based on at least one of the section types of the upper coding block or the section type of the adjacent coding block. In this case, the adjacent encoding block is adjacent to the current encoding block and may contain at least one of an adjacent block located above the current encoding block, an adjacent block located to the left of the current encoding block, or an adjacent block adjacent to the upper left corner of the current block coding. In this case, the partition type may include at least one of having a quaternary tree partition, having a binary tree partition, having a binary tree partition direction, having a ternary tree partition, or having a ternary tree partition direction.

[0138] For the purpose of determining the partition type of an encoding block, information indicating whether the encoding block is divided may be transmitted in a bit stream. The information is a 1-bit flag "split_cu_flag", and the flag value true indicates that the encoding block is split using a multi-tree splitting method.

[0139] When split_cu_flag is true, information indicating whether a coding block is split by a quaternary tree split may be transmitted in the bitstream. The information is a 1-bit flag "split_qt_flag", and when the flag is true, the encoding block can be split into four blocks.

[0140] For example, in FIG. 5 shows that four encoding blocks with a partition depth of one are generated because the code tree element is split by a quaternary tree-based partition. It is also shown that the quaternary tree division is reapplied to the first encoding block and the fourth encoding block of the four encoding blocks generated by performing the quaternary tree division. As a result, four coding blocks with a section depth of two can be generated.

[0141] Also, by reapplying the quaternary tree division to a coding block with a section depth of two, a coding block with a section depth of three can be generated.

[0142] When a quaternary tree division is not applied to a coding block, then whether a binary tree division or a ternary tree division should be performed on the coding block can be determined taking into account at least one of the encoding block size , whether the encoding block is located at the edge of the image, the maximum section depth, or the division aspect of an adjacent block. When it is determined that a binary tree division or a ternary tree division has been performed on an encoding block, information indicating the division direction may be transmitted in the bit stream. The information may be a 1-bit mtt_split_cu_vertical_flag. Based on the flag, it can be determined whether the dividing direction is vertical or horizontal. In addition, information indicating which of the binary tree division and the ternary tree division is applied to the encoding block may be transmitted in the bit stream. The information may be a 1-bit mtt_split_cu_binary_flag. Based on the flag, it can be determined whether a binary tree-based partition or a ternary tree-based partition is applied to the encoding block.

[0143] For example, in FIG. 5 shows that the binary tree-based vertical partitioning is applied to a coding block with a partition depth of one, the ternary tree-based vertical partitioning is applied to the left encoding block from the encoding blocks generated by the binary tree-based vertical partitioning, and the vertical partitioning on The binary tree base is applied to the right encoding block.

[0144] Inter-picture prediction is a predictive coding mode in which the current block is predicted using information about the previous picture. As an example, a block included in the previous image and located at the same location as the current block (hereinafter referred to as a compatible block) can be set as a prediction block of the current block. A prediction block generated based on a block located at the same location as the current block will be called a compatible prediction block.

[0145] Meanwhile, when an object present in the previous image moves to another location in the current image, the current block can be effectively predicted using the movement of the object. For example, when the motion direction and size of an object can be found by comparing a previous image with a current image, a prediction block (or prediction image) of the current block can be generated taking into account the motion information of the object. Hereinafter, the prediction block generated using the motion information may be called a motion prediction block.

[0146] A residual block may be generated by subtracting a prediction block from a current block. At this time, when motion of the object is present, the energy of the residual block can be reduced, and accordingly, the compression efficiency of the residual block can be improved by using a motion prediction block instead of a fusion prediction block.

[0147] As described above, generating a prediction block using motion information may be called motion-compensated prediction. In most inter-image predictions, a prediction block may be generated based on motion compensated prediction.

[0148] The motion information may comprise at least one of a motion vector, a reference image index, a prediction direction, or a bidirectional weight index. The motion vector indicates the direction of motion and the size of the object. The reference picture index indicates the reference picture of the current block of the reference pictures included in the reference picture list. The prediction direction indicates one of unidirectional L0 prediction, unidirectional L1 prediction, or bidirectional prediction (L0 prediction and L1 prediction). At least one of the movement information in the L0 direction or the movement information in the L1 direction may be used depending on the prediction direction of the current block. The bidirectional weight index indicates a weight applied to the L0 prediction block and a weight applied to the L1 prediction block.

[0149] In FIG. 6 is a flowchart of an inter-image prediction method according to an embodiment of the present invention.

[0150] As shown in FIG. 6, the inter-picture prediction method includes determining (S601) an inter-picture prediction mode of the current block, obtaining (S602) motion information of the current block according to the determined inter-picture prediction mode, and performing motion compensation prediction (S602). (S603) regarding the current block based on the received motion information.

[0151] In this case, an inter-picture prediction mode that represents various methods for determining motion information of the current block may include an inter-picture prediction mode using translational motion information and an inter-picture prediction mode using information about affine motion. As an example, an inter-image prediction mode using translational motion information may include a fusion mode and a motion vector prediction mode, and an inter-image prediction mode using affine motion information may include an affine fusion mode and a vector prediction mode. affine movement. Motion information of the current block can be determined based on information analyzed from the bit stream or blocks adjacent to the current block in accordance with the inter-picture prediction mode.

[0152] An inter-image prediction method using affine motion information will be described in detail below.

[0153] In FIG. 7 is a schematic diagram illustrating the nonlinear motion of an object.

[0154] The movement of an object in an image may occur non-linearly. For example, as in the example shown in FIG. 7, such nonlinear movement of an object as approaching, moving away, rotating and affine transformation can occur. When nonlinear motion of an object occurs, the translational motion vector cannot effectively represent the motion of the object. Accordingly, it is possible to improve the encoding efficiency by using affine motion in the part where nonlinear motion of the object occurred instead of translational motion.

[0155] In FIG. 8 is a flowchart of an inter-image prediction method based on affine motion according to an embodiment of the present invention.

[0156] Whether the inter-picture prediction method based on affine motion is applied to the current block can be determined based on information analyzed from the bitstream. In detail, whether the affine motion-based inter-image prediction method is applied to the current block can be determined based on at least one of a flag indicating whether the affine fusion mode is applied to the current block or a flag indicating whether mode for predicting the affine motion vector to the current block.

[0157] When the inter-image prediction method based on affine motion is applied to the current block, the affine motion model of the current block can be determined (S801). At least one of the 6-parameter affine motion model or the 4-parameter affine motion model may be defined as an affine motion model. A 6-parameter affine motion model expresses an affine model using six parameters, and a 4-parameter affine motion model expresses an affine model using four parameters.

[0158] Equation 1 expresses affine motion using six parameters. Affine motion represents translational motion for a given region defined by affine source vectors.

[0159] [Equation 1]

v _x = a x-by+e

v _y =cx+dy+j

[0160] When affine motion is expressed using six parameters, the complete motion can be expressed, but the number of bits required to encode the parameters may increase, thereby reducing the encoding efficiency. Accordingly, affine motion can be expressed using four parameters. Equation 2 expresses the affine motion using four parameters.

[0161] [Equation 2]

v _x = a x-by+e

v _y =bx+ a y+j

[0162] Information for determining the affine motion model of the current model may be encoded and transmitted in a bit stream. As an example, the information may be a 1-bit flag "affine_type_flag". A flag value of 0 may indicate that a 4-parameter affine motion model is applied, and a flag value of 1 may indicate that a 6-parameter affine motion model is applied. The flag may be encoded in slice, tile, or block elements (for example, encoding blocks or code tree elements). When a flag is passed at the slice level, the affine motion model defined at the slice level can be applied to all blocks belonging to the slice.

[0163] Alternatively, the affine motion model of the current block may be determined based on the inter-image affine prediction mode of the current block. As an example, when the affine merging mode is applied, the affine motion model of the current block may be determined to be a 4-parameter motion model. On the other hand, when the affine motion vector prediction mode is applied, information for determining the affine motion model of the current block can be encoded and transmitted in a bit stream. As an example, when the affine motion vector prediction mode is applied to the current block, the affine motion model of the current block can be determined based on the 1-bit flag “affine_type_flag”.

[0164] The affine source vector of the current block can then be obtained (S802). When a 4-parameter affine motion model is selected, motion vectors at two control points of the current block can be obtained. On the other hand, when the 6-parameter affine motion model is selected, the motion vectors at the three control points of the current block can be obtained. The motion vector at a control point can be called an affine source vector. The reference point may include at least one of the top left corner, top right corner, or bottom left corner of the current block.

[0165] In FIG. 9 is a schematic diagram illustrating the affine source vector for each affine motion model.

[0166] In the 4-parameter affine motion model, affine source vectors can be obtained for two of the upper left corner, upper right corner, or lower left corner. For example, as in the example shown in FIG. 9A, when the 4-parameter affine motion model is selected, the affine vector can be obtained using the affine source vector sv ₀ for the top left corner of the current block (for example, the top left sample (x ₀ , y ₀ )) and the affine source vector sv ₁ for top-right corner of the current block (for example, top-right count (x ₁ , y ₁ )). An affine source vector for the lower left corner can be used instead of an affine source vector for the upper left corner, or an affine source vector for the lower left corner can be used instead of an affine source vector for the upper right corner.

[0167] In the 6-parameter affine motion model, affine source vectors can be obtained for the upper left corner, upper right corner and lower left corner. For example, as in the example shown in FIG. 9B, when the 6-parameter affine motion model is selected, the affine vector can be obtained using the affine source vector sv ₀ for the top left corner of the current block (for example, the top left sample (x ₀ , y ₀ )), the affine source vector sv ₁ for the top right corner of the current block (for example, the top right sample (x ₁ , y ₁ )) and the affine source vector sv ₂ for the top left corner of the current block (for example, the top left sample (x ₂ , y ₂ )).

[0168] In the following embodiment, under the 4-parameter affine motion model, the affine source vectors for the top left control point and the top right control point will be called the first affine source vector and the second affine source vector, respectively. In further embodiments in which a first affine seed vector and a second affine seed vector are used, at least one of the first affine seed vector and the second affine seed vector may be replaced by an affine seed vector (third affine seed vector) for the lower left reference point, or affine source vector (fourth affine source vector) for the bottom right control point.

[0169] In addition, under the 6-parameter affine motion model, the affine source vectors of the top left control point, top right control point, and bottom left control point will be called the first affine source vector, the second affine source vector, and the third affine source vector, respectively. In further embodiments in which a first affine source vector, a second affine source vector, and a third affine source vector are used, at least one of the first affine source vector, the second affine source vector, and the third affine source vector may be replaced by an affine source vector (the fourth affine source vector) for the lower right control point.

[0170] An affine vector can be obtained (S803) for each sub-block using the affine source vectors. In this case, the affine vector indicates the translational vector obtained from the affine source vectors. An affine subblock vector may be called an affine subblock motion vector or a subblock motion vector.

[0171] In FIG. 10 is a schematic diagram illustrating the affine vectors of subblocks under a 4-parameter motion model.

[0172] The affine subblock vector can be obtained based on the reference point location, the subblock location, and the affine source vector. As an example, Equation 3 shows an example of obtaining an affine subblock vector.

[0173] [Equation 3]

[0174] In Equation 3, (x, y) represents the location of the sub-block. In this case, the location of the subblock indicates the location of the reference sample included in the subblock. The reference sample may be a sample that is located in the upper left corner of the subblock, or a sample that has at least an x-coordinate or a y-coordinate located in the center. (x ₀ , y ₀ ) represents the location of the first control point, and (sv _0x , sv _0y ) represents the first affine source vector. In addition, (x ₁ , y ₁ ) represents the location of the second control point, and (sv _1x , sv _1y ) represents the second affine source vector.

[0175] When the first reference point and the second reference point correspond to the upper left corner and upper right corner of the current block, respectively, x ₁ -x ₀ can be set to a value equal to the width of the current block.

[0176] Thereafter, motion compensated prediction can be performed (S804) for each sub-block using the affine vector of the corresponding sub-block. By performing motion-compensated prediction, a prediction block for each sub-block can be generated. Sub-block prediction blocks can be set as the current block prediction block.

[0177] The affine source vector of the current block can be obtained based on the affine source vector of the block adjacent to the current block. When the inter-image prediction mode of the current block is an affine fusion mode, an affine source vector of a fusion candidate included in the fusion candidate list can be defined as an affine source vector of the current block. In addition, when the inter-picture prediction mode of the current block is an affine fusion mode, motion information including at least one of a reference image index, a specific direction prediction flag, or a bidirectional weight of the current block may also be set to the same , as a candidate for merger.

[0178] A merge candidate may be obtained based on a neighboring block of the current block. The neighbor block may include at least one of a spatial neighbor block spatially adjacent to the current block and a temporary neighbor block included in an image different from the current image.

[0179] In FIG. 11 is a schematic diagram showing neighboring blocks that can be used to obtain a merge candidate.

[0180] An adjacent block of a current block may include at least one of an adjacent block (A) adjacent to the left side of the current block, an adjacent block (B) adjacent to the top side of the current block, an adjacent block (C) adjacent to the top right the corner of the current block, an adjacent block (D) adjacent to the lower left corner of the current block, or an adjacent block (E) adjacent to the upper left corner of the current block. When the top left sample of the current block has coordinates (x ₀ , y ₀ ), the left neighboring block A contains a sample located at (x ₀ -1, y ₀ +H-1), and the top neighboring block B contains a sample located at ( x ₀ +W-1, y ₀ -1). In this case, W and H represent the width and height of the current block, respectively. The upper right adjacent block C contains a sample located at (x ₀ +W, y ₀ -1), and the lower left adjacent block D contains a sample located at (x ₀ -1, y ₀ +H). The top left adjacent block E contains a sample located at (x ₀ -1, y ₀ -1).

[0181] When a neighbor block is encoded in an affine inter-picture prediction mode, the affine source vector of a merge candidate can be obtained based on the affine source vector of the corresponding neighbor block. Hereinafter, a neighbor block encoded in the affine inter-image prediction mode will be called an affine neighbor block, and a merge candidate obtained from an affine neighbor block will be called an affine merge candidate.

[0182] Neighboring blocks can be detected in a predefined scan order to generate an affine merge candidate for the current block. The scanning order can be preset in the encoder and decoder. As an example, neighboring blocks may be detected in the following order: A, B, C, D, and E. In addition, affine merge candidates may be sequentially obtained from the detected affine neighboring blocks. Alternatively, the scanning order may be adaptively determined based on at least one of the size, shape, or affine motion model of the current block. That is, scan orders for blocks that differ in at least one of size, shape, and affine motion pattern may differ from each other.

[0183] Alternatively, blocks located on top of the current block may be sequentially discovered to obtain one affine merge candidate from the affine neighboring block discovered first, and blocks located to the left of the current block may be sequentially discovered to obtain one affine merge candidate from the affine neighboring block found first. In this case, neighboring blocks located on top of the current block may include at least one of adjacent block E, adjacent block B, or adjacent block C, and blocks located to the left of the current block may include at least one of block A or block D. In this case, the adjacent block E can be classified as the block located to the left of the current block.

[0184] Although not shown, an affine merge candidate may be obtained from a temporary neighbor of the current block. In this case, the temporary neighboring block may include a block located at the same location in the fusion image as the current block or a block adjacent to the current block. In detail, when a temporal neighbor block of a current block is encoded in an affine inter-picture prediction mode, an affine merge candidate can be obtained based on the affine source vector of the temporal affine merge candidate.

[0185] A merge candidate list containing affine merge candidates may be generated, and an affine source vector of one of the merge candidates included in the merge candidate list may be determined to be the affine source vector of the current block. To achieve this, index information for identifying one of the merge candidates may be encoded and transmitted in a bit stream.

[0186] As another example, while neighboring blocks are detected in scan order, the affine source vector of the current block can be obtained from the affine source vector of the affine neighbor block found first.

[0187] The affine source vector of the current block can be obtained using the affine source vector of the neighboring block in affine merge mode.

[0188] When the inter-picture prediction mode of the current block is an affine motion vector prediction mode, the affine source vector of the motion vector prediction candidate included in the motion vector prediction candidate can be determined as the affine source vector prediction value of the current block. The affine source vector of the current block can be obtained by adding the difference value of the affine source vectors to the prediction value of the affine source vector.

[0189] An affine source vector prediction candidate may be obtained based on a neighboring block of the current block. In detail, the detection of neighboring blocks located on top of the current block is carried out in a predetermined scanning order, and the first candidate for affine seed vector prediction can be obtained from the affine neighboring block found first. Also, the detection of neighboring blocks located to the left of the current block is carried out in a predetermined scan order, and the second candidate for affine seed vector prediction can be obtained from the affine neighboring block found first.

[0190] Information for determining the difference value of the affine source vectors may be encoded and transmitted in the bit stream. The information may include size information indicating the size of the difference value of the affine source vectors, and sign information indicating the sign of the difference value of the affine source vectors. The difference value of the affine source vectors for each control point can be set to the same value. Alternatively, the difference value of the affine source vectors can be set differently depending on the reference point.

[0191] As described above, the affine source vector of the affine merge candidate or the affine source vector prediction candidate is obtained from the affine source vector of the affine neighboring block, and the affine source vector of the current block can be obtained using the obtained affine source vector of the affine merge candidate or candidate for predicting the affine source vector. Alternatively, after performing affine neighbor detection in a predetermined scanning order, the affine source vector of the current block can be obtained from the affine source vector of the affine neighbor block found first.

[0192] A method for obtaining an affine source vector of a current block, an affine merge candidate, or a prediction candidate affine source vector from an affine source vector of an affine neighboring block will be described in detail below. In further embodiments, obtaining the affine source vector of the current block may also be understood as obtaining the affine source vector of the affine merge candidate or obtaining the affine source vector of the prediction affine source vector.

[0193] In FIG. 12 is a schematic diagram showing an example of obtaining an affine source vector of a current block based on an affine source vector of an affine neighboring block.

[0194] When the first affine source vector nv ₀ for the top left control point and the second affine source vector nv ₁ for the top right control point are stored for the affine neighbor block, the third affine source vector nv ₂ for the bottom left control point of the affine neighbor block can be obtained based on the first affine seed vector and the second affine seed vector. Equation 4 shows an example of obtaining a third affine seed vector.

[0195] [Equation 4]

[0196] In Equation 4, (nv _0x , nv _0y ) represents the first affine source vector nv ₀ , (nv _1x , nv _1y ) represents the second affine source vector nv ₁ , and (nv _2x , nv _2y ) represents the third affine source vector nv ₂ . In addition, (x _n0 , x _n0 ) represents the location of the first control point, (x _n1 , x _n1 ) represents the location of the second control point, and (x _n2 , x _n2 ) represents the location of the third control point.

[0197] Thereafter, the affine source vector of the current block can be obtained using the first affine source vector, the second affine source vector, and the third affine source vector. Equation 5 shows an example of obtaining the first affine source vector v ₀ of the current block, and Equation 6 shows an example of obtaining the second affine source vector v ₁ of the current block.

[0198] [Equation 5]

[0199] [Equation 6]

[0200] In Equation 5 and Equation 6, (v _0x , v _0y ) represents the first affine source vector sv ₀ of the current block, and (v _1x , V _1y ) represents the second affine source vector sv ₁ of the current block. In addition, (x ₀ , y ₀ ) represents the location of the first control point, and (x ₁ , y ₁ ) represents the location of the second control point. For example, the first control point indicates the top left corner of the current block, and the second control point indicates the top right corner of the current block.

[0201] In the example described above, it was described that the affine source vectors of the current block are obtained using three affine source vectors for the affine neighboring block. As another example, the affine source vectors of the current block can be obtained using only two of the affine source vectors of an affine neighboring block.

[0202] Alternatively, the affine source vectors of the current block can be obtained using a fourth affine source vector for the lower right corner instead of the first affine source vector for the upper left corner, the second affine source vector for the upper right corner, or the third affine source vector for the lower left corner relative to affine neighboring block.

[0203] Specifically, when the top edge of the current block is adjacent to the top edge of a code tree element, and it is intended to use an affine source vector for the top reference point (e.g., top left corner or top right corner) of an affine neighbor block adjacent to the top side of the current block (hereinafter referred to as the upper affine neighbor block), this affine source vector must first be stored in a storage device. Accordingly, the problem of increasing the number of linear buffers may arise. Thus, when the top edge of the current block is adjacent to the top edge of a code tree element, the setting can be made such that the affine source vector for the bottom reference point (e.g. bottom left corner or bottom right corner) instead of the affine source vector for the top reference point used for the top affine neighbor block. As an example, the affine source vectors of the current block can be obtained using a third affine source vector for the lower left corner and a fourth affine source vector for the lower right corner relative to the upper affine adjacent block. In this case, the affine source vectors for the bottom corners can be obtained by copying the affine source vectors for the top corners or from the affine source vectors for the top corners. As an example, a fourth affine source vector for the lower right corner may be used by transforming or replacing the first affine source vector, the second affine source vector, or the third affine source vector.

[0204] Equation 7 and Equation 8 show examples of obtaining a first affine source vector and a second affine source vector of the current block using a third affine source vector for the lower left control point of an adjacent affine vector and a fourth affine source vector for the lower right control point of an adjacent affine vector .

[0205] [Equation 7]

[0206] [Equation 8]

[0207] In Equation 7 and Equation 8, (x _n2 , y _n2 ) represents the coordinates of the bottom left control point of the affine neighbor block, and (x _n3 , y _n3 ) represents the coordinates of the bottom right control point of the affine neighbor block. (x ₀ , y ₀ ) represents the coordinates of the top left control point of the current block, and (x ₁ , y ₁ ) represents the coordinates of the top right control point of the current block. (nv _2x , nv _20y ) represents the affine source vector for the bottom left reference point (i.e., the third affine source vector) of the affine neighbor block, and (nv _3x , nv _3y ) represents the affine source vector for the bottom right reference point (i.e. e. the fourth affine source vector) of the affine neighboring block. (v _0x , v _0y ) represents the affine source vector for the top left control point (i.e., the first affine source vector) of the current block, and (v _1x , v _1y ) represents the affine source vector for the top right control point (i.e. . second affine source vector) of the current block.

[0208] The division operation included in Equation 7 and Equation 8 may be changed to a bit shift operation. A bit shift operation can be performed based on the value obtained from the width (ie x _n3 -x _n2 ) between the bottom left reference point and the bottom right reference point.

[0209] As in the example presented above, the affine source vectors of the current block can be obtained from the affine source vectors of the affine neighboring block where encoding or decoding is performed. To do this, the affine source vectors of the affine neighboring block where encoding or decoding is performed must be stored in a storage device. However, since the affine source vectors of the affine neighbor block are stored in the memory in addition to the translational vectors (ie, affine vectors) of the sub-blocks included in the affine neighbor block, there is a problem of increasing the storage usage. To solve this problem, the affine source vector of the current block can be obtained using the motion vector of the subblock adjacent to the reference point of the affine neighbor block, instead of the affine source vector of the affine neighbor block. That is, the motion vector of a subblock adjacent to the reference point of an affine neighbor block can be set as the affine source vector of the affine neighbor block. In this case, a sub-block may be a block with a predetermined size or shape in the encoder and decoder, and may be a block with a basic size or shape where the motion vector is stored. As an example, the subblock may be a 4x4 square block. Alternatively, the motion vector for a particular sample location can be set to the affine source vector of the affine neighbor block.

[0210] In FIG. 13 is a schematic diagram showing an example of setting a motion vector of a sub-block as an affine source vector of an affine neighboring block.

[0211] The motion vector of a subblock adjacent to a reference point may be set as an affine source vector of the corresponding reference point. As an example, as in the example shown in FIG. 13, the motion vector (nv _4x , nv _4y ) of the sub-block adjacent to the lower left corner (lower left sub-block) of the affine neighboring block can be set as the affine source vector (nv _2x , nv _2x ) for the lower left control point, and the motion vector (nv _5x , nv _5y ) of the subblock adjacent to the lower right corner (lower right subblock) can be set as an affine source vector (nv _3x , nv _3y ) for the control point in the lower right corner. In this case, the lower left subblock may mean the subblock containing a sample adjacent to the lower left control point (x _n2 , y _n2 ) in an adjacent affine block, (for example, a sample located at (x _n2 , y _n2 -1)) and the lower right sub-block may mean a block containing a sample adjacent to the lower right control point (x _n3 , y _n3 ) in an adjacent affine block (for example, a sample located at (x _n3 -1, y _n3 -1)). When the affine source vectors of the current block are obtained based on Equation 7 and Equation 8, the third affine source vector of the affine neighboring block can be replaced by the motion vector of the lower left sub-block, and the fourth affine source vector can be replaced by the motion vector of the lower right sub-block.

[0212] In the following embodiment, a subblock used as an affine source vector of an affine neighbor block will be called an affine subblock.

[0213] According to an embodiment of the present invention, an affine subblock can be determined based on a sample located at a specific location. As an example, a subblock containing a sample located at a particular location may be set as an affine subblock. In the following, the sample located at a specific location will be called the reference sample. In addition, the reference sample used to determine the affine subblock for the lower left reference point will be referred to as the lower left reference sample, and the reference sample used to determine the affine subblock for the lower right reference point will be referred to as the lower right reference sample.

[0214] The lower left reference sample and the lower right reference sample may be selected from samples included in the affine adjacent block. As an example, at least one of the upper left sample, lower left sample, upper right sample, or lower left sample of the lower left subblock is set as the lower left reference sample, and at least one of the upper left sample, lower left sample, upper right reference or the lower left sample of the lower right subblock is set as the lower right reference sample. Accordingly, the motion vectors of the lower left sub-block containing the lower left reference sample and the lower right sub-block containing the lower right reference sample can be set as an affine source vector for the lower left reference point and an affine source vector for the lower right reference point, respectively.

[0215] As another example, at least one of the lower left reference sample or the lower right reference sample may be set to be a sample located outside the affine neighbor block. This will be described in detail below with reference to FIG. 14-16.

[0216] In FIG. 14-16 are schematic diagrams showing the location of the reference count.

[0217] As in the example shown in FIG. 14A, for the lower left reference point, the upper left sample of the lower left sub-block may be set as the reference sample (x _n4 , y _n4 ). Thus, the lower left subblock containing the reference sample (x _n4 , y _n4 ) can be set as an affine subblock for the lower left reference point.

[0218] For the lower right reference point, a sample located to the right of the upper right sample of the lower right sub-block may be set as the reference sample (x _n5 , y _n5 ). Thus, a subblock adjacent to the right side of the lower right subblock containing the reference sample (x _n5 , y _n5 ) can be set as an affine subblock for the lower right reference point.

[0219] Alternatively, as in the example shown in FIG. 14B, for the lower left reference point, a sample located to the left of the upper left sample of the lower left sub-block may be set as the reference sample (x _n4 , y _n4 ). Thus, a sub-block adjacent to the left side of the bottom-left sub-block containing the reference sample (x _n4 , y _n4 ) can be set as an affine sub-block for the bottom-left reference point.

[0220] For the lower right reference point, the upper right sample of the lower right subblock may be set as the reference sample (x _n5 , y _n5 ). Thus, the lower right subblock containing the reference sample (x _n5 , y _n5 ) can be set as an affine subblock for the lower right reference point.

[0221] Alternatively, as in the example shown in FIG. 15A, for the lower left reference point, the lower left count of the lower left sub-block may be set as the reference count (x _n4 , y _n4 ). Thus, the lower left subblock containing the reference sample (x _n4 , y _n4 ) can be set as an affine subblock for the lower left reference point.

[0222] For the lower right reference point, a sample located to the right of the lower right sample of the lower right subblock may be set as the reference sample (x _n5 , y _n5 ). Thus, a subblock adjacent to the right side of the lower right subblock containing the reference sample (x _n5 , y _n5 ) can be set as an affine subblock for the lower right reference point.

[0223] Alternatively, as in the example shown in FIG. 15B, for the lower left reference point, a sample located to the left of the lower left sample of the lower left sub-block may be set as the reference sample (x _n4 , y _n4 ). Thus, a sub-block adjacent to the left side of the bottom-left sub-block containing the reference sample (x _n4 , y _n4 ) can be set as an affine sub-block for the bottom-left reference point.

[0224] For the lower right reference point, the lower right sample of the lower right subblock may be set as the reference sample (x _n5 , y _n5 ). Thus, the lower right subblock containing the reference sample (x _n5 , y _n5 ) can be set as an affine subblock for the lower right reference point.

[0225] Alternatively, as in the example shown in FIG. 16A, for the lower left reference point, a sample located between the upper left sample and the lower left sample of the lower left subblock (eg, the middle left sample) may be set as the reference sample (x _n4 , y _n4 ). Thus, the lower left subblock containing the reference sample (x _n4 , y _n4 ) can be set as an affine subblock for the lower left reference point.

[0226] For the lower right reference point, a sample located to the right of a sample located between the upper right sample and the lower right sample of the lower right subblock (eg, the middle right sample) may be set as the reference sample (x _n5 , y _n5 ). Thus, a subblock adjacent to the right side of the lower right subblock containing the reference sample (x _n5 , y _n5 ) can be set as an affine subblock for the lower right reference point.

[0227] Alternatively, as in the example shown in FIG. 16B, for the lower left reference point, a sample located to the left of a sample located between the upper left sample and the lower left sample of the lower left subblock may be set as the reference sample (x _n4 , y _n4 ). Thus, a sub-block adjacent to the left side of the bottom-left sub-block containing the reference sample (x _n4 , y _n4 ) can be set as an affine sub-block for the bottom-left reference point.

[0228] For the lower right reference point, a sample located between the upper right sample and the lower right sample of the lower right subblock may be set as the reference sample (x _n5 , y _n5 ). Thus, the lower right subblock containing the reference sample (x _n5 , y _n5 ) can be set as an affine subblock for the lower right reference point.

[0229] When the affine source vectors of the current block are obtained based on Equation 7 and Equation 8, the third affine source vector of the affine neighboring block can be replaced by the motion vector of the affine sub-block for the lower left control point, and the fourth affine source vector can be replaced by the motion vector of the affine sub-block for the bottom right control point. In addition, the location of the lower left reference point may be replaced by the location of the lower left reference sample, and the location of the lower right reference point may be replaced by the location of the lower right reference sample.

[0230] Unlike described above with reference to FIG. 14-16, a subblock containing a sample adjacent to a reference sample may be set as an affine subblock. In detail, a sample located outside the affine neighbor sub-block may be set as a reference sample, and a sub-block contained in the affine neighbor block may be set as the affine sub-block. For example, in the example shown in FIG. 14A, a sample located to the right of the upper right sample of the lower right sub-block can be set as a reference sample (x _n5 , y _n5 ), and the lower right sub-block can be set as an affine sub-block for the lower right corner. Alternatively, in the example shown in FIG. 14B, a sample located to the left of the upper left sample of the lower left sub-block can be set as a reference sample (x _n4 , y _n4 ), and the lower left sub-block can be set as an affine sub-block for the lower left corner.

[0231] The above-described embodiments may also apply to FIGS. 15 and 16. That is, in the example shown in FIG. 15A or 16A, a sample located to the right of the lower right sample or middle right sample of the lower right subblock may be set as a reference sample (x _n5 , y _n5 ), and the lower right subblock may be set as an affine subblock for the lower right corner. Alternatively, in the example shown in FIG. 15B or 16B, a sample located to the left of the lower left sample or middle left sample of the lower left subblock may be set as a reference sample (x _n4 , y _n4 ), and the lower left subblock may be set as an affine subblock for the lower left corner.

[0232] As in the example presented above, the affine source vector of the affine neighbor block can be obtained using the motion vector of the affine subblock. For this purpose, in the case of a block where encoding or decoding is carried out, the motion vector can be stored in the elements of the subblocks.

[0233] As another example, after storing a minimum number of affine source vectors for an affine neighbor block, the motion vector of the affine sub-block can be obtained using the stored affine source vectors.

[0234] Equation 9 and Equation 10 show an example of obtaining the motion vector of an affine sub-block using the affine source vectors of the affine neighboring block.

[0235] [Equation 9]

[0236] [Equation 10]

[0237] In Equation 9 and Equation 10, (nv _4x , nv _4y ) represents the motion vector of the affine subblock for the lower left control point, and (nv _5x , nv _5y ) represents the motion vector of the affine subblock for the lower right control point. Since the motion vector of the affine subblock and the affine source vector of the control point are set equal to each other, (nv _4x , nv _4y ) can be replaced by the affine source vector (nv _2x , nv _2y ) for the bottom left control point, and (nv _5x , nv _5y ) can be replaced by an affine source vector (nv _3x , nv _3y ) for the lower right control point.

[0238] (x _n4 , y _n4 ) represents the location of the reference sample relative to the lower left subblock. Alternatively, the center location of the bottom left subblock or the location of the bottom left control point may be used instead. (x _n5 , y _n5 ) represents the location of the reference sample relative to the lower right subblock. Alternatively, the center location of the lower right subblock or the location of the lower right control point may be used instead.

[0239] Equation 9 and Equation 10 may be applied to the case in which the current block is not adjacent to the edge of a code tree element. When the current block is adjacent to the top edge of a code tree element, instead of using Equation 9 and Equation 10, the translation vector of the affine subblock determined based on the lower left reference sample can be set as the third affine source vector, and the translation vector of the affine subblock determined based on the lower right reference sample, can be set as the fourth affine source vector.

[0240] In Equation 7 and Equation 8, (x _n3 -x _n2 ) represents the width between the lower left reference point and the lower right reference point. As described above, _xn3 may be replaced by the lower right reference sample location _xn5 , and _xn2 may be replaced by the lower left reference sample location _xn4 . Next, (x _n3 -x _n2 ) or the value (for example, (x _n5 -x _n4 )) obtained by replacing the location in the above equation with the location of the reference samples will be specified as the variable W _seed , and the variable will be called the width of the original subvector.

[0241] Depending on the location of the reference samples, the width of the original subvector may not be an exponent of the number 2 (eg, 2 ⁿ ). As an example, when the lower left sample of the lower left sub-block is set as the lower left reference sample, and the lower right sample of the lower right sub-block is set as the lower right reference sample, the width of the original sub-vector is not a multiple of 2. When the width of the original sub-vector is not a power of 2 , the width of the original subvector may be converted to an exponent of 2. The transformation may involve adding an offset to or subtracting an offset from the width of the original subvector, or using a sample location adjacent to the reference sample instead of the reference sample location. As an example, the transformed width of the original subvector can be obtained by adding one to the width between the lower left reference sample and the lower right reference sample. Alternatively, the width between the lower left reference sample and an adjacent reference sample adjacent to the right side of the lower right reference sample may be set as the transformed width of the original subvector. After this, the affine source vector of the current block can be obtained by applying the transformed width of the source subvector to Equation 7 and Equation 8.

[0242] The division operation included in Equation 7 and Equation 8 may be changed to a bit shift operation. A bit shift operation can be performed based on the value obtained from the converted width of the original subvector (ie, the value expressed as an exponent of 2).

[0243] When the reference sample used to determine the affine subblock does not belong to the affine neighbor block, the affine source vector of the affine neighbor block can be obtained based on the sample adjacent to the reference sample from the samples included in the affine neighbor block. Specifically, a translation vector of a subblock containing a sample adjacent to a reference sample (hereinafter referred to as a neighboring reference sample) in an affine neighbor block can be set as an affine source vector of the affine neighbor block. As described above, a method for obtaining an affine source vector using a neighboring reference sample can be specified as a modified method for obtaining an affine fusion vector.

[0244] FIG. 17 is a schematic diagram showing an example in which a method for obtaining a modified affine fusion vector is applied.

[0245] When the lower right reference sample (x _n5 , y _n5 ) of the affine neighbor block E does not belong to the affine nearest block, the affine reference vector can be obtained based on the sample (x _n5 -1, y _n5 ) adjacent to the left side of the lower right reference sample, from the samples included in the affine adjacent block. In particular, the translation vector of the subblock containing the adjacent reference sample (x _n5 -1, y _n5 ) can be set as the affine source vector of the lower right reference point.

[0246] In the example shown in FIG. 17, it is shown that the sample adjacent to the right side of the upper right sample of the lower right subblock is the lower right reference sample. Even when a sample adjacent to the right side of the lower right sample of the lower right sub-block, or a sample adjacent to the right side of the middle right sample of the lower right sub-block, is set as the lower right reference sample, an affine source vector can be obtained based on the sample adjacent to the left side adjacent reference reading.

[0247] In addition, even when the lower left reference sample does not belong to an affine neighbor block, according to the embodiment described above, an affine source vector can be obtained based on a sample adjacent to the right side of the lower left reference sample.

[0248] Since the location of the reference sample and the sub-block for obtaining the affine source vector are set differently, the width of the source sub-vector can be set to an exponent of 2.

[0249] Alternatively, affine source vectors can be obtained based on the width of the adjacent block when the motion vectors of the lower left subblock and the lower right subblock are used.

[0250] FIG. 18 is a schematic diagram showing an example of obtaining an affine source vector of an affine merge candidate based on the motion vectors of subblocks belonging to a neighboring block.

[0251] When the top edge of the current block is adjacent to the edge of a code tree element, the affine source vector of the affine merge candidate can be obtained based on the motion vectors of the bottom left sub-block and the bottom right sub-block of the top adjacent block located above the current block. When the location of the upper left sample of the upper neighboring block is assumed to be (xNb, yNb), and the width and height of the upper neighboring block are NbW and NbH, respectively, the lower left sub-block may contain a sample located at (xNb, yNb+yNbH-1 ), and the lower right sub-block may contain a sample located at (xNb+NbW-1, yNb+NbH-1).

[0252] The affine source blocks of an affine merge candidate can be obtained based on the width of the neighboring block and the difference in coordinates between the current block and the neighboring block. As an example, affine source vectors of an affine fusion candidate can be obtained based on equations 11-13 below.

[0253] [Equation 11]

[0254] [Equation 12]

[0255] [Equation 13]

[0256] In equations 11-13, (v _0x , v _0y ) represents the first affine source vector, (v _1x , v _1y ) represents the second affine source vector, and (v _2x , v _2y ) represents the third affine source vector. V _LBx represents the horizontal motion component vector of the lower left sub-block, and V _LBy represents the vertical motion component vector of the lower left sub-block. V _RBx represents the horizontal motion component vector of the lower right sub-block, and V _RBy represents the vertical motion component vector of the lower right sub-block. NbW represents the width of the adjacent block, and xCbW and xCbH represent the width and height of the current block, respectively. xCb represents the x-coordinate of the top-left sample of the current block, and xNb represents the x-coordinate of the top-left sample of the adjacent block.

[0257] An affine merge candidate may be generated to combine the motion vector of the second adjacent block with the affine source vectors derived from the first adjacent block. For example, in the example shown in FIG. 18, the third affine source vector of the affine merge candidate may be obtained based on the motion vector of the subblock located to the left of the current block, rather than based on the motion vectors of the lower left subblock and the lower right subblock.

[0258] FIG. 19 is a schematic diagram showing an example of obtaining affine source vectors of an affine merge candidate based on the motion vectors of subblocks located to the left of the current block.

[0259] The affine source vectors of the affine merge candidate can be obtained based on the motion vectors of the lower left sub-block and the lower right sub-block of the upper neighboring block located above the current block. In detail, the first affine source vector and the second affine source vector of the affine merger candidate can be obtained based on the motion vectors of the lower left sub-block and the lower right sub-block. As an example, the first affine seed vector and the second affine seed vector of the affine fusion candidate can be obtained based on Equations 11 and 12.

[0260] The third affine source vector of the affine merge candidate may be obtained based on the motion vector of a neighboring block adjacent to the left side of the current block. As an example, the motion vector of an adjacent block adjacent to the left side of the current block, or an adjacent block adjacent to the lower left corner of the current block, can be set as the third affine source vector. An adjacent block adjacent to the left side of the current block may contain a sample located at (xCb-1, yCb+CbH-1), and an adjacent block adjacent to the lower left corner of the current block may contain a sample located at (xCb-1 , yCb+CbH). Equation 14 and Equation 15 show an example of setting the motion vector of a neighboring block located to the left of the current block as the third affine source vector.

[0261] [Equation 14]

[0262] [Equation 15]

[0263] In Equation 14, V _Leftx represents the horizontal motion vector of an adjacent block adjacent to the left side of the current block, and V _Lefty represents the vertical motion vector of an adjacent block adjacent to the left side of the current block. In Equation 15, V _LBx represents the horizontal motion vector of an adjacent block adjacent to the lower left corner of the current block, and V _LBy represents the vertical motion vector of an adjacent block adjacent to the lower left corner of the current block.

[0264] The third affine source vector can be obtained using the motion vector of an adjacent block adjacent to the lower left corner of the current block only when the motion vector of an adjacent block adjacent to the left side of the current block is not available.

[0265] The method for obtaining the third affine source vector of the affine merge candidate may be set differently depending on whether the motion vector of a neighboring block located to the left of the current block is available. As an example, when the motion vector of a neighboring block located to the left of the current block (e.g., a subblock adjacent to the left side of the current block, or a subblock adjacent to the lower left corner of the current block) is available, a third affine source vector can be obtained using Equation 14 or Equation 15. On the other hand, when the motion vector of a neighboring block located to the left of the current block is not available, a third affine source vector can be obtained using Equation 13.

[0266] The third affine source vector may be obtained based on the motion vector of a non-adjacent block having a reference reference that is not adjacent to the current block. To obtain the third affine source vector, whether to use the motion vector of an adjacent block having a reference reference adjacent to the current block or use the motion vector of a non-neighboring block having a reference reference which is not adjacent to the current block can be determined based on the location of the top neighbor block used to obtain the affine merge candidate.

[0267] FIG. 20 is a schematic diagram showing an example of deriving affine source vectors of an affine merge candidate based on motion information of an adjacent block or a non-adjacent block located to the left of the current block.

[0268] By comparing the location of the top left sample of the current block with the location of the top left sample of the top adjacent block used to obtain the affine merge candidate, the location of the block to obtain the third affine source vector can be determined. For example, as in the example shown in FIG. 20A, when the x-coordinate (xNb) of the top-left sample of the top adjacent block is less than the x-coordinate (xCb) of the top-left sample of the current block, a third affine source vector can be obtained based on the motion vector a non-adjacent block that is not adjacent to the left side of the current block. In detail, the third affine seed vector can be obtained based on the motion vector of the lower left non-adjacent block containing the sample located at (xNb, yCb+CbH) or the left non-adjacent block containing the sample located at (xNb, yCb+CbH-1), in this case, the samples have the same x coordinate as the upper left sample of the upper adjacent block. Equation 16 and Equation 17 show an example of obtaining a third affine seed vector based on the motion vector of a non-neighboring block.

[0269] [Equation 16]

[0270] [Equation 17]

[0271] In Equation 16, V _Left2x represents the horizontal motion vector of the left non-adjacent block, and V _Left2y represents the vertical motion vector of the left non-adjacent block. In Equation 17, V _LB2x represents the horizontal motion vector of the lower left non-adjacent block, and V _LB2y represents the vertical motion vector of the lower left non-adjacent block.

[0272] The third affine source vector can be obtained using the motion vector of the lower left non-adjacent block only when the motion vector of the left non-adjacent block is not available.

[0273] As in the example shown in FIG. 20B, when the x-coordinate (xNb) of the top left sample of the top adjacent block is greater than or equal to the x-coordinate (xCb) of the top left sample of the current block, the third affine source vector may be derived from the motion vector of an adjacent block that is adjacent to the lower left corner of the current block or its left side. In detail, the third affine source vector can be obtained based on the motion vector of the left neighbor block containing the sample (xCb, yCb+CbH-1) adjacent to the left side of the current block, or the lower left neighbor block containing the sample located at (xCb, yCb +CbH), adjacent to the lower left corner of the current block.

[0274] The location of the reference count of the left adjacent block or the left non-adjacent block may be set differently from that shown in FIGS. 18-20. As an example, a block containing a sample (xCb-1, yCb+subH-1) adjacent to the current block may be set as the left adjacent block, and a block containing a sample (xNb, yCb+subH-1) which is not adjacent to the current block, can be set as a left non-adjacent block. In this case, subH represents the minimum height of the block in which motion information is stored, and can be a natural number such as 2, 4, or 8.

[0275] FIG. 21 is a schematic diagram showing the location of a block for obtaining an affine source vector of an affine merge candidate.

[0276] For example, as in the example shown in FIG. 21A, when the x-coordinate (xNb) of the top left sample of the top adjacent block is less than the x-coordinate (xCb) of the top left sample of the current block, the third affine source vector can be obtained at based on the motion vector of a non-adjacent block that is not adjacent to the left side of the current block. In detail, the third affine source vector can be obtained based on the motion vector of the left non-adjacent block containing a sample (xNb, yCb+subH-1) which has the same x-coordinate as the top left sample of the upper adjacent block, and which is located on a predetermined distance (for example, subH) from the bottom left reference of the top adjacent block. Equation 18 shows an example of obtaining a third affine seed vector based on the motion vector of a non-adjacent block.

[0277] [Equation 18]

[0278] In Equation 18, V _LeftT2x represents the horizontal motion vector of the left non-adjacent block containing a sample located at a predetermined distance from the lower left sample of the upper adjacent block. V _LeftT2y represents the vertical motion vector of the left non-adjacent block containing a sample located at a predetermined distance from the lower left sample of the upper adjacent block.

[0279] Alternatively, as in the example shown in Fig. 21B, when the x-coordinate (xNb) of the top left sample of the top adjacent block is greater than or equal to the x-coordinate (xCb) of the top left sample of the current block, the third affine source vector can be obtained based on the motion vector of the left adjacent block, which is adjacent to the left side of the current block and which contains a sample (xCb-1, yCb+subH-1) horizontally located at a predetermined distance (for example, subH-1) from the top left reference of the current block. Equation 19 shows an example of obtaining a third affine seed vector based on the motion vector of an adjacent block.

[0280] [Equation 19]

[0281] In Equation 19, V _LeftTx represents a horizontal motion vector of a left adjacent block that is adjacent to the left side of the current block and that contains a sample vertically located at a predetermined distance from the top left sample of the current block. V _LeftTy represents the vertical motion vector of a left adjacent block that is adjacent to the left side of the current block and that contains a sample vertically located at a predetermined distance from the top left sample of the current block.

[0282] Alternatively, the reference count may be set at a location other than those shown in FIGS. 18-21. As an example, a sample (xCb-1, yCb) that is adjacent to the left side of the current block and which has the same y-coordinate as the top left sample of the current block, or a sample (xNb, yCb) that is not adjacent to left side of the current block and which has the same y-coordinate as the top left reference of the current block can be set as the reference reference.

[0283] As in the described example, the first affine source vector and the second affine source vector of the affine merge candidate can be obtained from a neighboring block located above the current block, and the third affine source vector of the affine merge candidate can be obtained from the neighboring block, located to the left of the current block. The first affine seed vector and the third affine seed vector of the affine merge candidate may be obtained from a neighboring block located to the left of the current block, and the second affine seed vector of the affine merge candidate may be obtained from a neighboring block located above the current block.

[0284] In the examples shown in Equations 11-13, the affine source vector of the affine merge candidate can be obtained based on the difference value between the motion vector V _LB of the bottom left subblock of the top adjacent block and the motion vector V _RB of the bottom right subblock of the top neighbor block. In this case, the affine source vector of the affine merge candidate can be obtained using the motion vector of another block instead of the motion vector of the lower left subblock or the lower right subblock. As an example, when the motion vector V _LB of the lower left subblock of the upper adjacent block has the same value as the motion vector V _RB of the lower right subblock of the upper adjacent block, V _RB can be replaced by the motion vector of the block containing the sample located to the right of the lower right subblock or under it. As an example, the motion vector of the affine merge candidate can be obtained using the motion vector of the block containing the sample located at (xNb+NbW, yNb+NbH-1) instead of using the motion vector V _RB of the lower right sub-block.

[0285] Alternatively, when the motion vector V _LB of the lower left subblock of the upper adjacent block has the same value as the motion vector V _RB of the lower right subblock of the upper adjacent block, V _LB can be replaced by the motion vector of the block containing the sample located to the left of the lower left subblock or below it. As an example, the motion vector of the affine merge candidate can be obtained using the motion vector of the block containing the sample located at (xNb, yNb+NbH-1+offset) instead of using the motion vector V _LB of the lower left sub-block. In this case, the offset can be an integer that is greater than or equal to 1.

[0286] Alternatively, V _RB may be replaced by the motion vector of a block containing a sample located to the left of or below the lower left subblock, or V _LB may be replaced by the motion vector of a block containing a sample located to the right of or below the lower right subblock.

[0287] A merge candidate can be obtained by combining the motion vectors of a plurality of neighboring vectors adjacent to the current block. A fusion candidate generated by combining the motion vectors of a plurality of neighboring vectors may be called a combined fusion candidate. When a combined merge candidate is obtained, the affine seed vector for a checkpoint can be obtained from one neighboring block adjacent to the checkpoint's neighborhood.

[0288] FIG. 22 is a schematic diagram to describe an example of combining motion vectors of multiple adjacent blocks to obtain a combined fusion candidate.

[0289] The plurality of neighbor blocks may include top neighbor blocks located above the current block and left neighbor blocks located to the left of the current block. The upper adjacent blocks may contain at least one of a block B0 containing a sample located at (xCb+CbW, yCb-1), a block B1 containing a sample located at (xCb+CbW-1, yCb-1), or a block B3 containing the sample located at (xCb, yCb-1). In this case, (xCb, yCb) represents the location of the top-left count of the current block, and CbW represents the width of the current block. The left adjacent blocks may contain at least one of a block A0 containing a sample located at (xCb-1, yCb+CbH), a block A1 containing a sample located at (xCb-1, yCb+CbH-1), or a block A2 containing the sample located at (xCb-1, yCb). In this case, CbH represents the height of the current block.

[0290] In addition, the plurality of neighboring blocks may further comprise temporary neighboring blocks of the current block.

[0291] A combined merge candidate may be obtained by combining parts of the motion information of two or more of a plurality of adjacent blocks. A combined merge candidate can be obtained by combining parts of the motion information of two or three adjacent blocks. In detail, the combined merge candidate may be obtained by combining at least two or more of motion information elements of an adjacent block adjacent to the first reference point, motion information elements of an adjacent block adjacent to the second reference point, or motion information elements of an adjacent block , adjacent to the third control point.

[0292] As an example, a combined merge candidate may be obtained by combining the motion vectors of at least two or more of an adjacent block A2, B2, or B3 adjacent to the top left reference point CP0, an adjacent block B1, or B0 adjacent to the top right control point CP1, or temporary adjacent block T for the lower left control point CP2.

[0293] In this case, the motion vector of an available neighboring block that is found first when neighboring blocks adjacent to a reference point are scanned in a predetermined scanning order can be set as the affine source vector for the reference point. As an example, the first affine source vector of the combined merge candidate for CP0 may be set equal to the motion vector of the available neighbor block that is found first when the neighbor blocks are found in the order B2, B3 and A2. The second affine source vector of the combined merge candidate for CP1 can be set equal to the motion vector of the available neighbor block that is found first when the neighbor blocks are found in the order B1 and B0. The third affine source vector of the combined merge candidate for CP2 can be set equal to the motion vector of the available neighbor block that is found first when the neighbor blocks are found in the order A1 and A0.

[0294] When a combined merge candidate is obtained, a neighboring block encoded by the affine motion model or a neighboring block obtained as an affine merge candidate may be set as unavailable.

[0295] FIG. 23 is a schematic diagram showing an inaccessible neighbor block.

[0296] When an affine merge candidate is obtained from a neighboring block B1 located above the current block, it may be determined that the neighboring block B1 is not available to obtain the combined merge candidate. Thus, when the affine source vector of the combined fusion candidate for CP1 is obtained, the motion vector B1 may not be used.

[0297] Alternatively, when a neighboring block is encoded in an affine motion model, a combined merge candidate can be obtained based on the affine vector of the subblock containing the reference sample. As an example, when a neighboring block containing a sample located in B1 is encoded in an affine motion model, a combined merge candidate can be obtained using the affine subblock vector of the neighboring block containing a sample located in B1.

[0298] Alternatively, the scan order may be determined based on whether the adjacent block is encoded using an affine motion model or whether the adjacent block is obtained as an affine merge candidate. As an example, a neighboring block encoded with an affine motion model or a neighboring block encoded with an affine merge candidate may be set equal to the last scanned one. As an example, when a neighboring block located in B1 is encoded using an affine motion model, the second affine source vector of the combined merge candidate for CP2 can be obtained by detecting neighboring blocks in the order of B0 and B1.

[0299] When neighboring blocks have different reference picture indices, the motion vectors may be scaled with respect to the reference picture with the largest index or the reference picture with the smallest index. Scaling may be performed based on a picture sequence number (PON) difference value between the reference picture and the current picture.

[0300] Alternatively, neighboring block combining can be performed using only neighboring blocks having the same reference image index from the neighboring blocks. As an example, when at least one of the neighboring blocks has a different reference picture index, the combine may be set to unavailable as a combined merge candidate. In addition, combination availability can be determined regardless of whether adjacent blocks have the same motion information or motion vector. As an example, a combined fusion candidate can be obtained by combining neighboring blocks having the same reference image index, even when the neighboring blocks have the same motion vector (eg, affine source vector CP0 and affine source vector CP1). Alternatively, the scanning order of neighboring blocks may be determined taking into account the reference image indices of the neighboring blocks. Alternatively, combining can be performed using only adjacent blocks having the same prediction direction from adjacent blocks.

[0301] A merge candidate may be generated in a predetermined combining order. As an example, the combination order can be set as follows.

[0302] 1. {affine source vector CP0, affine source vector CP1, affine source vector CP2}

[0303] 2. {affine source vector CP0, affine source vector CP1, affine source vector CP3}

[0304] 3. {affine source vector CP0, affine source vector CP2, affine source vector CP3}

[0305] 4. {affine source vector CP1, affine source vector CP2, affine source vector CP3}

[0306] 5. {affine source vector CP0, affine source vector CP1}

[0307] 6. {affine source vector CP0, affine source vector CP2}

[0308] Although six combination examples are shown, fewer or more combination examples can also be used.

[0309] A combined fusion candidate generated by combining three affine source vectors may be referred to as a 6-parameter combined fusion candidate, and a combined fusion candidate generated by combining two affine source vectors may be referred to as a 4-parameter combined fusion candidate.

[0310] The combining order may be preset in the encoder and decoder. Alternatively, the combining order may be determined based on at least one of the size, shape, partition shape or affine motion model of the current block, the location of the current block in a code tree element, or the POC of the reference image. As an example, when a 4-parameter affine motion model is applied to the current block, the setting may be made such that combination examples for a 4-parameter combined merge candidate have higher priority than combination examples for a 6-parameter combined merge candidate .

[0311] Although the combined merge candidate is generated in the order of combining, only the merge candidate generated by combining adjacent blocks having the same reference image can be determined to be available. Alternatively, when at least one of the neighboring blocks has a different reference image, the motion vectors may be scaled with respect to the highest index reference image or the lowest index reference image to obtain a merge candidate. Scaling can be performed based on the difference in POC between the reference image and the current image. Alternatively, when two of the adjacent blocks have the same reference image and the other adjacent block has a different reference image, the motion vector of the adjacent block having a different reference image can be scaled with respect to the reference image applied in common to the two adjacent blocks to obtain a combined candidate for the merger.

[0312] The number of combined merger candidates added to the merger candidate list may be determined based on at least one of the maximum number of merger candidates or the number of merger candidates previously included in the merger candidate list. Alternatively, depending on the affine motion model of the current block, only 6-parameter combined merge candidates may be added to the merge candidate list, or only 4-parameter combined merge candidates may be added to the combined merge candidate list.

[0313] As an example, when the number of combined merge candidates to be added to the merge candidate list is one, and a 6-parameter motion model is applied to the current block, one 6-parameter combined merge candidate can be added to the candidate list for merger in a pre-established combination order. In detail, when all of the affine source vector CP0, affine source vector CP1 and affine source vector CP2 are available, the combined fusion candidate {affine source vector CP0, affine source vector CP1 and affine source vector CP2} can be added to the list of fusion candidates. Alternatively, when affine seed vector CP2 is not available and affine seed vector CP3 is available, a combined fusion candidate {affine seed vector CP0, affine seed vector CP1, and affine seed vector CP3} may be added to the list of fusion candidates.

[0314] Alternatively, when the number of combined merge candidates to be added to the merge candidate list is one and a 6-parameter motion model is applied to the current block, one 4-parameter combined merge candidate may be added to the merge candidate list in a predetermined combination order.

[0315] As another example, the number of combined merge candidates to be added to the merge candidate list may be set differently depending on the affine motion pattern of the current block. As an example, when a 6-parameter motion model is applied to the current block, the combined merge candidate can be added to the list of merge candidates. On the other hand, when a 4-parameter motion model is applied to the current block, two combined merge candidates can be added to the list of merge candidates.

[0316] The combination fusion candidate to be added to the fusion candidate list may be selected depending on the number of available affinity fusion candidates. As an example, when the number of available affine fusion candidates is two or more, only 4-parameter combined fusion candidates can be added to the fusion candidate list. On the other hand, when the number of available affine fusion candidates is one or less, only 6-parameter combination fusion candidates can be added to the fusion candidate list. Alternatively, when the number of available affine fusion candidates is one or less, N 6-parameter combined fusion candidates and M 4-parameter combined fusion candidates can be added to the list of fusion candidates. In this case, N and M are natural numbers, and M can be obtained based on the maximum number of merge candidates minus N. Alternatively, when the number of available affine merge candidates is one or less, only 4-parameter combined merge candidates can be added to the list of merger candidates.

[0317] Alternatively, the order of combination may be determined based on the availability of affine source vectors. As an example, combined fusion candidates can be added to the fusion candidate list based on the availability of affine source vectors in the following order.

[0318] 1. A case in which the affine source vector CP0, the affine source vector CP1 and the affine source vector CP2 are available.

[0319] {affine source vector CP0, affine source vector CP1, affine source vector CP2}

[0320] 2. A case in which the affine source vector CP0, the affine source vector CP1 and the affine source vector CP3 are available.

[0321] {affine source vector CP0, affine source vector CP1, affine source vector CP3}

[0322] 3. A case in which the affine source vector CP0, the affine source vector CP2 and the affine source vector CP3 are available.

[0323] {affine source vector CP0, affine source vector CP2, affine source vector CP3}

[0324] 4. The case in which the affine source vector CP0, the affine source vector CP2 and the affine source vector CP3 are available.

[0325] {affine source vector CP1, affine source vector CP2, affine source vector CP3}

[0326] 5. The case in which the affine source vector CP0 and the affine source vector CP1 are available.

[0327] {affine source vector CP0, affine source vector CP1}

[0328] 6. The case in which the affine source vector CP0 and the affine source vector CP2 are available.

[0329] {affine source vector CP0, affine source vector CP2}

[0330] When the number of combined merge candidates to be added to the merge candidate list is one, the combined merge candidate corresponding to a condition that is satisfied by the first of the first to sixth conditions may be added to the merge candidate list. When none of the first to sixth candidates satisfy the condition, the combined merge candidate may not be added to the list of merge candidates.

[0331] As another example, the maximum number of combined fusion candidates to be added to the fusion candidate list may be determined depending on the number of available affinity fusion candidates.

[0332] When the number of merge candidates included in the merge candidate list is less than the maximum number, a zero merge candidate that has a motion vector of 0 may be added to the merge candidate list. Thus, a merge candidate can be obtained in affine merge mode in the following order.

[0333] 1. Obtaining an affine merge candidate.

[0334] 2. Combined merger candidate.

[0335] 3. Null merge candidate.

[0336] A prediction method performed between images using translational motion information will next be described in detail below.

[0337] Motion information of the current block can be obtained from motion information of a block other than the current block. In this case, the other block may be a block that is encoded or decoded by inter-picture prediction earlier than the current block. Setting the motion information of the current block to be the same as the motion information of another block can be set as a merge mode. In addition, setting the motion vector of another block as the motion vector prediction value of the current block can be set as a motion vector prediction mode.

[0338] FIG. 24 is a flowchart of a process for obtaining movement information of a current block in a merge mode.

[0339] A merge candidate of the current block may be obtained (S2401). The merge candidate of the current block may be obtained from a block that is encoded or decoded by inter-picture prediction earlier than the current block.

[0340] Candidate blocks used to obtain a merge block may include neighboring blocks containing a sample adjacent to the current block. As an example, when the coordinates of the top left sample of the current block are (0, 0), at least one of the block containing the reference sample located at (-1, H-1) of the block containing the reference sample located at (W -1, -1), a block containing a reference sample located at (W, -1), a block containing a reference sample located at (-1, H), or a block containing a reference sample located at (-1, -1), can be used as a candidate block.

[0341] Alternatively, by candidates, a block that does not belong to the same code tree element as the current block may be set as unavailable as a merge candidate. As an example, when the reference sample deviates from the top edge of the code tree element to which the current block belongs, the candidate block containing the reference sample may be set to unavailable as a merge candidate.

[0342] A merge candidate may be obtained from a temporary neighboring block included in an image other than the current block's image. As an example, a fusion candidate may be obtained from a fusion block included in a fusion image. One reference picture included in the reference picture list can be set as a fusion picture. Index information for identifying a fusion image from the reference images may be carried in a bit stream. Alternatively, a reference image having a predetermined index of the reference images may be determined as a fusion image.

[0343] The movement information of the merge candidate may be set equal to the movement information of the candidate block. As an example, at least one of a motion vector, a reference image index, a prediction direction, or a bidirectional weight index of a candidate block may be set as motion information of a fusion candidate.

[0344] A merge candidate list may be generated (S2402) containing the merge candidates.

[0345] The merge candidates from the list of merge candidates may be assigned indices in a predetermined order. As an example, indices may be sequentially assigned to a merge candidate obtained from a left neighbor block, a merge candidate obtained from a top neighbor block, a merge candidate obtained from a top right neighbor block, a merge candidate obtained from a bottom left neighbor block, a merge candidate obtained from the upper left neighbor block, and a merge candidate obtained from the temporary neighbor block.

[0346] When a plurality of merge candidates are included in the merge candidate, at least one of the plurality of merge candidates may be selected (S2403). Details of information to indicate one of a plurality of merge candidates may be carried in a bit stream. As an example, merge_idx information indicating the index of one of the merge candidates included in the list of merge candidates may be transmitted in a bit stream.

[0347] When the number of merge candidates included in the merge candidate list is less than a threshold value, the merge candidate included in the inter-site traffic information table may be added to the merge candidate list. In this case, the threshold value may be the maximum number of merge candidates that the merge candidate can contain, or a value obtained by subtracting the offset from the maximum number of merge candidates. The offset may be a natural number such as 1 or 2. The inter-hop traffic information table may contain a merge candidate derived from a block that is encoded or decoded before the current block.

[0348] The inter-section motion information table contains a merge candidate obtained from a block that is encoded or decoded based on the inter-picture prediction in the current picture. As an example, the motion information of a merge candidate included in the inter-section motion information table may be set equal to the motion information of a block encoded or decoded based on inter-picture prediction. In this case, the motion information may comprise at least one of a motion vector, a reference image index, a prediction direction, or a bidirectional weight index.

[0349] For convenience of description, a merge candidate included in the inter-site traffic information table will be referred to as an inter-site merge candidate.

[0350] The maximum number of merge candidates that can be included in the inter-hop traffic information table can be preset in the encoder and decoder. As an example, the maximum number of merge candidates that can be included in the inter-site traffic information table may be 1, 2, 3, 4, 5, 6, 7, 8 or more (eg, 16).

[0351] Alternatively, information indicating the maximum number of candidates for merging the inter-section traffic information table may be transmitted in the bitstream. Information can be transferred at the sequence, image or slice level.

[0352] Alternatively, the maximum number of candidates for merging the inter-section motion information table may be determined depending on the image size, slice size, or code tree element size.

[0353] The inter-section motion information table may be initialized in picture elements, slices, packets, code tree elements, or lines of code tree elements (rows and columns). As an example, when a slice is initialized, the inter-slice traffic information table is also initialized and may not contain a merge candidate.

[0354] Alternatively, information indicating whether the inter-section traffic information table should be initialized may be transmitted in the bit stream. Information can be transferred at the slice, tile, packet or block level. The preconfigured inter-section traffic information table may be used until the information indicates initialization of the inter-section traffic information table.

[0355] Alternatively, information regarding the initial inter-slice fusion candidate may be conveyed by an image parameter set or slice header. Even if the slice is initialized, the inter-partition traffic information table may contain an initial inter-partition merge candidate. Accordingly, an inter-slice merge candidate may be used for the block that is to be encoded or decoded first in the slice.

[0356] Blocks may be encoded or decoded in encoding or decoding order, and blocks encoded or decoded based on prediction performed between pictures can be sequentially set as fusion candidates between sections in encoding or decoding order.

[0357] FIG. 25 is a schematic diagram for describing an aspect of updating an inter-section traffic information table.

[0358] When inter-picture prediction is performed (S2501) on the current block, an inter-patch fusion candidate can be obtained (S2502) based on the current block. The movement information of a merge candidate between blocks may be set equal to the movement information of the current block.

[0359] When the inter-section traffic information table is empty (S2503), the inter-section merging candidate obtained based on the current block may be added (S2504) to the inter-section traffic information table.

[0360] When the inter-section traffic information table already contains an inter-section merging candidate (S2503), redundancy check may be performed on the movement information of the current block (or the inter-section merging candidate obtained based on the traffic information) (S2505 ). The redundancy check is designed to determine whether the movement information of the current block matches the movement information of an inter-section merger candidate previously stored in the inter-section movement information table. Redundancy checking may be performed on all inter-site merging candidates previously stored in the inter-site traffic information table. Alternatively, redundancy checking may be performed on inter-site merge candidates with indices greater than or equal to a threshold value, or indices less than or equal to a threshold value, among inter-site merge candidates previously stored in the inter-site traffic information table.

[0361] When an inter-picture prediction fusion candidate having the same motion information as the current block is not included, an inter-section fusion candidate obtained based on the current block may be added (S2508) to the motion information table between sites. Whether the inter-image prediction fusion candidates are identical can be determined based on whether the inter-image prediction fusion candidates have the same motion information (eg, motion vector and/or reference image index).

[0362] In this case, when the maximum number of inter-area merging candidates is previously stored in the inter-area traffic information table (S2506), the oldest inter-area merging candidate is deleted (S2507), and the inter-area merging candidate obtained based on current block may be added (S2508) to the inter-section traffic information table.

[0363] Merger candidates between regions can be identified by their indices. When an inter-site merge candidate obtained from the current block is added to the inter-site traffic information table, the smallest index (eg, 0) is assigned to the inter-site merge candidate, and the indices of previously stored inter-site merge candidates may be incremented by one. In this case, when the maximum number of inter-image prediction fusion candidates is previously stored in the inter-site motion information table, the inter-site fusion candidate with the largest index is removed.

[0364] Alternatively, when an inter-site merge candidate obtained from the current block is added to the inter-site traffic information table, the inter-site merge candidate may be assigned the highest index. As an example, when the number of inter-picture prediction fusion candidates previously stored in the inter-section motion information table is less than the maximum value, an index having the same value as the number of pre-stored inter-picture prediction fusion candidates is can be assigned to a cross-site merge candidate. Alternatively, when the number of prediction fusion candidates carried out between images previously stored in the inter-section motion information table is equal to the maximum value, the inter-section fusion candidate may be assigned an index having a maximum value of minus 1. In addition, the inter-section fusion candidate may be assigned an index having a maximum value of minus 1. In addition, the inter-section fusion candidate The sites with the smallest index are deleted, and the indices of other previously stored merge candidates between sites are decreased by 1.

[0365] FIG. 26 is a schematic diagram showing an aspect of updating a cross-site merge candidate table.

[0366] Since the inter-site merge candidate obtained from the current block is added to the inter-site merge candidate table, it is assumed that the largest index is assigned to the inter-site merge candidate. It is also assumed that the maximum number of cross-site merge candidates is stored in the cross-site merge candidate table.

[0367] When an inter-site merge candidate HmvpCand[n+1] obtained from the current block is added to the inter-site merge candidate table HmvpCandList, the inter-site merge candidate HmvpCand[0] that has the smallest index of the previously stored candidates for inter-chunk merge can be removed, and the indices of other inter-chunk merge candidates can be reduced by 1. Additionally, the index of the inter-chunk merge candidate HmvpCand[n+1], which is obtained from the current block, can be set as the maximum values (n in the example shown in Fig. 26).

[0368] When an inter-section fusion candidate identical to the inter-section fusion candidate obtained based on the current block is previously stored (S2505), the inter-section fusion candidate obtained based on the current block may not be added (S2509) to the information table about movement between sections.

[0369] Alternatively, while the inter-site fusion candidate obtained based on the current block is added to the inter-site traffic information table, a previously stored inter-site fusion candidate identical to the inter-site fusion candidate may be deleted. This has the same effect as updating the indexes of pre-stored cross-site merge candidates.

[0370] FIG. 27 is a schematic diagram showing an example in which the index of a previously stored inter-site merge candidate is updated.

[0371] When the index of a previously stored inter-picture prediction fusion candidate that is identical to the mvCand inter-section fusion candidate that is obtained based on the current block is equal to hIdx, the previously stored inter-picture prediction fusion candidate can be deleted. and inter-picture prediction fusion candidate indices greater than hIdx may be reduced by 1. As an example, FIG. 27 shows that HmvpCand[2], which is identical to mvCand, is removed from the inter-section motion information table HvmpCandList, and that the indices HmvpCand[3] - HmvpCand[n] are reduced by 1.

[0372] In addition, an inter-site fusion candidate mvCand that is obtained based on the current block may be added to the end of the inter-site traffic information table.

[0373] Alternatively, an index assigned to a previously stored inter-partition merge candidate that is identical to the inter-partition merge candidate obtained based on the current block may be updated. For example, the index of a previously stored inter-site merge candidate may be changed to a minimum value or a maximum value.

[0374] The setting may be made such that movement information of blocks included in a predetermined section is not added to the inter-section movement information table. As an example, an inter-section merging candidate obtained based on movement information of a block included in a merging processing section may not be added to the inter-section movement information table. Since the encoding or decoding order is not specified for the blocks included in the merging processing portion, it is inappropriate to use the motion information of one of the blocks to perform inter-picture prediction about other blocks. Accordingly, inter-site merge candidates obtained based on blocks included in the merge processing portion may not be added to the inter-site traffic information table.

[0375] When motion-compensated prediction is performed on subblock elements, a cross-section fusion candidate can be obtained based on the motion information of a representative subblock from a plurality of subblocks included in the current block. As an example, when a sub-block merge candidate is used for the current block, the inter-section merge candidate can be obtained based on the motion information of a representative sub-block of the sub-blocks.

[0376] The motion vectors of the sub-blocks can be obtained in the following order. First, one merge candidate may be selected from among the merge candidates included in the list of merge candidates of the current block, and an initial shift vector shVector may be obtained based on the motion vector of the selected merge candidate. Then, a shift subblock having a reference sample located at (xColSb, yColSb) can be obtained by adding the initial shift vector to the location (xSb, ySb) of a reference sample (eg, top left sample or middle sample) of each subblock in the encoding block. Equation 20 below gives the equation to obtain the shift subblock.

[0377] [Equation 20]

[0378] Then, the motion vector of the matching block corresponding to the center position of the subblock containing (xColSb, yColSb) can be set as the motion vector of the subblock containing (xSb, ySb).

[0379] A representative subblock may refer to a subblock containing the top left sample or center sample of the current block.

[0380] FIG. 28 is a schematic diagram showing the location of a representative subblock.

[0381] FIG. 28A shows an example in which a subblock located to the left of and above the current block is set as a representative subblock, and FIG. 28B shows an example in which a subblock located in the center of the current block is set as a representative subblock . When motion-compensated prediction is performed on subblock elements, a candidate for merging between portions of the current block can be obtained based on the motion vector of the subblock containing the top left sample of the current block or the subblock containing the center sample of the current block.

[0382] Whether to use the current block as an inter-partition fusion candidate can also be determined based on the inter-picture prediction mode of the current block. As an example, a block that is encoded or decoded based on an affine motion model may be set as unavailable as a cross-patch fusion candidate. Accordingly, although the current block is encoded or decoded by inter-picture prediction, the inter-picture prediction motion information table may not be updated based on the current block when the inter-picture prediction mode of the current block is an affine prediction mode.

[0383] Alternatively, a cross-section fusion candidate may be obtained based on a subblock vector of at least one subblock included in the block that is encoded or decoded based on the affine motion model. As an example, a cross-section merge candidate may be obtained using a subblock located in the upper left corner of the current block, a subblock located in the center of the current block, or a subblock located in the upper right corner of the current block. Alternatively, the average of the subblock vectors of the plurality of subblocks may be set as the motion vector of a cross-section fusion candidate.

[0384] Alternatively, a cross-section fusion candidate may be obtained based on the average of the affine source vectors of a block that is encoded or decoded based on the affine motion model. As an example, the average value of at least one of the first affine seed vector, the second affine seed vector, or the third affine seed vector of the current block may be set as the motion vector of the cross-section fusion candidate.

[0385] Alternatively, an inter-plot motion information table may be configured for each inter-picture prediction mode. As an example, at least one of an inter-lot motion information table for a block encoded or decoded by intra-block copying, an inter-lot motion information table for a block encoded or decoded based on a translational motion model, or an information table may be specified. about the motion between sites for a block encoded or decoded based on an affine motion model. One of the plurality of inter-section motion information tables may be selected according to the inter-picture prediction mode of the current block.

[0386] FIG. 29 shows an example in which a table of inter-section motion information is generated for each inter-picture prediction mode.

[0387] When a block is encoded or decoded based on a non-affine motion model, an inter-site fusion candidate mvCand that is obtained based on the block may be added to the non-affine inter-site motion information table HmvpCandList. On the other hand, when a block is encoded or decoded based on the affine motion model, the inter-site fusion candidate mvAfCand that is obtained based on the block may be added to the inter-site affine motion information table HmvpAfCandList.

[0388] The affine source vectors of the block may be stored in a cross-section fusion candidate derived from the block that is encoded or decoded based on the affine motion model. Accordingly, the cross-section merge candidate can be used as a merge candidate to obtain the affine source vector of the current block.

[0389] In addition to the inter-section traffic information table described above, an additional inter-section traffic information table may be specified. In addition to the above-described inter-section traffic information table (hereinafter referred to as the first inter-section traffic information table), a long-term traffic information table (hereinafter referred to as the second inter-section traffic information table) may be specified. In this case, the long-term traffic information table contains long-term merge candidates.

[0390] When both the first inter-site traffic information table and the second inter-site traffic information table are empty, inter-site merging candidates may preferably be added to the second inter-site traffic information table. Until the number of inter-site fusion candidates allowed for the second inter-site traffic information table reaches the maximum number, the inter-site fusion candidates may not be added to the first inter-site traffic information table.

[0391] Alternatively, one inter-image prediction fusion candidate may be added to both the first inter-site motion information table and the second inter-site motion information table.

[0392] In this case, the second inter-site traffic information table where the configuration has been made may no longer be updated. Alternatively, when the decoded slice is greater than or equal to the predetermined slice proportion, the second inter-slice motion information table may be updated. Alternatively, the second inter-site traffic information table may be updated every N code tree element lines.

[0393] On the other hand, the first inter-section motion information table may be updated whenever a block encoded or decoded by inter-picture prediction is generated. However, the setting may be made such that inter-site fusion candidates added to the second inter-site traffic information table are not used to update the first inter-site traffic information table.

[0394] Information for selecting one of the first inter-section traffic information table or the second inter-section traffic information table may be transmitted in a bit stream. When the number of merger candidates included in the merger candidate list is less than a threshold value, the merger candidates included in the inter-site traffic information table specified by the information may be added to the merger candidate list.

[0395] Alternatively, the inter-plot motion information table may be selected based on the size, shape or inter-picture prediction mode, the current block, the presence of bidirectional prediction, the presence of motion vector refinement, or the presence of a triangular partition.

[0396] Alternatively, when the number of merger candidates included in the merger candidate list is less than the maximum number associated with the merger, even if inter-site merger candidates included in the first inter-site traffic information table are added to the candidate list for merging, inter-site merging candidates included in the second inter-site traffic information table may be added to the merging candidate list.

[0397] FIG. 30 is a schematic diagram showing an example in which a cross-section merging candidate included in the long-term traffic information table is added to the merging candidate list.

[0398] When the number of merge candidates included in the merge candidate list is less than the maximum number, inter-site merge candidates included in the first inter-site traffic information table HmvpCandList may be added to the merge candidate list. When the number of fusion candidates included in the fusion candidate list is less than the maximum number, even if the inter-site fusion candidates included in the first inter-site traffic information table are added to the fusion candidate list, the inter-site fusion candidates included in long-term traffic information table HmvpLTCandList can be added to the list of merge candidates.

[0399] Table 1 shows the process of adding cross-section merge candidates included in the long-term traffic information table to the merge candidate list.

[0401] The setting may be made such that the cross-section merging candidate contains additional information as well as motion information. As an example, the size, shape, or block division information may be further stored for a cross-patch merge candidate. When configuring the fusion candidate list for the current block, only inter-image prediction fusion candidates having the same or similar size, shape, or partitioning information as the current block may be used from among the inter-image prediction merging candidates, or Prediction fusion candidates performed between images having the same or similar size, shape, or partition information as the current block may preferably be added to the fusion candidate list.

[0402] Alternatively, a table of inter-section movement information may be generated for each size, shape, or piece of block division information. A list of merge candidates of the current block may be generated using an inter-lot motion information table corresponding to the shape, size, or partition information of the current block from a plurality of inter-lot motion information tables.

[0403] When the number of merge candidates included in the merge candidate list of the current block is less than a threshold value, the inter-site merge candidate included in the inter-site traffic information table may be added to the merge candidate list. The adding process can be performed in ascending or descending order based on indices. As an example, a cross-region merge candidate with the highest index may be preferably added to the merge candidate list.

[0404] When an inter-site merging candidate included in the inter-site traffic information table is to be added to the merging candidate list, a redundancy check can be performed on the inter-site merging candidate relative to the merging candidates previously stored in the list merger candidates.

[0405] As an example, Table 2 shows the process of adding a cross-site merge candidate to the merge candidate list.

[0407] Redundancy checking can only be performed on some inter-site merging candidates included in the inter-site traffic information table. As an example, redundancy checking may be performed only on merge candidates between regions with indices greater than or equal to a threshold value, or indices less than or equal to a threshold value. Alternatively, redundancy checking may be performed only on the N merge candidates having the largest indices or having the smallest indices.

[0408] Alternatively, redundancy checking may be performed only on some of the merge candidates previously stored in the merge candidate list. As an example, redundancy checking may be performed only on a merge candidate with an index greater than or equal to a threshold, or an index less than or equal to the threshold, or a merge candidate obtained from a block at a particular location. In this case, the particular location may include at least one of a left neighbor block, a top neighbor block, a top right neighbor block, or a bottom left neighbor block of the current block.

[0409] FIG. 31 is a schematic diagram showing an example in which redundancy checking is performed on only some merge candidates.

[0410] When the inter-site merge candidate HmvpCand[j] is to be added to the list of merge candidates, a redundancy check may be performed on the inter-site merge candidate regarding the two merge candidates mergeCandList[NumMerge-2] and mergeCandList[ NumMerge-1], having the largest indices. In this case, NumMerge can represent the number of available spatial merge candidates and temporal merge candidates.

[0411] Unlike the example shown, where the inter-site merge candidate HmvpCand[j] is to be added to the list of merge candidates, redundancy checking may be performed on the inter-site merge candidate relative to at least two merge candidates, having the smallest indices. For example, you can check whether mergeCandList[0] and mergeCandList[1] are identical to HmvpCand[j]. Alternatively, redundancy checking may be performed only on the merge candidate received at a particular location. As an example, redundancy checking may be performed on at least one of a merge candidate derived from the nearest block located to the left of the current block or a merge candidate obtained from the nearest block located above the current block. When a merge candidate received at a particular location is not in the merge candidate list, cross-site merge candidates may be added to the merge candidate list without redundancy control.

[0412] When a fusion candidate identical to the first inter-site fusion candidate is found, the redundancy check on the fusion candidate identical to the first inter-site fusion candidate may be skipped while the redundancy check on the second inter-site fusion candidate merging between sections is performed.

[0413] FIG. 32 is a schematic diagram showing an example in which redundancy checks are omitted with respect to a particular merge candidate.

[0414] When the inter-site merge candidate HmvpCand[i] with index i is to be added to the merge candidate list, redundancy checking can be performed between the inter-site merge candidate and the merge candidates previously stored in the merge candidate list. In this case, when a merge candidate mergeCandList[j] is found identical to the inter-site merge candidate HmvpCand[i], the inter-site merge candidate HmvpCand[i] may not be added to the merge candidate list, and redundancy checking may be performed between a fusion candidate between sites HmvpCand[i-1] having index i-1 and fusion candidates. In this case, the redundancy check between the cross-section merge candidate HmvpCand[i-1] and the merge candidate mergeCandList[j] may be skipped.

[0415] For example, in the example shown in Fig. 32, HmvpCand[i] and mergeCandList[2] are determined to be identical to each other. Accordingly, HmvpCand[i] may not be added to the merge candidate list, and redundancy checking may be performed on HmvpCand[i-1]. In this case, the redundancy check between HvmpCand[i-1] and mergeCandList[2] may be skipped.

[0416] When the number of merge candidates included in the current block's merge candidate list is less than a threshold, at least one of the paired merge candidate or the null merge candidate may be further included in addition to the inter-block merge candidate . A pairwise fusion candidate refers to a fusion candidate with a motion vector equal to the average of the motion vectors of two or more fusion candidates, and a null fusion candidate refers to a fusion candidate with a motion vector equal to 0.

[0417] A merge candidate may be added to the merge candidate list of the current block in the following order.

[0418] Spatial Merge Candidate Temporal Merge Candidate Inter-Site Merge Candidate (Affine Inter-Site Merge Candidate) Pairwise Merge Candidate Null Merge Candidate.

[0419] A spatial merge candidate refers to a merge candidate obtained from at least one of an adjacent block or a non-neighboring block, and a temporal merge candidate refers to a merge candidate obtained from a previous reference image. An inter-region affine fusion candidate refers to an inter-region fusion candidate obtained from a block encoded or decoded by an affine motion model.

[0420] The inter-section motion information table can be used even in motion vector prediction mode. As an example, when the number of motion vector prediction candidates included in the motion vector prediction candidate list of the current block is less than a threshold value, an inter-section fusion candidate included in the inter-section motion information table may be set as a motion vector prediction candidate for the current block. In detail, the motion vector of the cross-section fusion candidate can be set as the motion vector prediction candidate.

[0421] When one of the motion vector prediction candidates included in the motion vector prediction candidate list of the current block is selected, the selected candidate may be set as the motion vector predictor of the current block. Subsequently, the motion vector of the current block can be obtained by decoding the residual motion vector value of the current block and then adding the motion vector predictor and the residual motion vector value.

[0422] The motion vector prediction candidate list of the current block may be configured in the following order.

[0423] Spatial motion vector prediction candidate Temporal motion vector prediction candidate - Inter-picture decoding section fusion candidate - (Inter-picture decoding section affine fusion candidate) - Null motion vector prediction candidate.

[0424] A spatial motion vector prediction candidate refers to a motion vector prediction candidate obtained from at least one of an adjacent block or a non-neighboring block, and a temporal motion vector prediction candidate refers to a motion vector prediction candidate obtained from a previous reference picture. An inter-region affine fusion candidate refers to an inter-region motion vector prediction candidate obtained from a block encoded or decoded by an affine motion model. The null motion vector prediction candidate represents a candidate having a motion vector equal to 0.

[0425] A coding block may be divided into a plurality of prediction elements, and the prediction elements may be subject to prediction. In this case, the prediction element specifies the base element for performing the prediction.

[0426] A coding block may be divided using at least one of a vertical line, a horizontal line, an oblique line, or a diagonal line. Information for determining at least one of the number, angles, or locations of lines that break up a coding block may be transmitted in the bit stream. As an example, information indicating one of a plurality of encoding block section type candidates may be transmitted in a bitstream, or information indicating one of a plurality of line candidates that divide an encoding block may be transmitted in a bitstream. Alternatively, information for determining the number or types of line candidates dividing a coding block may be carried in the bitstream. As an example, whether a slant line having a larger angle than the diagonal line and/or a slant line having a smaller angle than the diagonal line should be used as a line candidate can be determined using a 1-bit flag.

[0427] Alternatively, at least one of the number, angles, or locations of lines dividing a coding block may be adaptively determined based on at least one of an inter-picture prediction mode, a coding block, an inter-picture prediction mode, a location available a merge candidate or splitting aspect of an adjacent block.

[0428] When a coding block is divided into a plurality of prediction elements, intra or inter-picture prediction may be performed on the prediction elements.

[0429] FIG. 33 is a schematic diagram showing an example of dividing a coding block into a plurality of prediction elements using a diagonal line.

[0430] As in the examples shown in FIGS. 33A and 33B, a coding block may be divided into two triangular prediction elements using a diagonal line.

[0431] FIGS. 33A and 33B show that a coding block is divided into two prediction elements using a diagonal line connecting two vertices of the coding block. However, a coding block may be divided into two prediction elements using an inclined line having at least one end that does not pass through the top of the coding block.

[0432] FIG. 34 is a schematic diagram showing an example of dividing a coding block into two prediction elements.

[0433] As in the examples shown in FIGS. 34A and 34B, a coding block may be divided into two prediction elements using an inclined line, both ends of which are adjacent to the top edge and bottom edge of the coding block.

[0434] Alternatively, as in the examples shown in FIGS. 34C and 34D, a coding block may be divided into two prediction elements using an inclined line, both ends of which are adjacent to the left edge and right edge of the coding block.

[0435] Alternatively, the coding block may be split into two prediction blocks of different sizes. As an example, by setting a slant line dividing a coding block so that it is adjacent to two edges to form one vertex, the coding block can be split into two prediction elements of different sizes.

[0436] FIG. 35 shows examples in which a coding block is divided into a plurality of prediction blocks of different sizes.

[0437] As in the examples shown in FIGS. 35A and 35B, by setting a diagonal line connecting the upper left portion and the lower right portion of the encoding block to pass through the left edge, right edge, top edge, or bottom edge of the encoding block, and without passing through the upper left corner or lower right corner of the coding block, the coding block can be split into two prediction elements of different sizes.

[0438] Alternatively, as in the examples shown in FIGS. 35C and 35D, by setting a diagonal line connecting the top right portion and bottom left portion of the encoding block to pass through the left edge, right edge, top edge, or bottom edge of the encoding block , rather than passing through the upper left corner or lower right corner of the encoding block, the encoding block can be split into two prediction elements of different sizes.

[0439] Each of the prediction elements generated by dividing the coding block is called the "Nth prediction element". As an example, in the examples shown in FIGS. 33 to 35, PU1 may be set as the first prediction element, and PU2 may be set as the second prediction element. The first prediction element may refer to a prediction element containing a sample located in the lower left portion of the encoding block or a sample located in the upper left portion of the encoding block, and the second prediction element may refer to a prediction element containing a sample located in the upper right portion of the block encoding, or count, located at the bottom right of the encoding block.

[0440] Conversely, a prediction element containing a sample located in the upper right part of the encoding block or a sample located in the lower right part of the encoding block may be set as the first prediction element, and a prediction element containing a sample located in the lower left side of the encoding block, or a sample located at the upper left side of the encoding block may be specified as the second prediction element.

[0441] The following embodiments will be described by focusing on an example of division using a diagonal line. Specifically, dividing a coding block into two prediction elements using a diagonal line is called diagonal division or triangular division, and a prediction element generated based on the diagonal division is called a triangular prediction element. However, it should be understood that the following embodiments can be applied even to the example of dividing using an inclined line having an angle different from the angle of a vertical line, a horizontal line or a diagonal line.

[0442] Whether diagonal partitioning should be applied to a coding block may be determined based on at least one of slice type, maximum number of merge candidates, which may include a list of merge candidates, coding block size, coding block shape, mode predictive coding of the coding block or partitioning aspect of the parent node.

[0443] As an example, whether diagonal partitioning should be applied to a coding block can be determined based on whether the current slice is of type B. Diagonal partitioning can be enabled only when the current slice is of type B.

[0444] Alternatively, whether diagonal division should be applied to a coding block may be determined based on whether the maximum number of merge candidates included in the merge candidate list is two or more. A diagonal split may only be permitted when the maximum number of merger candidates included in the merger candidate list is two or more.

[0445] Alternatively, when at least one of the width or height is greater than 64 in the hardware implementation, it is disadvantageous that the processing element having a size of 64x64 is redundantly accessed. Accordingly, when at least one of the width or height of a coding block is larger than a threshold value, it may not be allowed to divide the coding block into a plurality of prediction blocks. As an example, when at least one of the encoding block width or height is greater than 64, the diagonal division may not be used.

[0446] Alternatively, whether diagonal division should be applied to a coding block may be determined based on at least one of whether the number of samples included in the coding block is less than or equal to the first threshold value, or whether the number of samples included in the coding block is less than or equal to the number of samples included in the coding block, the second threshold value. As an example, when the number of samples included in the encoding block is less than or equal to the first threshold value or when the number of samples included in the encoding block is greater than or equal to the second threshold value, the setting may be made such that the diagonal division is not applied to the encoding block . Alternatively, whether diagonal division should be applied to the encoding block may be determined based on whether the width-to-height ratio of the encoding block is smaller than the first threshold value or whether the width-to-height ratio of the encoding block is greater than the second threshold value. In this case, the width-height ratio whRatio of the encoding block can be defined as the ratio between the width CbW and the height CbH of the encoding block, as shown in Equation 21 below.

[0447] [Equation 21]

[0448] The second threshold value may be the inverse of the first threshold value. As an example, when the first threshold value is equal to k, the second threshold value may be equal to l/k.

[0449] Diagonal division can be applied to a coding block only when the width-to-height ratio of the coding block is between the first threshold and the second threshold.

[0450] Alternatively, the triangular partition may be used only when the width-to-height ratio of the coding block is less than the first threshold value or greater than the second threshold value. As an example, when the first threshold value is 16, diagonal division may not be allowed for a 64x4 or 4x64 encoding block.

[0451] Alternatively, whether to allow diagonal splitting may be determined based on the splitting aspect of the parent node. As an example, when an encoding block that represents a parent node is partitioned based on a quaternary tree partition, a diagonal partition may be applied to the encoding block that is a leaf node. On the other hand, when an encoding block that represents a parent node is partitioned based on a binary tree-based partition or a ternary tree-based partition, the setting can be made such that diagonal partitioning is not allowed for the encoding block that is a leaf node.

[0452] Alternatively, whether to allow diagonal division may be determined based on the predictive coding mode of the coding block. As an example, diagonal division can be enabled only when the encoding block is encoded using intra-prediction, when the encoding block is encoded using inter-picture prediction, or when the encoding block is encoded in a predetermined inter-picture prediction mode. In this case, the predetermined inter-image prediction mode may indicate at least one of a fusion mode, a motion vector prediction mode, an affine fusion mode, or an affine motion vector prediction mode.

[0453] Alternatively, whether to allow diagonal partitioning may be determined based on the size of the parallel processing area. As an example, when the size of the encoding block is larger than the size of the parallel processing area, diagonal division may not be used.

[0454] Whether or not diagonal division should be applied to a coding block may be determined by considering two or more of the above conditions.

[0455] In addition, information indicating whether to enable diagonal division may be transmitted in the bit stream. Information can be transferred at the sequence, image, slice or block level. The information indicates that diagonal division is allowed, and diagonal division can be applied to the encoding block only when at least one of the above conditions is satisfied.

[0456] When it is determined that a diagonal division should be applied to a coding block, information indicating the number or locations of lines dividing the coding block may be transmitted in the bit stream.

[0457] As an example, when a coding block is divided by a diagonal line, information indicating the direction of the diagonal line dividing the coding block may be transmitted in a bit stream. As an example, the triangle_partition_type_flag, which indicates the direction of the diagonal line, may be carried in the bit stream. The flag indicates whether the coding block is divided by a diagonal line connecting the upper left part and the lower right part, or a diagonal line connecting the upper right part and the lower left part. Dividing a coding block by a diagonal line connecting an upper left portion and a lower right portion may be called a left triangular partition type, and dividing a coding block by a diagonal line connecting an upper right portion and a lower left portion may be called a right triangular partition type. As an example, a flag having a value of 0 may indicate that the encoding block section type is a left triangular section type, and a flag having a value of 1 may indicate that the encoding block section type is a right triangular section type.

[0458] Information indicating the type of section of a coding block may be transmitted at the coding block level. Accordingly, a partition type can be determined for each encoding block to which the diagonal partition is applied.

[0459] As another example, information indicating the partition type for a sequence, image, slice, tile, or code tree element may be transmitted. In this case, the partition type of the encoding blocks to which the diagonal partition is applied can be set identically in a sequence, picture, slice, tile, or code tree element.

[0460] Alternatively, information may be encoded and transmitted to determine the section type of the first encoding element to which the diagonal division is applied in a code tree element, and the setting may be made such that the second and subsequent encoding elements to which the diagonal division is applied use the same section type as the first encoding element.

[0461] As another example, the section type of an encoding block may be determined based on the section type of an adjacent block. In this case, the adjacent block may contain at least one of an adjacent block adjacent to the upper left corner of the encoding block, an adjacent block adjacent to the upper right corner of the encoding block, an adjacent block adjacent to the lower left corner of the encoding block, an adjacent block located above coding block, or an adjacent block located to the left of the coding block. As an example, the section type of the current block can be set to the same as the section type of an adjacent block. Alternatively, the partition type of the current block may be determined based on whether the left triangular partition type applies to the top left adjacent block, or whether the right triangular partition type applies to the top right neighbor block or the bottom left neighbor block.

[0462] To perform motion prediction compensation for the first triangular prediction element and the second triangular prediction element, motion information of each of the first triangular prediction element and the second triangular prediction element can be obtained. In this case, motion information of the first triangular prediction element and the second triangular prediction element can be obtained from the fusion candidates included in the fusion candidate list. To distinguish the general fusion candidate list from the fusion candidate list used to obtain motion information of triangular prediction elements, the fusion candidate list for obtaining motion information of triangular prediction elements is called a triangular fusion candidate list, and the fusion candidate included in the list triangular fusion candidates are called triangular fusion candidates. However, using the above-described method for obtaining a merge candidate and a method for obtaining a list of merge candidates for a method for configuring triangular merge candidates and a list of triangular merge candidates is also included in the spirit of the present invention.

[0463] Information for determining the maximum number of triangular merge candidates that the list of triangular merge candidates can contain may be transmitted in the stream. The information may indicate the difference between the maximum number of merge candidates that the list of merge candidates can contain and the maximum number of triangular merge candidates that the list of triangular merge candidates can contain.

[0464] Triangular merge candidates can be obtained from spatial neighbors and temporal neighbors of a coding block.

[0465] FIG. 36 is a schematic diagram showing adjacent blocks used to obtain a triangular merge candidate.

[0466] A triangular fusion candidate may be obtained using at least one of an adjacent block located above the encoding block, an adjacent block located to the left of the encoding block, or a fusion block included in an image other than the encoding block image. The upper adjacent block may contain at least one of a block containing a sample (xCb+CbW-1, yCb-1) located above the encoding block, a block containing a sample (xCb+CbW, yCb-1) located above the encoding block, and a block containing a sample (xCb-1, yCb-1) located above the coding block. The left adjacent block may contain at least one of a block containing a sample (xCb-1, yCb+CbH-1) located to the left of the encoding block, or a block containing a sample (xCb-1, yCb+CbH) located to the left of coding block. The fusion block can be defined as one of a block containing a sample (xCb+CbW, yCb+CbH) adjacent to the upper right corner of the encoding block in the fusion image, or a block containing a sample (xCb/2, yCb/2) located in the center coding block in the combined image.

[0467] Neighboring blocks may be discovered in a predetermined order, and triangular merge candidates may form a list of triangular merge candidates in a predetermined order. As an example, triangular fusion candidates may be detected in the order B1, A1, B0, A0, C0, B2, and C1 to form a list of triangular fusion candidates.

[0468] Motion information of the triangular prediction elements may be obtained based on the list of triangular fusion candidates. That is, triangular prediction elements can be in the same list of triangular merging candidates.

[0469] Information for indicating at least one of the triangular fusion candidates included in the list of triangular fusion candidates may be transmitted in a bitstream for obtaining motion information of the triangular aurora element. As an example, index information merge_riangle_idx, which is intended to indicate at least one of the triangular merge candidates, may be carried in a bitstream.

[0470] The index information may indicate a combination of a first triangular prediction element fusion candidate and a second triangular prediction element fusion candidate. As an example, Table 3 presents combinations of merge candidates corresponding to the merge_triangle_idx index information.

[0471] [Table 3]

[0472] Index information merge_triangle_idx having a value of 1 represents that the motion information of the first triangular prediction element is obtained from the merge candidate at index 1, and that the motion information of the second triangular prediction element is received from the merge candidate at index 0. Using the merge_triangle_idx index information, a triangular merge candidate for obtaining motion information of the first triangular prediction element and a triangular merge candidate for obtaining motion information of the second triangular prediction element can be obtained.

[0473] The type of coding block section to which the diagonal division is applied can be determined using the index information. That is, the index information may indicate a combination of a first triangular prediction element fusion candidate, a second triangular prediction element fusion candidate, and a coding block division direction. When the partition type of an encoding block is determined using the index information, the triangle_partition_type_flag information, which indicates the direction of the diagonal line dividing the encoding block, may not be encoded. Table 4 represents the encoding block partition type in the merge_triangle_idx index information.

[0474] [Table 4]

[0475] A TriangleDir variable equal to 0 represents that the left triangular partition type is applied to the encoding block, and a TriangleDir variable equal to 1 represents that the right triangular partition type is applied to the encoding block. In the combination of Table 3 and Table 4, the index information merge_triangle_idx can be set to indicate a combination of a first triangular prediction element merge candidate, a second triangular prediction element merge candidate, and a coding block split direction.

[0476] As another example, only index information for one of the first triangular prediction element and the second triangular prediction element may be transmitted, and the triangular fusion candidate index for the other of the first triangular prediction element and the second triangular prediction element may be determined based on information about indexes. As an example, a triangular merging candidate of the first triangular prediction element may be determined based on the index information merge_triangle_idx, which indicates the index of one of the triangular merging candidates. In addition, a triangular merging candidate of the second triangular prediction element may be specified based on merge_triangle_idx. As an example, a triangular merge candidate of a second triangular prediction element can be obtained by adding an offset to or subtracting from the merge_triangle_idx index information. The offset may be an integer such as 1 or 2. As an example, a triangular merge candidate with an index equal to merge_traingle_idx plus 1 may be defined as a triangular merge candidate of the second triangular prediction element. When merge_traingle_idx indicates the triangular merge candidate with the highest index value of the triangular merge candidates, the motion information of the second triangular prediction element may be obtained from the triangular merge candidate with an index equal to 0 or the triangular merge candidate with an index equal to merge_traingle_idx minus 1.

[0477] Alternatively, the motion information of the second triangular prediction element may be obtained from a triangular fusion candidate having the same reference image as the triangular fusion candidate of the first triangular prediction element, which is indicated by the index information. In this case, a triangular fusion candidate having the same reference picture as the triangular fusion candidate of the first triangular prediction element may indicate a triangular fusion candidate in which at least one of the L0 reference picture or the L1 reference picture is the same. as with the triangular fusion candidate of the first triangular prediction element. When there are a plurality of triangular fusion candidates having the same reference image as the triangular fusion candidate of a first triangular prediction element, one of the plurality of triangular fusion candidates may be selected based on at least one of whether the fusion candidate contains bidirectional information, or the difference values between the index information and the merger candidate index.

[0478] As another example, index information may be transmitted to each of the first triangular prediction element and the second triangular prediction element. As an example, first index information 1st_merge_idx, which is for determining a triangular merging candidate of the first triangular prediction element, and second index information 2nd_merge_idx, which is for determining a triangular merging candidate of the second triangular prediction element, may be transmitted in a bit stream. The movement information of the first triangular prediction element may be obtained from the triangular merger candidate determined based on the first index information 1st_merge_idx, and the movement information of the second triangular prediction element may be obtained from the triangular merger candidate determined based on the second index information 2nd_merge_idx .

[0479] The first index information 1st_merge_idx may indicate the index of one of the triangular merge candidates included in the triangular merge candidate list. The triangular merging candidate of the first triangular prediction element may be determined to be the triangular merging candidate indicated by the first index information 1st_merge_idx.

[0480] The setting may be made such that the triangular merge candidate indicated by the first index information 1st_mergez_idx cannot be used as the triangular merge candidate of the second triangular prediction element. Accordingly, the second index information 2nd_merge_idx of the second triangular prediction element may indicate the index of one of the remaining triangular merge candidates other than the triangular merge candidate indicated by the first index information. When the value of the second index information 2nd_merge_idx is less than the value of the first index information 1st merge idx, the triangular merge candidate of the second triangular prediction element may be determined to be a triangular merge candidate having the index information indicated by the second index information 2nd_merge_idx. On the other hand, when the value of the second index information 2nd_merge_idx is greater than or equal to the value of the first index information 1st_merge_idx, the triangular merging candidate of the second triangular prediction element may be determined to be a triangular merging candidate with the index having the value obtained by adding units to the value of the second index information 2nd_merge_idx.

[0481] Alternatively, whether to transmit the second index information may be determined according to the number of triangular merge candidates included in the triangular merge candidate list. As an example, when the maximum number of triangular merge candidates that the triangular merge candidate list can contain does not exceed two, transmission of the second index information may be skipped. When the transmission of the second index information is omitted, a second triangular merge candidate can be obtained by adding an offset to or subtracting from the first index information. As an example, when the maximum number of triangular merge candidates that the triangular merge candidate list can contain is two, and the first index information indicates an index equal to 0, the second triangular merge candidate can be obtained by adding one to the first index information indices. Alternatively, when the maximum number of triangular merge candidates that the list of triangular merge candidates can contain is two, and the first index information indicates one, the second triangular merge candidate can be obtained by subtracting one from the first index information.

[0482] Alternatively, when transmission of the second index information is skipped, the second index information may be set as a default value. In this case, the default value may be zero. The second triangular merge candidate can be obtained by comparing the first index information and the second index information. As an example, the merge candidate at index 0 may be set as the second triangular merge candidate when the second index information is less than the first index information, and the merge candidate at index 1 may be set as the second triangular merge candidate at merge when the second index information is greater than or equal to the first index information.

[0483] When the triangular fusion candidate has unidirectional motion information, the unidirectional motion information of the triangular fusion candidate can be set as the motion information of the triangular prediction element. On the other hand, when the triangular fusion candidate has bidirectional motion information, only one of the L0 motion information or the L1 motion information can be set as the motion information of the triangular prediction element. Whether to select the L0 motion information or the L1 motion information can be determined based on the index of the triangular fusion candidate or the motion information of other triangular prediction elements.

[0484] As an example, when the index of the triangular fusion candidate is an even number, the motion information L0 of the triangular prediction element may be set to zero, and the motion information L1 of the triangular fusion candidate may be set to the motion information L1 of the triangular prediction element . On the other hand, when the index of the triangular fusion candidate is an odd number, the motion information L1 of the triangular prediction element may be set to zero, and the motion information L0 of the triangular fusion candidate may be set to zero. Conversely, the motion information L0 of the triangular fusion candidate may be set as the motion information L0 of the triangular prediction element when the index of the triangular fusion candidate is an even number, and the motion information L1 of the triangular fusion candidate may be set as the information about motion L1 of the triangular prediction element when the index of the triangular fusion candidate is an odd number. Alternatively, the motion information L0 of the triangular fusion candidate may be set to the motion information L0 of the first triangular prediction element when the triangular fusion candidate for the first triangular prediction element is an even number, and the motion information L1 of the triangular fusion candidate may be set to as motion information L1 of the second triangular prediction element when the triangular fusion candidate for the second triangular prediction element is an odd number.

[0485] Alternatively, when the first triangular prediction element has L0 motion information, the L0 motion information of the second triangular prediction element may be set to zero, and the L1 motion information of the triangular fusion candidate may be set as the L1 information of the second triangular prediction element. . On the other hand, when the first triangular prediction element has motion information L1, the motion information L1 of the second triangular prediction element can be set to zero, and the motion information L0 of the triangular fusion candidate can be set as the motion information L0 of the second triangular element predictions.

[0486] The setting may be made such that the list of triangular fusion candidates for obtaining motion information of the first triangular prediction element is different from the list of triangular fusion candidates for obtaining motion information of the second triangular prediction element.

[0487] As an example, when a triangular fusion candidate for obtaining motion information of the first triangular prediction element is specified from a list of triangular fusion candidates based on the index information for the first triangular prediction element, the motion information of the second triangular prediction element can be obtained with using a list of triangular merge candidates containing the remaining triangular merge candidates other than the triangular merge candidate specified by the index information. Details of the motion of the second triangular prediction element can be obtained from one of the remaining triangular fusion candidates.

[0488] Accordingly, the maximum number of triangular fusion candidates that can contain the list of triangular fusion candidates of the first triangular prediction element may be different from the maximum number of triangular fusion candidates that can contain the list of triangular fusion candidates of the second triangular prediction element. As an example, when the triangular fusion candidate list of the first triangular prediction element contains M fusion candidates, the triangular fusion candidate list of the second triangular prediction element may contain M-1 fusion candidates other than the triangular fusion candidate indicated by the triangular fusion candidate information. indices of the first triangular prediction element.

[0489] As another example, while triangular prediction element fusion candidates are obtained based on neighboring blocks adjacent to an encoding block, the availability of neighboring blocks may be determined based on the shapes or locations of the triangular prediction elements.

[0490] FIG. 37 is a schematic diagram for describing an example of determining the availability of a neighboring block for each triangular prediction element.

[0491] An adjacent block that is not adjacent to the first triangular prediction element may be set as unavailable to the first triangular prediction element, and an adjacent block that is not adjacent to the second triangular prediction element may be set as unavailable to the second triangular element predictions.

[0492] For example, as in the example shown in FIG. 37A, when a left triangular partition type is applied to a coding block, it can be determined that blocks A1, A0, and A2 adjacent to the first triangular prediction element are from blocks adjacent to coding blocks are available for the first triangular prediction element, and that blocks B0 and B1 are not available for the first triangular prediction element. Accordingly, the triangular merge candidate list for the first triangular prediction element may include triangular merge candidates derived from blocks A1, A0, and A2, but may not contain triangular merge candidates derived from blocks B0 and B1.

[0493] As in the example shown in FIG. 37B, when the left triangular partition type is applied to the encoding block, it can be determined that blocks B0 and B1 adjacent to the second triangular prediction element are available for the second triangular prediction element, and that blocks A1, A0 and A2 are not available to the second triangular prediction element. Accordingly, the triangular merge candidate list for the second triangular prediction element may include triangular merge candidates derived from blocks B0 and B1, but may not contain triangular merge candidates derived from blocks A1, A0, and A2.

[0494] Therefore, the number or range of triangular fusion candidates that a triangular prediction element can use can be determined based on at least one of the location of the triangular prediction element or the type of encoding block partition.

[0495] As another example, the fusion mode may be applied to only one of the first triangular prediction element and the second triangular prediction element. In addition, the motion information of the other of the first triangular prediction element and the second triangular prediction element can be set to be the same as the motion information of the triangular prediction element to which the merging mode is applied, or can be obtained by refining the motion information of the triangular prediction element. to which merge mode is applied.

[0496] As an example, the reference image index and motion vector of the first triangular prediction element may be obtained based on the triangular fusion candidate, and the motion vector of the second triangular prediction element may be obtained by refining the motion vector of the first triangular prediction element. As an example, the motion vector of the second triangular prediction element may be obtained by adding or subtracting the refined motion vector {Rx, Ry} to the motion vector {mvD1LXx, mvD1LXy} of the first triangular prediction element. The reference image index of the second triangular prediction element may be set equal to the reference image index of the first triangular prediction element.

[0497] Information for determining a refined motion vector indicating the difference between the motion vector of the first triangular prediction element and the motion vector of the second triangular prediction element may be transmitted in the bit stream. The information may comprise at least one of information indicating the size of the refined motion vector or information indicating the sign of the refined motion vector.

[0498] Alternatively, the sign of the refined motion vector may be obtained based on at least one of the partition type applied to the encoding block, the location of the triangular prediction element, or the index of the triangular prediction element.

[0499] As another example, a motion vector and a reference image index of one of the first triangular prediction element and the second triangular prediction element may be transmitted. The motion vector of the other of the first triangular prediction element and the second triangular prediction element can be obtained by refining the transmitted motion vector.

[0500] As an example, the reference picture index and motion vector of the first triangular prediction element may be determined based on information transmitted from the bitstream. In addition, the motion vector of the second triangular prediction element can be obtained by refining the motion vector of the first triangular prediction element. As an example, the motion vector of the second triangular prediction element may be obtained by adding or subtracting the refined motion vector {Rx, Ry} to the motion vector {mvD1LXx, mvD1LXy} of the first triangular prediction element. The reference image index of the second triangular prediction element may be set equal to the reference image index of the first triangular prediction element.

[0501] Motion compensated prediction may be performed on the encoding block based on the motion information of the first triangular prediction element and the motion information of the second triangular prediction element. In this case, degradation of image quality may occur at the boundary between the first triangular prediction element and the second triangular prediction element. As an example, the continuity of image quality may deteriorate near an edge present at the boundary between the first triangular prediction element and the second triangular prediction element. To reduce image quality degradation at the edge, a prediction sample can be obtained using a weighted prediction or a smoothing filter.

[0502] The prediction sample in the encoding block to which the diagonal division is applied may be obtained based on a weighted sum operation of the first prediction sample selected based on the motion information of the first triangular prediction element and the second prediction sample selected based on the motion information of the second triangular divination element. Alternatively, the prediction sample of the first triangular prediction element may be obtained from the first prediction block determined based on the motion information of the first triangular prediction element, and the prediction sample of the second triangular prediction element may be obtained from the second prediction block determined based on the motion information of the second triangular element predictions. In this case, the prediction sample located in the boundary portion between the first triangular prediction element and the second triangular prediction element can be obtained based on a weighted sum operation of the first prediction sample included in the first prediction block and the second prediction sample included in the second block. predictions. As an example, Equation 22 below presents an example of obtaining prediction samples of the first triangular prediction element and the second triangular prediction element.

[0503] [Equation 22]

[0504] In Equation 22, P1 represents the first prediction sample, and P2 represents the second prediction sample, w1 represents a weight applied to the first prediction sample, and (1-w1) represents a weight applied to the second prediction sample. As in the example shown in Equation 22, the weight applied to the second prediction sample can be obtained by subtracting the weight applied to the first prediction sample from a constant value.

[0505] When the left triangular partition type is applied to a coding block, the boundary portion may contain prediction samples having the same x-coordinate and the same y-coordinate. On the other hand, when a right triangular partition type is applied to a coding block, the boundary portion may contain prediction samples each having an x-coordinate and a y-coordinate whose sum is greater than or equal to the first threshold and less than or equal to the second threshold.

[0506] The size of the boundary region may be determined based on at least one of a coding block size, a coding block shape, triangular prediction element motion information, a triangular prediction element motion vector difference value, a POC of a reference image, or a difference value between the first sample prediction and a second prediction count in the diagonal boundary.

[0507] FIGS. 38 and 39 are schematic diagrams showing an example of obtaining a prediction sample based on a weighted sum operation of the first prediction sample and the second prediction sample. FIG. 38 illustrates a case in which the left triangular section type is applied to the encoding block, and FIG. 39 illustrates the case in which the right triangular section type is applied to the encoding block. In addition, FIGS. 38A and 39A are schematic views showing a prediction aspect for a luma component, and FIGS. 38B and 39B are schematic views showing a prediction aspect for a chrominance component.

[0508] The drawings show that a number recorded in a prediction sample located adjacent to the boundary between the first prediction element and the second prediction element indicates a weight applied to the first prediction sample. As an example, when the number recorded in the prediction sample is N, the prediction sample can be obtained by applying a weight of N/8 to the first prediction sample and applying a weight of 1-(N/8) to the second prediction sample.

[0509] In the non-boundary region, the first prediction sample or the second prediction sample may be determined to be the prediction sample. As shown in the example of FIG. 38, in the portion belonging to the first triangular prediction element, of the portions where the absolute difference between the x-coordinate and the y-coordinate is greater than the threshold value, the first prediction count obtained based on the motion information of the first triangular element prediction, can be set as a prediction reference. On the other hand, in the portion belonging to the second triangular prediction element, from the portions where the difference between the x coordinate and the y coordinate is larger than the threshold value, a second prediction sample obtained based on the motion information of the second triangular prediction element can be determined in as a prediction reference.

[0510] As shown in the example of FIG. 39, in a portion where the sum of the x coordinate and the y coordinate is less than the first threshold value, the first prediction sample obtained based on the motion information of the first triangular prediction element may be set as the sample predictions. On the other hand, in a portion where the sum of the x coordinate and the y coordinate is greater than the second threshold value, the second prediction sample obtained based on the motion information of the second triangular prediction element may be set as the prediction sample.

[0511] The threshold value for determining a non-edge region may be determined based on at least one of a coding block size, a coding block shape, or a color component. As an example, when the threshold value for the luma component can be set to N, the threshold value for the chrominance component can be set to N/2.

[0512] The prediction samples included in the boundary portion may be obtained based on a weighted sum operation of the first prediction sample and the second prediction sample. In this case, the weight applied to the first prediction sample and the second prediction sample may be determined based on at least one of a location of the prediction sample, a coding block size, a coding block shape, or a color component.

[0513] For example, as in the example shown in FIG. 38A, prediction samples located at the same x-coordinate and the same y-coordinate can be obtained by applying the same weight to the first prediction sample and the second prediction sample. Prediction samples in which the absolute difference between the x coordinate and the y coordinate is one can be obtained by setting the ratio between the weights applied to the first prediction sample and the second prediction sample to (3:1) or (1:3). In addition, prediction samples in which the absolute difference between the x-coordinate and the y-coordinate is two can be obtained by setting the ratio between the weights applied to the first prediction sample and the second prediction sample to (7:1) or (1:7 ).

[0514] Alternatively, as in the example shown in FIG. 38B, prediction samples located at the same x-coordinate and the same y-coordinate can be obtained by applying the same weight to the first prediction sample and the second prediction sample, and the prediction samples in which the absolute difference between the x-coordinate and the y-coordinate is equal to one, can be obtained by setting the ratio between the weights applied to the first prediction sample and the second prediction sample to be (7:1) or (1:7).

[0515] For example, as in the example shown in FIG. 39A, prediction samples in which the sum of the x-coordinate and the y-coordinate are one less than the width or height of the encoding block can be obtained by applying the same weight to the first prediction sample and the second prediction count. Prediction samples in which the sum of the x-coordinate and the y-coordinate are equal to or two less than the width or height of the encoding block can be obtained by setting the ratio between the weights applied to the first prediction sample and the second prediction sample to (3:1) or (1:3). Prediction samples in which the sum of the x-coordinate and the y-coordinate are one more or three less than the width or height of the encoding block can be obtained by setting the ratio between the weights applied to the first prediction sample and the second prediction sample to (7: 1) or (1:7).

[0516] Alternatively, as in the example shown in FIG. 39B, prediction samples in which the sum of the x-coordinate and y-coordinate are one less than the width or height of the encoding block can be obtained by applying the same weight to the first prediction sample and the second prediction count. Prediction samples in which the sum of the x-coordinate and the y-coordinate are equal to or two less than the width or height of the encoding block can be obtained by setting the ratio between the weights applied to the first prediction sample and the second prediction sample to (7:1) or (1:7).

[0517] As another example, the weighting factor may be determined taking into account the location of the prediction sample or the shape of the coding block. Equations 23-25 show examples of obtaining a weight coefficient when the left triangular section type is applied to a coding block. Equation 23 shows an example of obtaining a weight applied to the first prediction sample when the coding block is square.

[0518] [Equation 23]

[0519] In Equation 23, x and y represent the location of the prediction reference. When the coding block is non-square, the weight applied to the first prediction sample can be obtained using Equation 24 or Equation 25 below. Equation 24 represents a case in which the width of a coding block is larger than the height, and Equation 25 represents a case in which the width of a coding block is smaller than the height.

[0520] [Equation 24]

[0521] [Equation 25]

[0522] When the right triangular partition type is applied to a coding block, the weight applied to the first prediction sample can be determined using Equations 26-28. Equation 26 shows an example of obtaining a weight applied to the first prediction sample when the coding block is square.

[0523] [Equation 26]

[0524] In Equation 26, CbW represents the coding block width. When the coding block is non-square, the weight applied to the first prediction sample can be obtained using Equation 27 or Equation 28 below. Equation 27 represents a case in which the width of a coding block is larger than the height, and Equation 28 represents a case in which the width of a coding block is smaller than the height.

[0525] [Equation 27]

[0526] [Equation 28]

[0527] In Equation 27, CbH represents the height of the coding block.

[0528] As in the illustrated example, the prediction sample included in the first triangular prediction element from the prediction samples in the boundary portion may be obtained by assigning a greater weight to the first prediction sample than the second prediction sample, and the prediction sample included in the second triangular prediction element may be obtained by assigning a greater weight to the second prediction sample than to the first prediction sample.

[0529] When diagonal division is applied to a coding block, the setting may be made such that a combined prediction mode, which is a combination of an intra prediction mode and a merging mode, is not applied to the coding block.

[0530]

[0531] Intra prediction is designed to predict the current block using a reconstructed sample that has been encoded or decoded and that is adjacent to the current block. In this case, the reconstructed sample before applying the filter in the loop can be used for intra-prediction of the current block.

[0532] The intra prediction method includes matrix-based intra prediction and directivity-based normal intra prediction with respect to the nearest reconstructed sample. Information indicating the intra prediction method of the current block may be carried in the bit stream. The information may be a 1-bit flag. Alternatively, the intra prediction method of the current block may be determined based on at least one of the location, size or shape of the current block or the intra prediction method of the adjacent block. As an example, when the current block is present at the edge of the image, the setting may be made such that the matrix-based intra prediction is not applied to the current block.

[0533] Matrix-based intra prediction may be a method of deriving a block prediction of a current block based on a matrix product between a matrix previously stored in the encoder and decoder and reconstructed samples adjacent to the current block. Information to indicate one of a plurality of pre-stored matrices may be transmitted in a bit stream. The decoder may determine a matrix for intra prediction of the current block based on the information and size of the current block.

[0534] Normal intra prediction is a method of obtaining a prediction block for a current block based on an undirected intra prediction mode or a directed intra prediction mode. The process of performing intra prediction based on normal intra prediction will be described in detail below with reference to drawings.

[0535] FIG. 40 is a flowchart of an intra prediction method according to an embodiment of the present invention.

[0536] The reference sample line of the current block can be determined (S4001). A reference line refers to a set of reference samples included in

k ^-th line from the top and/or left side of the current block. The reference samples can be obtained from the reconstructed samples that have been encoded or decoded and are adjacent to the current block.

[0537] Index information for identifying a reference sample line of a current block from a plurality of reference sample lines may be transmitted in a bit stream. As an example, index information intra_luma_ref_idx, which is intended to indicate the reference line of the current block, may be carried in the bit stream. Index information may be carried in encoding block elements.

[0538] The plurality of reference sample lines may comprise at least one of a first line, a second line, a third line, or a fourth line from above and/or to the left of the current block. A reference sample line consisting of a row adjacent to the top side of the current block and a column adjacent to the left side of the current block among a plurality of reference lines may be called an adjacent reference line, and other reference lines may be called non-adjacent reference lines.

[0539] Only some of the plurality of reference sample lines can be selected as the reference sample line of the current block. As an example, the remaining reference sample lines other than the third non-adjacent reference sample line of the plurality of reference sample lines may be set as candidate reference sample lines. Table 5 presents the indices assigned to the candidate reference sample lines.

[0540] [Table 5]

[0541] More candidate reference sample lines or fewer candidate reference sample lines may be set than described above. Moreover, the number or locations of non-adjacent reference sample lines set as candidate reference sample lines is not limited to the example described above. As an example, the first non-adjacent reference sample line and the third non-adjacent reference sample line may be set as candidate reference sample lines, and the second non-adjacent reference sample line and the third non-adjacent reference sample line may be set as candidate reference sample lines. Alternatively, all of the first non-adjacent reference sample line, the second non-adjacent reference sample line, and the third non-adjacent reference sample line may be set as candidate reference sample lines.

[0542] The number or types of candidate reference sample lines may be determined based on at least one of the size, shape, or location of the current block, the presence of subblock division, or the intra prediction mode of the current block.

[0543] The reference sample line of the current block may be determined based on at least one of the location, size or shape of the current block or the predictive coding mode of an adjacent block. As an example, when the current block is adjacent to the edge of an image, tile, slice, or code tree element, the adjacent reference sample line may be defined as the current block's reference line.

[0544] Alternatively, when the current block is non-square, the adjacent reference line may be defined as the current block's reference line. Alternatively, when the width-to-height ratio of the current block is equal to or greater than a threshold value, or equal to or less than a threshold value, an adjacent reference sample line may be defined as the reference sample line of the current block.

[0545] The reference sample line may include top reference samples located above the current block and left reference samples located to the left of the current block. The top reference samples and left reference samples can be obtained from the reconstructed samples adjacent to the current block. The reconstructed samples may be in the state before the filter was applied to the loop.

[0546] Next, the intra prediction mode of the current block can be determined (S4002). At least one of an undirected intra prediction mode or a directed intra prediction mode may be determined as the intra prediction mode of the current block. The non-directional intra prediction mode includes planar mode and DC mode, and the directional intra prediction mode includes 33 or 65 modes from diagonal down and left direction to diagonal up and right direction.

[0547] FIG. 41 is a schematic diagram showing intra prediction modes.

[0548] FIG. 41A shows 35 intra prediction modes, and FIG. 41B shows 67 intra prediction modes.

[0549] More intra prediction modes or fewer intra prediction modes may be specified than shown in FIG. 41.

[0550] The most likely mode (MPM) may be set based on the intra prediction mode of a block adjacent to the current block. Here, the neighbor block may include a left neighbor block adjacent to the left side of the current block and an upper neighbor block adjacent to the top side of the current block.

[0551] The number of MPMs included in the MPM list may be preset in the encoder and decoder. As an example, the number of MRMs may be three, four, five or six. Alternatively, information indicating the number of MPMs may be transmitted in the bit stream. Alternatively, the amount of MRM may be determined based on at least one of a predictive coding mode of an adjacent block or the size, shape, or line index of the current block's reference samples. As an example, when an adjacent reference sample line is defined as the current block reference line, N MRMs may be used, and when a non-adjacent reference sample line is defined as the current block reference line, M MRMs may be used. M represents a natural number that is less than N. As an example, N can be equal to six, and M can be equal to five, four or three. Accordingly, one of the six candidate intra prediction modes is determined to be the intra prediction mode of the current block when the reference sample line index of the current block is zero and the MPM flag is true, and one of the five candidate intra prediction modes is determined to be the intra prediction mode of the current block. block, when the index of the line of reference samples of the current block is greater than zero, and the MPM flag is true.

[0552] Alternatively, a fixed number (eg, six or five) MPM candidates may be used regardless of the reference sample line index of the current block.

[0553] An MPM list containing a plurality of MPMs is created, and information indicating whether an MPM identical to the intra prediction mode of the current block is included in the MPM list may be transmitted in the bit stream. The information is a 1-bit flag, and the flag may be called an MPM flag. When the MPM flag indicates that an MPM identical to the current block is included in the MPM list, index information to indicate one of the MPMs may be transmitted in the bit stream. As an example, index information intra_luma_ref_idx, which is intended to indicate one of a plurality of MRMs, may be carried in a bit stream. The intra prediction mode of the current block may be set to the MPM indicated by the index information. When the MPM flag indicates that an MPM identical to the current block is not included in the MPM list, residual mode information indicating one of the remaining intra prediction modes other than the MPM may be transmitted in the bit stream. The residual mode information indicates an index value corresponding to the intra prediction mode of the current block when the indices are reassigned to the remaining intra prediction modes other than MPM. To determine the intra prediction mode of the current block, the decoder may sort the MPM in ascending order and compare the MPM with the residual mode information. As an example, when the residual mode information is less than or equal to the MPM, the intra prediction mode of the current block can be obtained by adding one to the residual mode information.

[0554] When obtaining the intra prediction mode of the current block, the comparison between the residual mode information and some of the MPM may be skipped. As an example, MPMs that are non-directional intra prediction modes from the MPM may be excluded from comparison. When the non-directional intra prediction modes are set as the MPM, it is apparent that the residual mode information indicates the directed intra prediction mode. Thus, the intra prediction mode of the current block can be obtained by comparison between the residual mode information and the remaining MPMs other than the non-directed intra prediction modes. Instead of excluding the non-directional intra prediction modes from the comparison, the number of non-directional intra prediction modes can be added to the residual mode information, and thus the resulting value can be compared with the remaining MPMs.

[0555] Instead of setting the MPM as the default mode, information indicating whether the intra prediction mode of the current block is the default mode may be carried in the bit stream. The information may be a 1-bit flag, and the flag may be called a default mode flag. The default mode flag can be transmitted only when the MPM flag indicates that an MPM identical to the current block is included in the MPM list. As described above, the default mode may include at least one of planar mode, DC mode, vertical mode, or horizontal mode. As an example, when the default mode is set to planar mode, the default mode flag may indicate whether the intra prediction mode of the current block is planar mode. When the default mode flag indicates that the intra prediction mode of the current block is not the default mode, one of the MPMs indicated by the index information may be set as the intra prediction mode of the current block.

[0556] When the default mode flag is used, the setting may be made such that an intra prediction mode identical to the default mode is not set as the MPM. As an example, when the default mode flag indicates whether the intra prediction mode of the current block is planar, the intra prediction mode of the current block can be obtained using five MPMs excluding the MPM corresponding to planar.

[0557] When multiple intra prediction modes are set as default modes, index information indicating one of the default modes may be transmitted. The intra prediction mode of the current block can be set as the default mode indicated by the index information.

[0558] When the reference line index of the current block is non-zero, the setting can be made such that the default mode is not used. As an example, when a non-adjacent reference sample line is determined to be the reference sample line of the current block, the setting may be made such that a non-directional intra prediction mode such as a DC mode or a planar mode is not used. Accordingly, when the index of the reference sample line is non-zero, the default mode flag may not be transmitted, and the value of the default mode flag may be set to a preset value (ie, false).

[0559] When the intra prediction mode of the current block is determined, prediction samples for the current block can be obtained (S4003) based on the determined intra prediction mode.

[0560] When the DC mode is selected, prediction samples for the current block can be generated based on the average of the reference samples. In detail, the values of all samples in the prediction block may be generated based on the average value of the reference samples. The average may be obtained using at least one of the top reference samples located above the current block and the left reference samples located to the left of the current block.

[0561] The number or range of reference samples used to obtain the average may vary depending on the shape of the current block. As an example, when the current block is a non-square block in which the width is greater than the height, the average can be calculated using only the top reference samples. On the other hand, when the current block is a non-square block in which the width is less than the height, the average can be calculated using only the left reference samples. That is, when the width and height of the current block are different from each other, the average can be calculated using only the reference samples adjacent to the larger of the width and height. Alternatively, whether to calculate the average using the top reference samples or calculate the average using the left reference samples can be determined based on the width-to-height ratio of the current block.

[0562] When the planar mode is selected, a prediction sample can be obtained using a horizontal prediction sample and a vertical prediction sample. In this case, the horizontal prediction sample is obtained based on the left reference sample and the right reference sample located on the same horizontal line as the prediction sample, and the vertical prediction sample is obtained based on the upper reference sample and the lower reference sample located on the same vertical line , which is the prediction count. Here, the right reference sample may be generated by copying a reference sample adjacent to the upper right corner of the current block, and the lower reference sample may be generated by copying the reference sample adjacent to the lower left corner of the current block. The horizontal prediction count can be obtained based on the weighted sum operation of the left reference sample and the right reference sample, and the vertical prediction count can be obtained based on the weighted sum operation of the upper reference sample and the lower reference sample. In this case, the weight assigned to each reference sample can be determined depending on the location of the prediction sample. The prediction sample may be obtained based on an averaging operation or a weighted sum operation of the horizontal prediction sample and the vertical prediction sample. When performing the weighted sum operation, weights assigned to the horizontal prediction sample and the vertical prediction sample can be determined based on the location of the prediction sample.

[0563] When a directional prediction mode is selected, a parameter indicating the prediction direction (or prediction angle) of the selected directional prediction mode can be determined. Table 6 below presents the intraPredAng intra direction parameter for each intra prediction mode.

[0564] [Table 6]

[0565] Table 6 shows the intra direction parameter for each intra prediction mode with an index from 2 to 34 when 35 intra prediction modes are specified. When more than 33 directed intra prediction modes are specified, Table 6 can be subdivided to set the intra direction parameters for each directed intra prediction mode.

[0566] After arranging the top reference samples and the left reference samples of the current block in a line, a prediction sample can be obtained based on the value of the inner direction parameter. In this case, the left reference samples and the top reference samples can be arranged in a line when the value of the inner direction parameter is negative.

[0567] FIGS. 42 and 43 are schematic diagrams showing an example of a one-dimensional array in which reference samples are arranged in a line.

[0568] FIG. 42 shows an example of a one-dimensional vertical array in which the reference samples are located vertically, and FIG. 43 shows an example of a one-dimensional horizontal array in which the reference samples are located horizontally. The embodiments shown in FIGS. 42 and 43 will be described assuming that 35 intra prediction modes are specified.

[0569] A one-dimensional horizontal array, in which the top reference samples are rotated counterclockwise, can be used when the intra prediction mode index is between 11 and 18, and a one-dimensional vertical array, in which the left reference samples are rotated clockwise, can be used when , when the intra prediction mode index is between 19 and 25. When the reference samples are arranged in a line, the intra prediction mode angle can be taken into account.

[0570] Based on the internal direction parameter, a reference determination parameter can be determined. The reference sample determination parameter may comprise a reference sample index to indicate the reference sample and a weight parameter to determine the weight to be applied to the reference sample.

[0571] The reference index iIdx and the weighting parameter _ifact can be obtained using Equations 30 and 31 below.

[0572] [Equation 30]

[0573] [Equation 31]

[0574] In equations 30 and 31, Pang represents the internal direction parameter. The reference sample, indicated by the reference sample index iIdx, corresponds to an integer pixel.

[0575] At least one reference sample may be specified to obtain the prediction sample. The details of the location of the reference sample to be used in obtaining the prediction sample may be specified taking into account the slope of the prediction mode. As an example, a reference sample to use in obtaining a prediction sample may be specified using the reference sample index iIdx.

[0576] In this case, when the slope of the intra prediction mode cannot be expressed by a single reference sample, the prediction sample can be generated by interpolating multiple reference samples. As an example, when the slope of the intra prediction mode is a value between the slope between the prediction sample and the first reference sample and the slope between the prediction sample and the second reference sample, the prediction sample can be obtained by interpolating the first reference sample and the second reference sample. That is, the corner line corresponding to the intra prediction angle does not pass through a reference sample located in the entire pixel, the prediction sample can be obtained by interpolating the reference samples adjacent to the left, right, top or bottom sides of the location through which the corner line passes.

[0577] Below, Equation 32 shows an example of obtaining a prediction sample based on the reference samples.

[0578] [Equation 32]

[0579] In Equation 32, P represents a prediction sample, and Ref_1D represents one of the reference samples located in one dimension. In this case, the location of the reference sample can be determined from the reference sample index iIdx and the location (x, y) of the prediction sample.

[0580] When the slope of the intra prediction mode can be expressed by a single reference sample, the weighting parameter i _fact can be set to zero. Accordingly, Equation 32 can be simplified to Equation 33 below.

[0581] [Equation 33]

[0582] Intra prediction may be performed on the current block based on a variety of intra prediction modes. As an example, an intra prediction mode may be obtained for each prediction sample, and a prediction sample may be obtained based on the intra prediction mode assigned to the corresponding prediction sample.

[0583] Alternatively, an intra prediction mode may be obtained for each portion, and intra prediction may be performed on the corresponding portion based on the intra prediction mode assigned to the corresponding portion. In this case, the region may contain at least one sample. At least one of the size or shape of the patch may be adaptively determined based on at least one of the size, shape, or intra prediction mode of the current block. Alternatively, at least one of the size or shape of the region may be preset in the encoder and decoder regardless of the size or shape of the current block.

[0584] Alternatively, intra prediction may be performed multiple times, and the final prediction sample may be obtained based on an averaging operation or a weighted sum operation of multiple prediction samples obtained by performing intra prediction multiple times. As an example, a first prediction sample may be obtained by performing intra prediction based on a first intra prediction mode, and a second prediction sample may be obtained by performing intra prediction based on a second intra prediction mode. Then, the final prediction sample can be obtained based on an averaging operation or a weighted sum operation of the first prediction sample and the second prediction sample. In this case, the weights assigned to the first prediction sample and the second prediction sample may be determined taking into account at least one of whether the first intra prediction mode is an undirected prediction mode or a directional prediction mode, whether the second intra prediction mode is a undirected prediction or directed prediction mode, or intra-neighbor block prediction mode.

[0585] The plurality of intra prediction modes may be a combination of an undirected intra prediction mode and a directed intra prediction mode, a combination of directed intra prediction modes, or a combination of undirected intra prediction modes.

[0586] FIG. 44 is a schematic diagram illustrating the angles formed between a straight line parallel to the x-axis and directional intra prediction modes.

[0587] In the example shown in FIG. 44, directional prediction modes may range from diagonal down and left to diagonal up and right. Where the following description assumes the angle formed between the x-axis and the directional prediction mode, the directional prediction modes may range from 45 degrees (diagonally down and left) to -135 degrees (diagonally up and right).

[0588] When the current block has a non-square shape, a prediction sample may be obtained using a reference sample far from the prediction sample from reference samples located on the corner line corresponding to the intra prediction angle, instead of a reference sample near the prediction sample depending on the intra prediction mode of the current block.

[0589] FIG. 45 is a schematic view showing an aspect in which a prediction count is obtained when the current block has a non-square shape.

[0590] For example, as in the example shown in FIG. 45A, it is assumed that the current block has a non-square shape in which the width is greater than the height, and that the intra prediction mode of the current block is a directional intra prediction mode having an angle from 0 degrees to 45 degrees. In this case, when obtaining the prediction sample A near the right column of the current block, the left reference sample L away from the prediction sample from the reference samples located on the corner line corresponding to the corner may be used instead of the upper reference sample T near the prediction sample.

[0591] As another example, as in the example shown in FIG. 45B, it is assumed that the current block has a non-square shape in which the height is greater than the width, and that the intra prediction mode of the current block is a directional intra prediction mode having an angle from -90 degrees to -135 degrees. In this case, when obtaining a prediction sample A near the bottom line of the current block, the top reference sample T away from the prediction sample from the reference samples located on the corner line corresponding to the corner may be used instead of the left reference sample L near the prediction sample.

[0592] To solve this problem, when the current block has a non-square shape, the intra prediction mode of the current block can be replaced by an intra prediction mode in the opposite direction. Thus, for non-square shaped blocks, directional prediction modes having an angle that is larger or smaller than those of the directional prediction modes shown in FIG. 41 can be used. Such a directional intra prediction mode may be set as a wide-angle intra prediction mode. The wide-angle intra prediction mode represents the directional intra prediction mode, which is not in the range of 45 degrees to -135 degrees.

[0593] FIG. 46 is a schematic diagram showing wide-angle intra prediction modes.

[0594] In the example shown in FIG. 46, wide-angle intra prediction modes are indicated by intra prediction modes indexed from -1 to -14 and intra prediction modes indexed from 67 to 80.

[0595] FIG. 46 illustrates 14 wide-angle intra prediction modes having an angle that is greater than 45 degrees (-1 to -14) and 14 wide-angle intra prediction modes having an angle that is less than -135 degrees (67 to 80 ), however, more or fewer wide-angle intra prediction modes can be specified.

[0596] When the wide-angle intra prediction mode is used, the length of the top reference samples can be set to 2W+1, and the length of the left reference samples can be set to 2H+1.

[0597] When the wide-angle intra prediction mode is used, sample A shown in FIG. 45A can be predicted using reference sample T, and sample A shown in FIG. 45B can be predicted using reference sample L.

[0598] By adding N wide-angle intra prediction modes to the existing intra prediction modes, a total of 67+N intra prediction modes can be used. As an example, Table 7 shows the inner direction parameters of the intra prediction modes when 20 wide-angle intra prediction modes are specified.

[0599] [Table 7]

[0600] When the current block has a non-square shape and the intra prediction mode of the current block obtained in step S4002 is within the conversion range, the intra prediction mode of the current block can be converted to a wide-angle intra prediction mode. The transformation range may be determined based on at least one of the size, shape, or ratio of the current block. In this case, the ratio can indicate the relationship between the width and height of the current block.

[0601] When the current block has a non-square shape in which the width is greater than the height, the conversion range can be set to a range from the intra prediction mode index in the diagonal up and right direction (for example, 66) to the intra prediction mode index in the diagonal up and right direction minus N. In this case, N can be determined based on the ratio of the current block. When the intra prediction mode of the current block is within the conversion range, the intra prediction mode can be converted to a wide-angle intra prediction mode. The conversion may be the subtraction of a predetermined value from the intra prediction mode, and the predetermined value may be the total number (eg, 67) of intra prediction modes excluding wide-angle intra prediction modes.

[0602] According to the above-described embodiment, intra prediction modes No. 66 to No. 53 can be converted to wide-angle intra prediction modes No. -1 to No. -14, respectively.

[0603] When the current block has a non-square shape in which the height is greater than the width, the transformation range can be set to the diagonal down-left intra prediction mode index (eg, 2) to the diagonal down-left intra prediction mode index plus M. In this case, M can be determined based on the ratio of the current block. When the intra prediction mode of the current block is within the conversion range, the intra prediction mode can be converted to a wide-angle intra prediction mode. The transformation may be adding a predetermined value to the intra prediction mode, and the predetermined value may be the total number (eg, 65) of directional intra prediction modes excluding wide-angle intra prediction modes.

[0604] According to the above-described embodiment, intra prediction modes No. 2 to No. 15 can be converted to wide-angle intra prediction modes No. 67 to No. 80, respectively.

[0605] Intra prediction modes that are within the conversion range will be referred to as alternative wide-angle intra prediction modes.

[0606] The conversion range can be determined based on the ratio of the current block. As an example, Table 8 and Table 9 show the conversion range when 35 intra prediction modes excluding wide-angle intra prediction modes are specified, and the prediction range when 67 intra prediction modes are specified.

[0607] [Table 8]

[0608] [Table 9]

[0609] In the examples shown in Table 8 and Table 9, the number of alternative wide-angle intra prediction modes within the conversion range may differ depending on the ratio of the current block.

[0610] The conversion range, as shown in Table 10 below, can be set by subdividing the ratio of the current block.

[0611] [Table 10]

[0612] When a non-adjacent reference sample line is specified as the reference sample line of the current block, or when a multi-line intra prediction coding method is used to select one of a plurality of reference sample lines, the setting may be made such that the wide-angle intra prediction mode is not used . That is, even when the current block has a non-square shape, and the intra prediction mode of the current block is within the conversion range, the intra prediction mode of the current block may not be converted to the wide-angle intra prediction mode.

[0613] Alternatively, when the intra prediction mode of the current block is determined to be a wide-angle intra prediction mode, the setting may be made such that non-adjacent reference sample lines are not available as the current block's reference sample line, or the setting may be made such that the method No multi-line intra prediction coding is used to select one of a plurality of reference sample lines. When a multi-line intra prediction coding method is not used, an adjacent reference sample line can be defined as the reference sample line of the current block.

[0614] When the wide-angle intra prediction mode is not used, refW and refH may be set to the sum of nTbW and nTbH. Accordingly, with the exception of the top left reference samples, a non-adjacent reference sample located at distance i from the current block may include (nTbW + nTbH + offsefX[i]) top reference samples and (nTbW + nTbH + offsetY[i]) left reference samples . That is, a non-adjacent reference sample located at distance i from the current block may include (2nTbW + 2nTbH + offsetX[i] + offsetY[i] + 1) reference samples. For example, when the value of whRatio is greater than 1, the setting may be made such that the value of offsetX is greater than the value of offsetY. As an example, the offsetX value may be set to 1, and the offsetY value may be set to 0. On the other hand, when the whRatio value is less than 1, the setting may be made such that the offsetY value is greater than the offsetX value. As an example, the value of offsetX can be set to 0, and the value of offsetY can be set to 1.

[0615] Since wide-angle intra prediction modes are used in addition to existing intra prediction modes, resources required to encode wide-angle intra prediction modes may increase, and thus coding efficiency may decrease. Accordingly, coding efficiency can be improved by encoding alternative intra prediction modes for wide-angle intra prediction modes instead of encoding wide-angle intra prediction modes as is.

[0616] As an example, when the current block is encoded in the wide-angle intra prediction mode No. 67, the wide-angle intra prediction mode No. 2, which is an alternative wide-angle intra prediction mode for No. 67, may be encoded into the intra prediction mode of the current block. In addition, when the current block is encoded in the wide-angle intra prediction mode No.-1, the wide-angle intra prediction mode No. 66, which is an alternative wide-angle intra prediction mode for No.-1, can be encoded into the intra prediction mode of the current block.

[0617] The decoder may decode the intra prediction mode of the current block and may determine whether the decoded intra prediction mode is within the transform range. When the decoded intra prediction mode is an alternative wide-angle intra prediction mode, the intra prediction mode can be converted to a wide-angle intra prediction mode.

[0618] Alternatively, when the current block is encoded in the wide-angle intra prediction mode, the wide-angle intra prediction mode may be encoded as is.

[0619] Intra prediction mode coding may be performed based on the above-described MPM list. In detail, when a neighboring block is encoded in the wide-angle intra prediction mode, the MPM can be set based on the alternative wide-angle intra prediction mode corresponding to the wide-angle intra prediction mode.

[0620] The residual image can be obtained by subtracting the prediction image from the original image. In this case, when the afterimage is changed to the frequency domain, the subjective image quality is not significantly reduced even if the high-frequency components are removed from the frequency components. Accordingly, when the values of the high-frequency components are changed to small values, or when the values of the high-frequency components are set to 0, it is possible to improve the compression efficiency without causing significant visual distortion. Considering the characteristics described above, the current block can be transformed to decompose the residual image into two-dimensional frequency components. The conversion can be performed using a conversion method such as discrete cosine transform (DCT) or discrete sine transform (DST).

[0621] In DCT, a cosine transform is used to decompose (or transform) the residual image into (in) two-dimensional frequency components, and in DST, a sine transform is used to decompose (or transform) the residual image into (in) two-dimensional frequency components. As a result of the residual image transformation, the frequency components can be expressed using a basic circuit. As an example, when DCT is performed on an N×N block, N ² base circuit components can be obtained. The size of each of the base circuit components included in the N×N block can be obtained using a transformation. According to the conversion method used, the component size of the underlying circuit may be called the DCT coefficient or the DST coefficient.

[0622] The DCT transformation method is mainly used to transform an image in which many non-zero low-frequency components are distributed. The DST conversion method is mainly used for an image in which many high-frequency components are distributed.

[0623] The residual image may be transformed using a transformation method other than DCT or DST.

[0624] The conversion of the residual image into two-dimensional frequency components will be referred to as two-dimensional conversion. In addition, the size of the base circuit components obtained by the transformation result will be called the transformation factor. As an example, the conversion coefficient may be called the DCT coefficient or the DST coefficient. When both of the first transform and the second transform, which will be described below, are applied, the transform factor may relate to the size of the base circuit component generated as a result of the second transform.

[0625] The conversion method may be specified in block elements. The transformation method may be determined based on at least one of a predictive encoding mode of the current block or a size of the current block. As an example, when the current block is encoded in the intra prediction mode, and the size of the current block is less than N×N, the conversion may be performed using the DST conversion method. On the other hand, when the above condition is not satisfied, the conversion can be performed using the DCT conversion method.

[0626] Two-dimensional transformation may not be performed on some afterimage blocks. Failure to perform a two-dimensional transformation may be called a skip transformation. When skip conversion is applied, quantization can be applied to the residual values on which no conversion is performed.

[0627] After transforming the current block using DCT or DST, the transformed current block may be re-transformed. In this case, a transformation based on DCT or DST may be specified as the first transformation, and a retransformation of the block subject to the first transformation may be specified as the second transformation.

[0628] The first transformation may be performed using one of a plurality of candidate transformation bases. As an example, the first conversion may be performed using one of DCT2, DCT8 and DCT7.

[0629] Different transformation bases may be used in the horizontal direction and the vertical direction. Information indicating the combination of a vertical transform base and a horizontal transform base may be transmitted in the bit stream.

[0630] The first transformation and the second transformation may be performed in different elements. As an example, the first transformation may be performed on an 8×8 block, and the second transformation may be performed on a 4×4 subblock of the transformed 8×8 block. In this case, the conversion factor of the residual portions in which the second conversion is not performed can be set to 0.

[0631] Alternatively, the first transformation may be performed on the 4x4 block and the second transformation may be performed on the portion of the 8x8 block containing the transformed 4x4 block.

[0632] Information indicating whether to perform the second conversion may be transmitted in the bit stream.

[0633] The decoder may inverse the second transform (second inverse transform) and may inverse the first transform (first inverse transform) with respect to the result of the second inverse transform. By performing the second deconversion and the first deconversion, residual signals for the current block can be obtained.

[0634] Quantization is performed to reduce the block energy, and the quantization process involves dividing the conversion factor by a specific constant value. This constant value can be obtained by a quantization parameter, and the quantization parameter can be specified as a value from 1 to 63.

[0635] When the encoder performs the transform and quantization, the decoder may obtain the residual block by inverse quantization and inverse transform. The decoder can obtain a reconstructed block for the current block by adding the prediction block and the residual block.

[0636] When receiving the reconstructed block of the current block, the loss of information that occurs during quantization and encoding can be reduced by using in-loop filtering. The filter in the loop may comprise at least one of a deblocking filter, an adaptive sample offset filter (SAO), or an adaptive loop filter (ALF). The reconstructed block before applying the filter in the loop will be called the first reconstructed block, and the reconstructed block after applying the filter in the loop will be called the second reconstructed block.

[0637] The second reconstructed block may be obtained by applying at least one of a deblocking filter, SAO, or ALF to the first reconstructed block. In this case, SAO or ALF can be applied after applying the deblocking filter.

[0638] The deblocking filter is designed to mitigate image quality degradation (blocking artifact) that occurs at the edge of a block when performing quantization on block elements. To apply a deblocking filter, a block quality score (BS) between the first reconstructed block and the adjacent reconstructed block can be determined.

[0639] FIG. 47 is a flowchart showing a block quality determination process.

[0640] In the example shown in FIG. 47, P represents the first reconstructed block and Q represents the adjacent reconstructed block. In this case, the adjacent reconstructed block may be a block adjacent to the current block to the left or above.

[0641] In the example shown in FIG. 47, the block quality is determined taking into account the P and Q predictive coding modes, whether non-zero transform coefficients are included, whether inter-picture prediction is performed using the same reference image, or whether the difference value between motion vectors is greater than or equal to a threshold value.

[0642] Based on the quality indicator of the block, it can be determined whether to apply a deblock filter. As an example, filtering may not be performed when the block quality score is 0.

[0643] SAO is designed to reduce the ringing artifact that occurs when performing frequency domain quantization. SAO can be performed by adding or subtracting an offset determined given the design of the first reconstructed image. The method for determining the offset includes edge offset (EO) or stripe offset (BO). EO specifies a method for determining the offset of the current sample according to the pattern of nearby pixels. VO specifies a method for applying a common offset to a set of pixels having similar brightness values in a region. In detail, the brightness of the pixels is divided into 32 equal sections, and pixels having similar brightness values can be set as one set. As an example, four adjacent bands of 32 bands can be set as one group, and the same offset value can be applied to the samples belonging to the four bands.

[0644] ALF is a method of generating a second reconstructed image by applying a filter of a predetermined size or shape to a first reconstructed image or to a reconstructed image to which a deblocking filter has been applied. Equation 34 below provides an example application of ALF.

[0645] [Equation 34]

[0646] One of the predefined candidate filters may be selected in pixels, code tree elements, encoding blocks, prediction blocks, or transform blocks. Candidate filters may vary in size or shape.

[0647] FIG. 48 shows predefined candidate filters.

[0648] In the example shown in FIG. 48, at least one of a 5x5 diamond shape, a 7x7 diamond shape, or a 9x9 diamond shape may be selected.

[0649] For the chrominance component, only a 5x5 diamond shape can be used.

[0650]

[0651] The application of embodiments that have been described as directed to a decoding process or an encoding process to a decoding process or an encoding process is within the scope of the present invention. Changing the predetermined order in which the embodiments have been described to an order different from that described above is also within the scope of the present invention.

[0652] Although the above embodiments have been described in terms of a number of steps or sequences of operations, the steps or sequences of operations do not limit the order of the present invention in a time series and may be performed simultaneously or in other orders as necessary. In addition, each of the components (eg, elements, modules, etc.) constituting a block diagram in the above-described embodiments may be implemented by a hardware device or software, and a plurality of components may be combined and implemented by a single hardware device or software provision. The above-described embodiments may be implemented in the form of program instructions, which may be executed by various computer means and recorded in a computer-readable storage medium. The computer-readable storage medium may comprise a program instruction, a data file, a data structure, and the like. separately or in combination. Examples of computer readable storage media include magnetic media such as hard disk, floppy disk and magnetic tape, optical media such as compact disc read only memory (CD-ROM) and digital versatile disk (DVD), magneto-optical media such as flexible magneto-optical disk, and hardware devices such as ROM, random access memory (RAM), and flash memory, which are specifically designed to store and execute software commands. The hardware device may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

INDUSTRIAL APPLICABILITY

[0653] The present invention can be applied to an electronic device for encoding or decoding video.

Claims (41)

1. A video decoding method applied to a video decoder, including: determining whether to partition the encoding block so that it contains a first prediction element and a second prediction element; determining a partition type of the encoding block when determining that the encoding block should be divided; determining the index of the second merger candidate based on a comparison between the value of the second index information and the value of the first index information; obtaining first motion information with respect to the first prediction element in the encoding block and second motion information with respect to the second prediction element in the encoding block; And sampling a prediction sample in the encoding block based on the first motion information and the second motion information, wherein first motion information regarding the first prediction element is obtained from the first merge candidate indicated by the first index information from the plurality of merge candidates included in the merge candidate list, and second motion information regarding the second prediction element is obtained from a second merge candidate indicated by the second index information from a plurality of merge candidates included in the merge candidate list. 2. The video decoding method according to claim 1, wherein the index of the second merge candidate is equal to the value of the second index information plus one when the value of the second index information is greater than or equal to the value of the first index information. 3. The video decoding method according to claim 1, characterized in that the index of the second merge candidate is equal to the value of the second index information when the value of the second index information is less than the value of the first index information. 4. The video decoding method according to claim 1, characterized in that the prediction sample is obtained based on a weighted sum operation of the first prediction sample obtained based on the first motion information and the second prediction sample obtained based on the second motion information, when the prediction sample included in the boundary region between the first prediction element and the second prediction element. 5. The video decoding method according to claim 4, characterized in that the first weight coefficient applied to the first prediction sample is determined based on the x-coordinate and y-coordinate of the prediction sample. 6. The video decoding method of claim 5, wherein the second weight applied to the second prediction sample is obtained by subtracting the first weight from a constant value. 7. A video encoding method applied to a video encoder, including: determining whether to partition the encoding block so that it contains a first prediction element and a second prediction element; determining a partition type of the encoding block when determining that the encoding block should be divided; determining the index of the second merger candidate based on a comparison between the value of the second index information and the value of the first index information; obtaining first motion information with respect to the first prediction element in the encoding block and second motion information with respect to the second prediction element in the encoding block; And sampling a prediction sample in the encoding block based on the first motion information and the second motion information, wherein first motion information regarding the first prediction element is obtained from the first merge candidate of the plurality of merge candidates included in the list of merge candidates, second motion information with respect to the second prediction element is obtained from a second fusion candidate from a plurality of fusion candidates included in the fusion candidate list, and respectively encoding first index information to indicate a first merger candidate and second index information to indicate a second merger candidate. 8. The video encoding method according to claim 7, wherein the index of the second merge candidate is equal to the value of the second index information plus one when the value of the second index information is greater than or equal to the value of the first index information. 9. The video encoding method according to claim 7, wherein the index of the second merge candidate is equal to the value of the second index information when the value of the second index information is less than the value of the first index information. 10. Video decoding device containing: an inter-picture prediction element configured to determine whether the encoding block should be divided so that it contains a first prediction element and a second prediction element, determining a partition type of the encoding block when it is determined that the encoding block should be divided, determining an index of the second fusion candidate based on a comparison between the value of the second index information and the value of the first index information, obtaining first motion information with respect to the first prediction element in the encoding block and second motion information with respect to the second prediction element in the encoding block, and sampling the prediction sample in encoding block based on the first motion information and the second motion information, wherein the first motion information with respect to the first prediction element is obtained from the first fusion candidate indicated by the first index information of the plurality of fusion candidates included in the list of fusion candidates, and the second motion information with respect to the second prediction element is obtained from the second a merger candidate indicated by the second index information from a plurality of merger candidates included in the merger candidate list. 11. The video decoding apparatus as set forth in claim 10, wherein the index of the second merge candidate is equal to the value of the second index information plus one when the value of the second index information is greater than or equal to the value of the first index information. 12. The video decoding apparatus according to claim 10, characterized in that the index of the second merge candidate is equal to the value of the second index information when the value of the second index information is less than the value of the first index information. 13. Video encoder containing: processor, wherein the processor is configured to: determining whether to partition the encoding block so that it contains a first prediction element and a second prediction element; determining a partition type of the encoding block when it is determined that the encoding block should be divided; determining the index of the second merger candidate based on a comparison between the value of the second index information and the value of the first index information; obtaining first motion information with respect to the first prediction element in the encoding block and second motion information with respect to the second prediction element in the encoding block; And samples a prediction sample in the encoding block based on the first motion information and the second motion information, wherein the first motion information regarding the first prediction element is obtained from the first merge candidate of the plurality of merge candidates included in the merge candidate list, second motion information regarding the second prediction element is obtained from a second fusion candidate of the plurality of fusion candidates included in the fusion candidate list, and first index information for indicating the first merger candidate and second index information for indicating the second merger candidate are respectively encoded. 14. The video encoder of claim 13, wherein the index of the second merge candidate is equal to the value of the second index information plus one when the value of the second index information is greater than or equal to the value of the first index information. 15. The video encoder according to claim 13, wherein the index of the second merge candidate is equal to the value of the second index information when the value of the second index information is less than the value of the first index information.