Method and apparatus for hybrid intra prediction
By using a hybrid intra prediction method in video encoding, combining the intra prediction directions of the current block and adjacent blocks for weighted sum, the discontinuity problem at the block boundary in intra prediction is solved, and the encoding efficiency is improved.
Patent Information
- Application Number
- CN201980025828.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2018-05-09
- Filing Date
- 2019-04-29
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2039-04-29
AI Technical Summary
The existing video encoding technology has the problem of prediction discontinuity at block boundaries in intra prediction, especially in the case of object boundaries, curved edges or changing textures, which makes it difficult to effectively smooth the transition, resulting in low encoding efficiency.
By adopting the hybrid intra prediction method, the mixed intra prediction sub is constructed by weighting the intra prediction direction of the current block and the intra prediction direction of the adjacent block, and the corresponding weight decreases as the pixel distance increases, and the mixed intra prediction sub is constructed to smoothly. Transition at block boundaries.
It effectively reduces the prediction discontinuity at the block boundary, improves the efficiency of intra prediction, and enhances the performance of video encoding and decoding.
Smart Images

Figure CN112106372B_ABST
Abstract
Description
Technical Field
[0001] At least one of the present embodiments relates generally to methods or apparatus for video encoding and decoding, and more particularly to methods or apparatus for performing intra prediction using a hybrid intra predictor for pixels of a current block being encoded and decoded, wherein the hybrid intra predictor is determined by summing a predictor using an intra prediction direction to be used for the current block and one or more predictors using one or more corresponding intra prediction directions of one or more neighboring blocks using corresponding weights. Background Art
[0002] In order to achieve high compression efficiency, image and video coding schemes usually use prediction and transformation to supplement (1everage) spatial and temporal redundancy in video content. Usually, intra-frame or inter-frame prediction is used to exploit intra-frame or inter-frame correlation, and then the difference between the original image block and the predicted image block (usually expressed as a prediction error or prediction residual) is transformed, quantized, and entropy encoded. In order to reconstruct the video, the compressed data is decoded by the inverse process corresponding to prediction, transformation, quantization, and entropy coding.
[0003] Intra-frame prediction in video compression refers to the spatial prediction of a block of pixels using information from causal neighboring blocks (i.e., neighboring blocks that have been encoded or decoded in the same frame). Intra-frame prediction is a powerful coding tool because it allows high compression efficiency within a frame as well as between frames. Therefore, intra-frame prediction has been included as a core coding tool in many video compression standards, including but not limited to H.264 / AVC (Advanced Video Coding), H.265 / HEVC (High Efficiency Video Coding), and H.266 developed by JVET (Joint Video Exploration Team). Summary of the invention
[0004] According to a general aspect of at least one embodiment, a method for encoding video data is proposed, comprising: obtaining an intra-frame prediction direction for a current block to be encoded; obtaining one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; obtaining a weighted sum of hybrid intra-frame predictors of pixels of the current block, wherein the weighted sum is obtained by summing a predictor using the intra-frame prediction direction of the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and encoding the current block using the hybrid intra-frame predictor with the weighted sum.
[0005] According to another general aspect of at least one embodiment, a method for decoding video data is proposed, comprising: obtaining an intra-frame prediction direction of a current block to be decoded; obtaining one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; obtaining a weighted sum of hybrid intra-frame predictors of pixels of the current block, wherein the weighted sum is obtained by summing a predictor using the intra-frame prediction direction of the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and decoding the current block using the hybrid intra-frame predictor with the weighted sum.
[0006] According to another general aspect of at least one embodiment, a device for encoding video data is proposed, comprising one or more processors, wherein the one or more processors are configured to: obtain an intra-frame prediction direction of a current block to be encoded; obtain one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; obtain a weighted sum of hybrid intra-frame predictors of pixels of the current block, wherein the weighted sum is obtained by summing a predictor using the intra-frame prediction direction of the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and encode the current block using the hybrid intra-frame predictor with the weighted sum.
[0007] According to another general aspect of at least one embodiment, a device for decoding video data includes one or more processors, wherein the one or more processors are configured to: obtain an intra-frame prediction direction of a current block to be decoded; obtain one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; obtain a weighted sum of mixed intra-frame predictors of pixels of the current block, wherein the weighted sum is obtained by summing a predictor using the intra-frame prediction direction of the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and decode the current block using the mixed intra-frame predictor with the weighted sum.
[0008] According to another general aspect of at least one embodiment, a bitstream is formatted to include encoded blocks of video data, wherein the encoded blocks of video data are encoded in the following manner: obtaining an intra-frame prediction direction of the video data block to be encoded; obtaining one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the video data block; obtaining a weighted sum of hybrid intra-frame predictors for pixels of the video data block, wherein the weighted sum is obtained by summing a predictor using the intra-frame prediction direction of the video data block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks with corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and encoding the video data block using the hybrid intra-frame predictor having the weighted sum.
[0009] One or more of the embodiments further provides a computer-readable storage medium on which instructions for encoding or decoding video data according to the above method are stored. The embodiment further provides a computer-readable storage medium on which a bit stream generated according to the above method is stored. The embodiment further provides a method and device for transmitting or receiving a bit stream generated according to the above method. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] Figure 1 A block diagram of an exemplary video encoder is illustrated.
[0011] Figure 2A is a diagrammatic example depicting reference samples used for intra prediction in HEVC, Figure 2B is a diagrammatic example depicting intra prediction directions and corresponding modes in HEVC, Figure 2C is a graphical example depicting intra prediction directions and corresponding modes in the JEM (Joint Exploration Model) being developed by JVET.
[0012] Figure 3 A block diagram of an exemplary video decoder is illustrated.
[0013] Figure 4 The five possible neighbor candidates used in JEM for Most Probable Mode (MPM) encoding and decoding are illustrated graphically.
[0014] Figure 5 The use of a prediction block for intra prediction of a picture is illustrated.
[0015] Figure 6 It illustrates how a picture is subdivided into smaller coding blocks of different sizes depending on the picture content.
[0016] Figure 7The boundary prediction filters for intra modes 34 and 30 - 33 in HEVC are illustrated graphically.
[0017] Figure 8 Conceptually and graphically hybrid intra prediction using current and neighboring predictors and their corresponding directions is illustrated.
[0018] Fig. 9A 1 illustrates a graph of weighted values (on the Y-axis) versus pixel distances in number of pixels from corresponding neighboring candidates (on the X-axis) according to an exemplary embodiment, Fig. 9B Also illustrated is another graph of weighted values (on the Y-axis) versus pixel distances in number of pixels from corresponding neighboring candidates (on the X-axis) according to another exemplary embodiment.
[0019] Fig.10 Hybrid intra prediction for a picture with sub-blocks according to an embodiment is illustrated.
[0020] Fig.11 Also illustrated is hybrid intra prediction for a picture with sub-blocks according to an embodiment.
[0021] Fig.12 An exemplary method of video encoding according to an embodiment is illustrated.
[0022] Fig.13 An exemplary method of video decoding according to an embodiment is illustrated.
[0023] Fig.14 Another exemplary method of video encoding according to an embodiment is illustrated.
[0024] Fig.15 An exemplary method of video decoding according to an embodiment is illustrated.
[0025] Fig.16 An exemplary method for determining whether to use hybrid intra prediction according to an embodiment is illustrated.
[0026] Fig.17 Illustrated is hybrid intra prediction applied to a block that has been subdivided into 4 different sub-blocks according to an embodiment.
[0027] Fig.18 A block diagram is illustrated of an exemplary device in which various aspects of the exemplary embodiments may be implemented. DETAILED DESCRIPTION
[0028] Figure 1An exemplary video encoder 100 is illustrated, such as a HEVC encoder. HEVC is a compression standard developed by the Joint Collaborative Team on Video Coding (JCT-VC) (e.g., see "ITU-T H.265 TELECOMMUNICATION STANDARDIZATION SECTOR OF ITU (10 / 2014), SERIES H: AUDIOVISUAL AND MULTIMEDIASYSTEMS, Infrastructure of audiovisual services Coding of moving video, Highefficiency video coding, Recommendation ITU-T H.265"). Figure 1 It is also possible to illustrate an encoder in which the HEVC standard is improved or an encoder adopting a technique similar to HEVC, such as an encoder based on or improved on the JEM (Joint Exploration Model) being developed by JVET.
[0029] In this application, the terms "reconstruction" and "decoding" can be used interchangeably, the terms "pixel" and "sample" can be used interchangeably, and the terms "picture" and "frame" can be used interchangeably. Usually, but not necessarily, the term "reconstruction" is used on the encoder side, while "decoding" is used on the decoder side.
[0030] The HEVC specification distinguishes between "blocks" and "units", where a "block" addresses a specific area in the sample array (e.g., luma, Y), while a "unit" includes a concatenated block of all coded color components (Y, Cb, Cr, or monochrome), syntax elements, and prediction data associated with the block (e.g., motion vectors).
[0031] For encoding, the picture is partitioned into square coding tree blocks (CTBs) of configurable size, and a continuous set of coding tree blocks is grouped into slices. The coding tree unit (CTU) contains the CTB of the coded color component. The CTB is the root of a quadtree partitioned into coding blocks (CBs), and the coding block can be partitioned into one or more prediction blocks (PBs) and forms the root of a quadtree partitioned into transform blocks (TBs). Corresponding to the coding blocks, prediction blocks, and transform blocks, the coding unit (CU) includes a prediction unit (PU) and a tree-structured transform unit (TU) set, the PU includes prediction information for all color components, and the TU includes a residual coding syntax structure for each color component. The size of the CB, PB, and TB of the luminance component is applicable to the corresponding CU, PU, and TU. In this application, the term "block" may be used to refer to any one of CTU, CU, PU, TU, CB, PB, and TB. In addition, "block" may also be used to refer to macroblocks and partitions specified in H.264 / AVC or other video coding standards, and more generally refers to data arrays of various sizes.
[0032] exist Figure 1 In an exemplary encoder 100, as described below, a picture is encoded by an encoder element. The picture to be encoded is processed in units of CUs. Each CU is encoded using intra or inter mode. When a CU is encoded in intra mode, it performs intra prediction (160). In inter mode, motion estimation (175) and compensation (170) are performed. The encoder determines (105) which of intra mode or inter mode is used to encode the CU, and indicates the intra / inter determination through a prediction mode flag. The prediction residual is calculated by subtracting (110) the predicted block from the original image block.
[0033] The prediction residual is then transformed (125) and quantized (130). The quantized transform coefficients as well as the motion vectors and other syntax elements are entropy encoded (145) to output a bitstream. The encoder can also skip the transform and apply quantization directly to the untransformed residual signal on a 4x4 TU basis. The encoder can also bypass both transform and quantization, i.e., directly encode the residual without applying a transform or quantization process. In direct PCM encoding, no prediction is applied and the coding unit samples are encoded directly into a bitstream.
[0034] The encoder decodes the encoded block to provide a reference for further prediction. The quantized transform coefficients are dequantized (140) and inverse transformed (150) to decode the prediction residual. The image block is reconstructed by combining (155) the decoded prediction residual and the predicted block. An in-loop filter (165) is applied to the reconstructed picture to perform, for example, deblocking / SAO (sample adaptive offset) filtering to reduce coding artifacts. The filtered image is stored in a reference picture buffer (180).
[0035] Figure 3 A block diagram of an exemplary video decoder 300 (eg, a HEVC decoder) is shown. In the exemplary decoder 300, a bitstream is decoded by decoder elements as described below. The video decoder 300 generally performs the same Figure 1 The encoding pass shown corresponds to a decoding pass that performs video decoding as part of encoding the video data. Figure 3 It is also possible to illustrate a decoder in which the HEVC standard is improved or a decoder adopting a technique similar to HEVC, for example, a decoder based on JEM or improved on the basis of JEM.
[0036] In particular, the input to the decoder includes a video bitstream that may be generated by the video encoder 100. The bitstream is first entropy decoded (330) to obtain transform coefficients, motion vectors, and other encoded information. The transform coefficients are dequantized (340) and inverse transformed (350) to decode the prediction residual. The image block is reconstructed by combining (355) the decoded prediction residual and the predicted block. The predicted block can be obtained (370) from intra-frame prediction (360) or motion compensated prediction (i.e., inter-frame prediction) (375). Advanced motion vector prediction (AMVP) and merge mode techniques can be used to derive motion vectors for motion compensation, which can use interpolation filters to calculate interpolated values for sub-integer samples of reference blocks. An in-loop filter (365) is applied to the reconstructed image. The filtered image is stored in a reference picture buffer (380).
[0037] In order to exploit spatial redundancy, a CU in intra mode is predicted from reconstructed neighboring samples within the same slice. When considering the encoding / decoding of the current CU, the causal neighboring CU has already been encoded / decoded. To avoid mismatches, the encoder and decoder have the same prediction. Therefore, both the encoder and decoder use information from the reconstructed / decoded neighboring causal CU to form a prediction for the current CU.
[0038] The intra prediction process in HEVC consists of three steps: (1) reference sample generation, (2) intra sample prediction, and (3) post-processing of the predicted samples. Figure 2AAn exemplary HEVC reference sample for intra prediction is illustrated in FIG, where the reference sample candidates are the top row (shaded) and left column (also shaded) of the target block (unshaded). The pixel value at the coordinate (x, y) is shown by P(x, y) in FIG. 2 . For a CU of size N×N, the top row of 2N decoded samples is formed by the decoded CU. Similarly, the left column of 2N samples is formed by the decoded CU. The corner pixels from the top left decoded CU are used to fill the gap between the top row and left column references. If some samples are not available, for example, when the corresponding CU is not in the same slice or the current CU is at the frame boundary, reference sample replacement is performed, where the lost samples are copied from the available samples in a clockwise direction. The reference samples are then filtered using the specified filter, depending on the current CU size and prediction mode.
[0039] The next step, intra sample prediction, consists in predicting the pixels of the target CU based on the reference samples. To efficiently predict different kinds of content, HEVC supports multiple prediction modes. In particular, planar and DC prediction modes are used to predict smooth and gradually changing areas, while directional prediction modes (also called "angular prediction modes") are used to capture different directional structures. HEVC supports 33 directional prediction modes indexed from 2 to 34. These prediction modes correspond to Figure 2B Different prediction directions are shown, where the numbers (ie, 2, 3, ..., 34) represent the intra prediction mode index.
[0040] In JEM, the current proposal increases the number of intra prediction modes compared to HEVC. Figure 2C As shown in FIG. 1 , in addition to planar mode 0 and DC mode 1, JEM 3.0 uses 65 directional intra prediction modes. Figure 2B These directional intra prediction modes are numbered in increasing order from 2 to 66 in the same way as from 2 to 34 in HEVC as shown. These 65 directional prediction modes include the 33 directional prediction modes specified in HEVC, plus 32 additional directional prediction modes corresponding to angles between the two original angles. In other words, the prediction direction in JEM has twice the angular resolution of HEVC. A greater number of prediction modes has been proposed to exploit the possibility of a finer angular structure with the proposed larger block size.
[0041] Since there are multiple available intra prediction modes, the decoder needs mode information to form a prediction for an intra-coded CU. In HEVC, two different options can be used to encode the applicable luma intra prediction mode for the current block. If the applicable mode is included in the constructed list of three most probable modes (MPM), the mode is signaled by the index in the MPM list. Otherwise, the mode is signaled by a fixed-length binarization of the mode index.
[0042] In JEM, the MPM method is based on an algorithm in which five adjacent candidate positions are considered, namely, left (L), top (A), bottom left (BL), top right (AR), and top left (AL), as shown in Figure 4 shown.
[0043] Some prediction modes, such as DC mode and direct horizontal (i.e., mode 10 in HEVC, mode 18 in JEM, also referred to as "HOR") and direct vertical mode (i.e., mode 26 in HEVC, mode 50 in JEM, also referred to as "VER"), may result in discontinuities at the CU boundaries of the predicted samples. Therefore, in HEVC as well as JEM, such prediction modes are followed by a post-processing step in which the boundaries of the predicted samples are smoothed using a low-pass filter. Here, direct horizontal mode refers to a prediction mode when the reference samples to the left of the target block are repeated horizontally to the right for intra-frame prediction. Similarly, direct vertical mode refers to a prediction mode when the reference samples above the target block are repeated vertically downward for intra-frame prediction.
[0044] Various aspects of the present exemplary embodiment relate to improving intra-frame prediction for video encoding and decoding. For example, the present embodiment extends intra-frame prediction beyond the boundaries of the current block to smooth the transition between blocks with different prediction directions. It is proposed to create a hybrid mode that mixes the prediction used for the direction of the current block with the prediction used for the direction used by the adjacent blocks. The present embodiment improves the efficiency of intra-frame prediction for blocks that bring supplementary information to their causal neighbors by combining the mode of the neighborhood with the current mode. The present embodiment aims to reduce prediction discontinuities at block boundaries. Note that, as previously mentioned, block boundary discontinuities are currently addressed in JEM by intra-frame block boundary filters, but JEM does not take into account the prediction directions of adjacent blocks.
[0045] In existing codecs, intra prediction aims to construct a first version of the current block, called a predictor, which minimizes the energy of the residual signal to be transmitted in the bitstream, e.g. Figure 5 The residual is transformed in the frequency domain and then quantized and encoded. Therefore, the goal is usually not to provide the best visually pleasing predictor, but to use the available samples in a way that minimizes the cost of the encoded or transmitted data.
[0046] For example, in the case where a block is located at an object boundary, curved edge, or varying texture, the encoder may have the following options:
[0047] - Use the best match directional mode,
[0048] - Use DC or flat, or
[0049] - Where possible, split the block so that smaller sub-blocks are better predicted at the cost of more syntax elements.
[0050] For example, a curved boundary of an object will cause the encoder to select smaller patches along the edge so that the straight mode can approximate the shape, e.g. Figure 6 As shown. Therefore, curved edges and changing textures are mostly encoded using small blocks. However, existing directional models consist in propagating or replicating reference samples along a fixed direction. Thus, for example, a curve is predicted by multiple straight lines. Therefore, the prediction is generally not very effective.
[0051] As mentioned before, for certain directions of intra prediction, in order to smooth prediction discontinuities at block boundaries, filters are applied along edges that are not used for prediction. In HEVC, after the intra prediction blocks have been generated for VER and HOR intra modes, the leftmost column and topmost row of prediction samples are further adjusted, respectively. This approach has been further extended in JEM to several diagonal intra modes, and up to four columns or rows of boundary samples are further adjusted using a two-tap (for intra modes 2 and 34) or a three-tap filter (for intra modes 3-6 and 30-33). Examples of boundary prediction filters for intra modes 34 and 30-33 are shown in FIG. Figure 7 , and the boundary prediction filters for intra modes 2 and 3-6 are similar but not shown. It should be noted that Figure 7 In FIG. 5 , filtering for HEVC modes is shown. In more recent versions of JEM, the method has been extended to 67 modes as described above.
[0052] In order to better predict the spatial direction of changes in objects and textures and smooth the transition between blocks with different prediction directions, this embodiment uses hybrid intra prediction to derive a hybrid intra predictor for the current block being encoded or decoded. Figure 8 Hybrid intra prediction extends intra prediction beyond block boundaries to create a hybrid mode that mixes classical prediction according to the direction of the current block (e.g., as used in H.264 or HEVC, and Figure 8 a), and prediction based on the direction used by the neighboring blocks (e.g., Figure 8 The prediction directions of neighboring blocks are used to create multiple prediction candidates, which are then combined with the classic current block prediction using a weighted sum that favors a given prediction for pixels that are closer to a given relevant neighborhood (e.g., as Figure 8 as shown in c).
[0053] According to aspects of the present embodiment, in order to construct the current intra-frame prediction, the intra-frame prediction direction of the adjacent block that has been encoded or decoded is used to generate a temporary prediction. The classical prediction of the current block with the encoding or decoding mode is also calculated as the temporary prediction. Then, these temporary predictions are combined as a weighted sum so that the corresponding weights for prediction decrease as the distance between the pixels of the current block being encoded or decoded and the corresponding adjacent blocks increases, and the sum of the weights is equal to 1. Therefore, the present embodiment ensures the generation of smooth transitions between blocks, and gives more importance to the classical prediction as the pixels further leave the neighborhood that has been encoded or decoded.
[0054] In one embodiment, for a current block of size W×H, the following operations are performed:
[0055] - Identify relevant neighborhood locations such as for MPM derivation (e.g. Figure 4 shown),
[0056] - Compute a prediction P for each neighborhood n∈{left, above, right, below} that has been encoded / decoded and has intra prediction n , where computing the prediction means doing a classic prediction for the current block, but with the prediction direction used / coded for block n,
[0057] -Use the prediction direction encoded for the current block to calculate the classic prediction P of the current block c ,
[0058] - Calculate the final prediction P by combining the predictions as follows final :
[0059] P final (x, y) = w c (x, y)×P c (x,y)+∑ n∈{left,above...} w n (x, y)×P n (x, y) (1)
[0060] w i is the weight, i∈{c, left, above, right, below}, for the current block w i =w c , i = current, w i (i≠c) depends on the distance from the pixel at coordinate (x, y) ((0, 0) is the upper left pixel of the current block) to the neighborhood i, and ∑ i w i (x, y) = 1 is used to predict P i :
[0061] ○For i=left, Where W is the width of the current block
[0062] ○For i=up, Where H is the height of the current block
[0063] ○For i=right,
[0064] ○For i=down,
[0065] ○ For the current block prediction, w c (x, y) = 1 - ∑ i≠c w i (x, y)
[0066] Fig. 9A The weight value (on the Y axis) based on the weighted calculation is graphically illustrated versus the pixel distance (on the X axis) to the corresponding neighboring candidate in terms of the number of pixels, where w left (x, y) and W = 16. Four neighborhoods are used here. It should be noted that a different set of neighborhood positions can be used, and more or fewer neighborhoods can also be used.
[0067] Note that the weighting can be implemented using fixed-point arithmetic, for example, by substituting all w i Multiply by M = 2 m (e.g., m=5) and then divide the result by M, where the division operation can be achieved by right shifting:
[0068]
[0069] Using these equations, the width of the combined region (where w i >0) depends on the block size: 1 pixel for size 4, 2 pixels for size 8, and so on.
[0070] In another non-limiting exemplary embodiment, another set of weighted calculations may be calculated as follows:
[0071] ○ where w c (x, y) = 1 - ∑ i≠c w i (x, y) is used for current block prediction.
[0072] Fig. 9B Illustrated is a graph of weighted values (on the Y axis) based on weighted calculation according to the above-described second exemplary embodiment versus pixel distances from corresponding neighboring candidates in units of the number of pixels (on the X axis).
[0073] In yet another non-limiting exemplary embodiment, another set of weighted calculations may be calculated as follows:
[0074] ○w0=.5, w1=.3, w2=.1, w k =0, if k>2 where w c (x, y) = 1 - ∑ i≠c w i (x, y) is used for current block prediction,
[0075] In addition, exemplary aspects of this embodiment illustrate the possibility of having sub-blocks for the current block under consideration and / or for neighboring blocks used as part of the weighted sum. In an exemplary embodiment, prediction is performed per sub-block. The basic principle is to deal with blocks where the neighborhood is split along a boundary or border. Fig.10 As shown, the current block may have several neighbors along the above block edge, each with its own prediction mode. In this case, the prediction P n Defined by the closest neighbor, i.e., the sub-block immediately above the upper edge of the current block, or the block or sub-block immediately to the left of the left edge of the current block or sub-block. Fig.10 , for pixels in sub-block 1 of the current block, the "Above 1" block will be used as the upper neighboring block, and for pixels in sub-block 2 of the current block, the "Above 2" block will be used as the upper neighboring block.
[0076] Note that the left edge of sub-block 2 may not need to consider the left neighboring sub-block, because the sub-block on the left has the same direction as sub-block 2 (in fact, it is in the same block).
[0077] Fig.11 Another exemplary embodiment is shown, which is Fig.10 The previous embodiment shown provides additional advantages. As mentioned before, hybrid intra prediction can be done by sub-blocks when the current block has several neighbors along a block edge, each with its own prediction mode. However, at sub-block boundaries, in order to avoid abrupt changes between hybrid intra predictions from different neighbors (e.g., upper 1 and upper 2), these predictions will also be mixed near sub-block boundaries.
[0078] In this case, a temporal prediction is established for each neighboring subblock of the entire block. Fig.10 As in the case shown, for the current block, three temporary predictions are calculated using the directions of the neighborhood and one temporary prediction is calculated using the direction of the current block. These temporary predictions are then summed together as described above, without any changes to sub-block 1. For sub-block 2, the weights or weights are instead calculated as follows:
[0079] ○ For i = 1
[0080] ○ For i = upper 2
[0081] In yet another non-limiting exemplary embodiment, for sub-block 1, the weighting or weight may instead be calculated as follows without additionally considering the above 2 neighboring blocks:
[0082] ○ For i = upper 2
[0083] ○ For i = upper 1
[0084] ○ For i = left
[0085] Based on the above and Fig.11 As shown, since the current hybrid intra prediction considers and mixes the different directions of the top neighboring sub-blocks, the boundary between the top sub-block 1 and sub-block 2 is now smoothed. This reduces the abrupt change between the hybrid predictions from the top 1 and top 2 neighboring blocks at the top of the current block near the sub-block boundary.
[0086] Fig.12 FIG. 1 illustrates a method for encoding video data by an encoder (eg, Figure 1 1. An exemplary encoding method 1200 performed by the encoder 100 in FIG. 12. In step 1210, the encoder determines the intra prediction direction to be used for the current block to be encoded. In step 1220, the encoder determines one or more corresponding intra prediction directions of one or more neighboring blocks of the current block. In step 1230, the encoder determines a weighted sum of hybrid intra predictors for pixels of the current block, wherein the weighted sum is determined by summing a predictor using the intra prediction direction to be used for the current block with one or more predictors using one or more corresponding intra prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel from the one or more neighboring blocks increases. In step 1240, the encoder encodes the current block using the hybrid intra predictor with the weighted sum. In step 1250, the intra prediction direction of the current block is also encoded.
[0087] Fig.13 FIG. 1 illustrates a method for decoding video data by a decoder (eg, Figure 313. The exemplary decoding method 1300 performed by the decoder 300 in FIG. 13. At step 1310, the decoder determines an intra prediction direction to be used for a current block to be decoded. At step 1320, the decoder determines one or more corresponding intra prediction directions of one or more neighboring blocks of the current block. At step 1330, the decoder determines a weighted sum of hybrid intra predictors for pixels of the current block, wherein the weighted sum is determined by summing a predictor using an intra prediction direction to be used for the current block with one or more predictors using one or more corresponding intra prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel from the one or more neighboring blocks increases. At step 1340, the decoder decodes the current block using the hybrid intra predictor with the weighted sum.
[0088] like Fig.12 and Fig.13 As shown, for one block used for hybrid intra prediction, only one intra prediction mode is still encoded or decoded. However, neighboring intra prediction directions are also used in hybrid intra prediction to smooth block boundary discontinuities.
[0089] In one non-limiting exemplary embodiment, a signaling flag or syntax element indicating the use of hybrid intra prediction is encoded at the encoder and sent to the decoder. Fig.14 Such an exemplary encoding method 1400 performed by an encoder is illustrated. Fig.14 At step 1410, the encoder determines whether the left edge and / or the top edge of the current block are qualified for hybrid intra prediction. If the determination is no, then at step 1420, hybrid intra prediction is not used, and no hybrid intra prediction mode flag or syntax element is required or encoded. On the other hand, if the determination is yes, then at step 1430, hybrid intra prediction is used, and the hybrid intra prediction mode flag or syntax element is encoded and sent to the decoder.
[0090] Fig.15 An exemplary decoding method performed by a decoder is illustrated. Fig.15 At step 1510, the decoder determines whether the left edge and / or top edge of the current block is qualified for hybrid intra prediction, for example, there is an adjacent intra prediction. If the determination is no, then at step 1520, hybrid intra prediction is not used. On the other hand, if the determination is yes, then at step 1530, the hybrid intra prediction flag is decoded. At step 1540, a determination is made to see if the decoded hybrid intra prediction flag is true (e.g., =1). If no, then hybrid intra prediction is not used in step 1550. If yes, then hybrid intra prediction is used for the left / top edge in step 1560.
[0091] In another non-limiting exemplary embodiment, the above hybrid intra prediction mode flag or syntax element is signaled for each coding unit. In another non-limiting exemplary embodiment, hybrid intra prediction is always used when a neighboring block or subblock uses intra prediction and therefore a dedicated signaling flag or syntax element indicating the use of hybrid intra prediction is not required.
[0092] In another non-limiting exemplary embodiment, hybrid intra prediction is only applied if the difference between the intra prediction direction for the current block and one of the one or more corresponding intra prediction directions for one or more neighboring blocks is greater than a threshold. In yet another non-limiting exemplary embodiment, hybrid intra prediction is only applied if the height or width of the current block is greater than a threshold. For example, hybrid intra prediction is only applied on an edge if the edge length is greater than or equal to 8 pixels, as defined in Table 1 below:
[0093] Table 1
[0094] Block size (WxH) Hybrid intra prediction uses 4x4 No hybrid intra prediction 4x8 Hybrid intra prediction for left edge only 8x4 Hybrid intra prediction for top edge only 8x8 Hybrid intra prediction of two edges 16x4 Hybrid intra prediction for top edge only
[0095] Fig.16 Another exemplary method 1600 incorporating aspects of the present embodiment is illustrated. The method 1600 may be used on the encoder or decoder side. At step 1610, a determination is made to see whether the left and / or upper neighboring blocks are intra-coded. If the determination is no, then at step 1620, hybrid intra prediction is not used. On the other hand, if the determination is yes, then at step 1630, another determination is made to see whether the difference between the intra prediction direction for the current block and one or more corresponding intra prediction directions of one or more neighboring blocks is greater than a threshold, where the threshold is 4 directional mode differences (e.g., the difference between JEM intra prediction angle mode 7 and mode 2 will be 5, which is greater than the threshold 4). If the determination at step 1630 is no, then at step 1640, hybrid intra prediction is not used. On the other hand, if the determination at step 1630 is yes, then at step 1650, another determination is made to see whether the width or height of the current block is greater than or equal to 8, as shown in Table 1 above. If the determination at step 1650 is no, then at step 1660, hybrid intra prediction is not used. On the other hand, if the determination at step 1650 is yes, then at step 1670 , hybrid intra prediction is applied to the qualifying edges of the current block.
[0096] In another non-limiting exemplary embodiment, for example, the current block is subdivided into four different sub-blocks, but the adjacent blocks immediately above and to the left are not subdivided, such as Fig.17 This embodiment is interesting when the direction of prediction changes for each sub-block within the current block.
[0097] Therefore, the present embodiment differs from the previously proposed and known overlapped block intra prediction (OBIP) in that the present embodiment provides weights for one or more neighboring blocks that decrease as the distance of the considered pixel from the one or more neighboring blocks increases. In addition, the present embodiment illustrates the possibility of sub-blocks for the current block being considered or its reference neighboring blocks.
[0098] Various methods are described above, and each method includes one or more steps or actions for implementing the described method. Unless a specific order of steps or actions is required for the correct operation of the method, the order and / or use of specific steps and / or actions can be modified or combined.
[0099] The various methods according to the present embodiment described above can be used to modify, for example, Figure 1 and Figure 3 1 and 2. Intra prediction modules (160, 360) of JVET or HEVC encoder 100 and decoder 300 are shown in FIG. 1 , respectively. In addition, the present embodiment is not limited to JVET or HEVC and can be applied to other standards, recommendations and extensions thereof. The various aspects and / or embodiments described above can be used alone or in combination.
[0100] Various numerical values are used in the present application, for example, the mode difference threshold for turning on / off hybrid intra prediction, and the constants used in the weight calculation equation. It should be noted that the specific values are for exemplary purposes, and the present embodiment is not limited to these specific values.
[0101] Fig.18 1800 is a block diagram of an exemplary system in which various aspects of the exemplary embodiments may be implemented. System 1800 may be implemented as a device including the various components described below and configured to perform the above-described processes. Examples of such devices include, but are not limited to, personal computers, laptop computers, smart phones, tablet computers, digital multimedia set-top boxes, digital television receivers, personal video recording systems, connected home appliances, and servers. System 1800 may be communicatively coupled to other similar systems and communicated via, for example, Fig.18 A communication channel as shown and as known to those skilled in the art is coupled to the display to implement the exemplary video system described above.
[0102] The system 1800 may include at least one processor 1810, which is configured to execute instructions loaded therein for implementing various processes as described above. The processor 1810 may include embedded memory, input and output interfaces, and various other circuits known in the art. The system 1800 may also include at least one memory 1820 (e.g., volatile memory device, non-volatile memory device). The system 1800 may further include a storage device 1820, which may include non-volatile memory, including but not limited to EEPROM, ROM, PROM, RAM, DRAM, SRAM, flash memory, disk drive, and / or optical drive. As a non-limiting example, the storage device 1840 may include an internal storage device, an attached storage device, and / or a network accessible storage device. The system 1800 may also include an encoder / decoder module 1830, which is configured to process data to provide encoded video or decoded video.
[0103] The encoder / decoder module 1830 represents a module(s) that can be included in the device to perform encoding and / or decoding functions. As is well known, the device can include one or both of the encoding and decoding modules. In addition, the encoder / decoder module 1830 can be implemented as a separate element of the system 1800, or can be incorporated into the processor 1810 as a combination of hardware and software known to those skilled in the art.
[0104] Program code to be loaded onto the processor 1810 to perform the various processes described above may be stored in the storage device 1840 and subsequently loaded onto the memory 1820 for execution by the processor 1810. According to an exemplary embodiment, one or more of the processor(s) 1810, the memory 1820, the storage device 1840, and the encoder / decoder module 1830 may store one or more of the various items during execution of the processes discussed above herein, including but not limited to input video, decoded video, bitstream, equations, formulas, matrices, variables, operations, and operation logic.
[0105] The system 1800 may also include a communication interface 1850 that enables communication with other devices via a communication channel 1860. The communication interface 1850 may include, but is not limited to, a transceiver configured to transmit and receive data from the communication channel 1860. The communication interface may include, but is not limited to, a modem or a network card, and the communication channel may be implemented in a wired and / or wireless medium. The various components of the system 1800 may be connected or communicatively coupled together using a variety of suitable connections, including, but not limited to, internal buses, wires, and printed circuit boards.
[0106] The exemplary embodiments may be performed by computer software, or hardware, or a combination of hardware and software implemented by the processor 1810. As a non-limiting example, the exemplary embodiments may be implemented by one or more integrated circuits. As a non-limiting example, the memory 1820 may be of any type suitable for the technical environment and may be implemented using any appropriate data storage technology, such as an optical storage device, a magnetic storage device, a semiconductor-based storage device, a fixed memory, and a removable memory. As a non-limiting example, the processor 1810 may be of any type suitable for the technical environment and may include one or more of a microprocessor, a general-purpose computer, a special-purpose computer, and a processor based on a multi-core architecture.
[0107] The implementation described herein can be implemented, for example, with a method or process, device, software program, data stream, or signal. Even if only discussed in the context of a single form of implementation (e.g., only discussed as a method), the implementation of the discussed feature can also be implemented in other forms (e.g., device or program). The device can be implemented, for example, with appropriate hardware, software, and firmware. The method can be implemented in a device such as a processor, which generally refers to a processing device, including, for example, a computer, a microprocessor, an integrated circuit, or a programmable logic device. The processor also includes a communication device, such as a computer, a mobile phone, a portable / personal digital assistant ("PDA"), and other devices that contribute to the information communication between end users.
[0108] According to a general aspect of at least one embodiment, a method for encoding video data is proposed, comprising: determining an intra-frame prediction direction for a current block to be encoded; determining one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; determining a weighted sum of hybrid intra-frame predictors for pixels of the current block, wherein the weighted sum is determined by summing a predictor using the intra-frame prediction direction of the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and encoding the current block using the hybrid intra-frame predictor with the weighted sum.
[0109] According to another general aspect of at least one embodiment, a method for decoding video data is proposed, comprising: determining an intra-frame prediction direction for a current block to be decoded; determining one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; determining a weighted sum of hybrid intra-frame predictors for pixels of the current block, wherein the weighted sum is determined by summing a predictor using the intra-frame prediction direction of the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and decoding the current block using the hybrid intra-frame predictor with the weighted sum.
[0110] According to another general aspect of at least one embodiment, a device for encoding video data is proposed, comprising one or more processors, wherein the one or more processors are configured to: determine an intra-frame prediction direction for a current block to be encoded; determine one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; determine a weighted sum of hybrid intra-frame predictors for pixels of the current block, wherein the weighted sum is determined by summing a predictor using the intra-frame prediction direction of the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and encode the current block using the hybrid intra-frame predictor with the weighted sum.
[0111] According to another general aspect of at least one embodiment, a device for encoding video data is proposed, comprising: a component for determining an intra-frame prediction direction for a current block to be encoded; a component for determining one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; a component for determining a weighted sum of hybrid intra-frame predictors for pixels of the current block, wherein the weighted sum is determined by summing a predictor using the intra-frame prediction direction of the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and a component for encoding the current block using the hybrid intra-frame predictor having the weighted sum.
[0112] According to another general aspect of at least one embodiment, a device for decoding video data includes one or more processors, wherein the one or more processors are configured to: determine an intra-frame prediction direction for a current block to be decoded; determine one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; determine a weighted sum of hybrid intra-frame predictors for pixels of the current block, wherein the weighted sum is determined by summing a predictor using the intra-frame prediction direction to be used for the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and decode the current block using the hybrid intra-frame predictor with the weighted sum.
[0113] According to another general aspect of at least one embodiment, a device for decoding video data includes: a component for determining an intra-frame prediction direction for a current block to be decoded; a component for determining one or more corresponding intra-frame prediction directions of one or more neighboring blocks of the current block; a component for determining a weighted sum of hybrid intra-frame predictors for pixels of the current block, wherein the weighted sum is determined by summing a predictor using an intra-frame prediction direction to be used for the current block and one or more predictors using one or more corresponding intra-frame prediction directions of one or more neighboring blocks using corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and a component for decoding the current block using the hybrid intra-frame predictor having the weighted sum.
[0114] According to one embodiment, the current block represents a current sub-block that has been subdivided from a larger block, and the one or more neighboring blocks represent one or more neighboring sub-blocks that have been subdivided from one or more larger neighboring blocks, respectively.
[0115] According to another embodiment, the one or more neighboring blocks include one or more blocks above the current block.
[0116] According to another embodiment, the one or more neighboring blocks include one or more blocks to the left of the current block.
[0117] According to another embodiment, the one or more neighboring sub-blocks include a neighboring sub-block immediately above the current sub-block.
[0118] According to another embodiment, the one or more adjacent sub-blocks further include a sub-block immediately above and to the left of the current sub-block.
[0119] According to another embodiment, the one or more neighboring sub-blocks further include a sub-block immediately to the left of the current sub-block.
[0120] According to another embodiment, if the difference between the intra-frame prediction direction of the current block and one of the one or more corresponding intra-frame prediction directions of the one or more neighboring blocks is greater than a threshold, only the hybrid intra-frame predictor with the weighted sum is applied to decode or encode the current block.
[0121] According to another embodiment, if the height or width of the current block is greater than a threshold, only the hybrid intra predictor with the weighted sum is applied to decode or encode the current block.
[0122] According to another general aspect of at least one embodiment, a bitstream is formatted to include an encoded block of video data, wherein the encoded block of video data is encoded by: determining an intra prediction direction of the video data block to be encoded; determining one or more corresponding intra prediction directions of one or more neighboring blocks of the video data block; determining a weighted sum of hybrid intra predictors for pixels of the video data block, wherein the weighted sum is determined by summing a predictor using the intra prediction direction of the video data block and one or more predictors using one or more corresponding intra prediction directions of one or more neighboring blocks with corresponding weights, and the corresponding weights of the one or more neighboring blocks decrease as the distance of the pixel to the one or more neighboring blocks increases; and encoding the video data block using the hybrid intra predictor having the weighted sum. .
[0123] References to "one embodiment" or "an embodiment" or "an implementation" or "an implementation", and other variations thereof, mean that a particular feature, structure, characteristic, etc. described in connection with the embodiment is included in at least one embodiment. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" or "in one implementation" or "in an implementation", and any other variations thereof, appearing in various places throughout the specification do not necessarily all refer to the same embodiment.
[0124] Furthermore, this application or its claims may refer to "determining" various pieces of information. Determining information may include, for example, one or more of estimating information, calculating information, predicting information, or retrieving information from a memory.
[0125] Furthermore, this application or its claims may refer to "accessing" various pieces of information. Accessing information may include, for example, one or more of receiving information, retrieving information (e.g., from a memory), storing information, processing information, transmitting information, moving information, copying information, erasing information, calculating information, determining information, predicting information, or estimating information.
[0126] Furthermore, the application or its claims may refer to "receiving" various pieces of information. Like "accessing," receiving is intended to be a broad term. Receiving information can include, for example, one or more of accessing information, or retrieving information (e.g., from a memory). Furthermore, "receiving" is often referred to in one way or another during operations such as storing information, processing information, transmitting information, moving information, copying information, erasing information, calculating information, determining information, predicting information, or estimating information.
[0127] It will be apparent to those skilled in the art that various implementations can generate various signals that are formatted to carry information that can be, for example, stored or transmitted. The information can include, for example, instructions for executing the method, or data generated by one of the implementations. For example, a signal can be formatted to carry a bitstream of the embodiment. Such a signal can be formatted as, for example, an electromagnetic wave (e.g., using the radio frequency portion of the spectrum) or a baseband signal. Formatting can include, for example, encoding a data stream and modulating a carrier with the encoded data stream. The information carried by the signal can be, for example, analog or digital information. As is known, the signal can be transmitted by various wired or wireless links. The signal can be stored on a processor readable medium.
Claims
1. A method for encoding video data, include: Obtaining an intra prediction direction of a current block to be encoded; Obtaining a plurality of adjacent blocks of the current block along a first block edge of the current block, wherein the first block edge has a first direction; Obtaining a neighboring block of the current block along a second block edge of the current block, wherein the second block edge has a second direction; Obtaining a plurality of sub-blocks of the current block based on the plurality of neighboring blocks along an edge of the first block; For a first subblock of the current block adjacent to both the first block edge and the second block edge, obtain a first weighted sum of first hybrid intra-predictors of first pixels of the first subblock, wherein the first weighted sum is obtained by summing (1) a predictor using an intra-prediction direction of the current block, (2) a predictor using an intra-prediction direction of a first adjacent neighboring block along the first block edge, and (3) a predictor using an intra-prediction direction of the neighboring block along the second block edge with corresponding weights, and the corresponding weights decrease as the distance of the first pixel from the neighboring block increases; For a second subblock of the current block that is adjacent to the first subblock but not adjacent to an edge of the second block, obtain a second weighted sum of second mixed intra predictors of second pixels of the second subblock, wherein the second weighted sum is obtained by summing (1) another predictor using the intra prediction direction of the current block and (2) a predictor using the intra prediction direction of a second adjacent neighboring block along the edge of the first block with corresponding weights; encoding the current block using the first hybrid intra predictor having the first weighted sum and the second hybrid intra predictor having the second weighted sum, wherein if a difference between an intra prediction direction of the current block and at least one of corresponding intra prediction directions of the neighboring blocks is greater than a threshold, only applying the hybrid intra predictor to encode the current block; and The intra prediction direction of the current block is encoded. 2 . The method of claim 1 , wherein for at least a portion of the second subblock adjacent to a block boundary between the first subblock and the second subblock, intra prediction does not consider an intra prediction mode of the first subblock.
3. The method of claim 1, wherein the plurality of neighboring blocks represent one or more neighboring sub-blocks that have been respectively subdivided from one or more larger neighboring blocks. 4 . The method of claim 1 , wherein if the height or width of the current block is greater than a second threshold, encoding the current block using a hybrid intra predictor is only applied.
5. A method for decoding video data, include: Obtaining the intra prediction direction of the current block to be decoded; Obtaining a plurality of adjacent blocks of the current block along a first block edge of the current block, wherein the first block edge has a first direction; Obtaining a neighboring block of the current block along a second block edge of the current block, wherein the second block edge has a second direction; Obtaining a plurality of sub-blocks of the current block based on the plurality of neighboring blocks along an edge of the first block; For a first subblock of the current block adjacent to both the first block edge and the second block edge, obtain a first weighted sum of first hybrid intra-predictors for first pixels of the first subblock, wherein the first weighted sum is obtained by summing (1) a predictor using an intra-prediction direction of the current block, (2) a predictor using an intra-prediction direction of a first adjacent neighboring block along the first block edge, and (3) a predictor using an intra-prediction direction of a neighboring block along the second block edge using corresponding weights, and the corresponding weights decrease as the distance of the first pixel from the neighboring block increases; For a second subblock of the current block that is adjacent to the first subblock but not adjacent to an edge of the second block, obtain a second weighted sum of second mixed intra predictors of second pixels of the second subblock, wherein the second weighted sum is obtained by summing (1) another predictor using the intra prediction direction of the current block and (2) a predictor using the intra prediction direction of a second adjacent neighboring block along the edge of the first block with corresponding weights; and The current block is decoded using the first hybrid intra-frame predictor having the first weighted sum and the second hybrid intra-frame predictor having the second weighted sum, wherein if the difference between the intra-frame prediction direction of the current block and at least one of the corresponding intra-frame prediction directions of the neighboring blocks is greater than a threshold, only the hybrid intra-frame predictor is applied to decode the current block. 6 . The method of claim 5 , wherein for at least a portion of the second subblock adjacent to a block boundary between the first subblock and the second subblock, intra prediction does not consider an intra prediction mode of the first subblock.
7. The method of claim 5, wherein the plurality of neighboring blocks represent one or more neighboring sub-blocks that have been respectively subdivided from one or more larger neighboring blocks. 8 . The method of claim 5 , wherein if the height or width of the current block is greater than a second threshold, only applying the hybrid intra predictor to decode the current block.
9. An apparatus for encoding video data, comprising one or more processors, wherein the one or more processors are configured to: Obtaining an intra prediction direction of a current block to be encoded; Obtaining a plurality of adjacent blocks of the current block along a first block edge of the current block, wherein the first block edge has a first direction; Obtaining a neighboring block of the current block along a second block edge of the current block, wherein the second block edge has a second direction; Obtaining a plurality of sub-blocks of the current block based on the plurality of neighboring blocks along an edge of the first block; For a first subblock of the current block adjacent to both the first block edge and the second block edge, obtain a first weighted sum of first hybrid intra-predictors of first pixels of the first subblock, wherein the first weighted sum is obtained by summing (1) a predictor using an intra-prediction direction of the current block, (2) a predictor using an intra-prediction direction of a first adjacent neighboring block along the first block edge, and (3) a predictor using an intra-prediction direction of the neighboring block along the second block edge with corresponding weights, and the corresponding weights decrease as the distance of the first pixel from the neighboring block increases; For a second subblock of the current block that is adjacent to the first subblock but not adjacent to an edge of the second block, obtain a second weighted sum of second mixed intra predictors of second pixels of the second subblock, wherein the second weighted sum is obtained by summing (1) another predictor using the intra prediction direction of the current block and (2) a predictor using the intra prediction direction of a second adjacent neighboring block along the edge of the first block with corresponding weights; encoding the current block using the first hybrid intra predictor having the first weighted sum and the second hybrid intra predictor having the second weighted sum, wherein if a difference between an intra prediction direction of the current block and at least one of corresponding intra prediction directions of the neighboring blocks is greater than a threshold, only applying the hybrid intra predictor to encode the current block; and The intra prediction direction of the current block is encoded. 10 . The apparatus of claim 9 , wherein for at least a portion of the second subblock adjacent to a block boundary between the first subblock and the second subblock, intra prediction does not consider an intra prediction mode of the first subblock.
11. The apparatus of claim 9, wherein the plurality of neighboring blocks represent one or more neighboring sub-blocks that have been respectively subdivided from one or more larger neighboring blocks. 12 . The apparatus of claim 9 , wherein if a height or a width of the current block is greater than a second threshold, encoding the current block using a hybrid intra predictor is only applied.
13. An apparatus for decoding video data, comprising one or more processors, wherein the one or more processors are configured to: Obtaining the intra prediction direction of the current block to be decoded; Obtaining a plurality of adjacent blocks of the current block along a first block edge of the current block, wherein the first block edge has a first direction; Obtaining a neighboring block of the current block along a second block edge of the current block, wherein the second block edge has a second direction; Obtaining a plurality of sub-blocks of the current block based on the plurality of neighboring blocks along an edge of the first block; For a first subblock of the current block adjacent to both the first block edge and the second block edge, obtain a first weighted sum of first hybrid intra-predictors of first pixels of the first subblock, wherein the first weighted sum is obtained by summing (1) a predictor using an intra-prediction direction of the current block, (2) a predictor using an intra-prediction direction of a first adjacent neighboring block along the first block edge, and (3) a predictor using an intra-prediction direction of the neighboring block along the second block edge with corresponding weights, and the corresponding weights decrease as the distance of the first pixel from the neighboring block increases; For a second subblock of the current block that is adjacent to the first subblock but not adjacent to an edge of the second block, obtain a second weighted sum of second mixed intra predictors of second pixels of the second subblock, wherein the second weighted sum is obtained by summing (1) another predictor using the intra prediction direction of the current block and (2) a predictor using the intra prediction direction of a second adjacent neighboring block along the edge of the first block with corresponding weights; and The current block is decoded using the first hybrid intra-frame predictor having the first weighted sum and the second hybrid intra-frame predictor having the second weighted sum, wherein if the difference between the intra-frame prediction direction of the current block and at least one of the corresponding intra-frame prediction directions of the neighboring blocks is greater than a threshold, only the hybrid intra-frame predictor is applied to decode the current block. 14 . The apparatus of claim 13 , wherein for at least a portion of the second subblock adjacent to a block boundary between the first subblock and the second subblock, intra prediction does not consider an intra prediction mode of the first subblock.
15. The apparatus of claim 13, wherein the plurality of neighboring blocks represent one or more neighboring sub-blocks that have been respectively subdivided from one or more larger neighboring blocks. 16 . The apparatus of claim 13 , wherein if a height or a width of the current block is greater than a second threshold, only decoding the current block using a hybrid intra predictor is applied.
