WO2019107997A1 - Method and device for processing video signal - Google Patents

Abstract

The present invention relates to a method and a device for processing a video signal on the basis of an intra mirror prediction (IMP) mode, and specifically, relates to a method and a device therefor, the method comprising the steps of: splitting a current block into two blocks, wherein the current block is split in a vertical direction on the basis of an IMP mode of the current block being a vertical directional mode, and the current block is split in a horizontal direction on the basis of the IMP mode of the current block being a horizontal directional mode; acquiring mode information indicating the intra prediction direction of the current block; on the basis of the acquired mode information and a reference sample of a neighboring block adjacent to the current block, acquiring a prediction value of a first block, among the two blocks; and on the basis of the prediction value of the specific block, acquiring a prediction value of a second block, among the two blocks.

Description

Method and apparatus for processing video signal

The present invention relates to a video processing method, and more particularly, to a video signal processing method using intra prediction and an apparatus therefor.

As the digital video processing technology rapidly develops, digital multimedia services using various media such as high-definition digital broadcasting, digital multimedia broadcasting and Internet broadcasting are being activated. Various high-definition digital broadcasting services are being developed and various service applications are being developed. Speed video processing techniques for video images are required. To this end, standards for coding video signals such as H.265 / HEVC (High Efficiency Video Coding) and H.264 / AVC (Advanced Video Coding) have been actively discussed.

It is an object of the present invention to provide a method and apparatus for efficiently processing a video signal.

It is another object of the present invention to improve coding efficiency and prediction performance by allowing intra prediction to be performed without dividing a current block into sub-blocks.

The technical problems to be solved by the present invention are not limited to the technical problems and other technical problems which are not mentioned can be understood by those skilled in the art from the following description.

There is provided, as a first aspect of the present invention, a method for decoding a video signal based on an Intra Mirror Prediction (IMP) mode, the method comprising dividing a current block into two blocks, The current block is divided in the vertical direction based on the IMP mode being the vertical direction mode, and the current block is divided in the horizontal direction based on the IMP mode of the current block being the horizontal mode; Obtaining mode information indicating an intra prediction direction of the current block; Obtaining a predicted value of a first block among the two blocks based on the acquired mode information and a reference sample of a neighboring block adjacent to the current block; Obtaining a predicted value of a second block among the two blocks based on a predicted value of the specific block; And reconstructing the current block based on the obtained predicted value.

As a second aspect of the present invention, there is provided an apparatus for decoding a video signal based on an Intra Mirror Prediction (IMP) mode, the apparatus comprising: a memory; And a processor coupled in operation to the memory, wherein the processor divides the current block into two blocks, and based on the IMP mode of the current block being in a vertical direction mode, the current block is divided vertically , The current block is divided horizontally based on the IMP mode of the current block being a horizontal direction mode; Acquiring mode information indicating an intra prediction direction of the current block; Obtain a predicted value of a first block among the two blocks based on the acquired mode information and a reference sample of a neighboring block adjacent to the current block; Obtain a predicted value of a second block among the two blocks based on the predicted value of the specific block; And reconstruct the current block based on the obtained predicted value.

The predicted value of the second block may be generated by symmetrically copying the predicted value of the first block.

The predicted value of the second block may be generated using a weighted sum based on the predicted value of the first block and the reference samples of the neighboring block adjacent to the current block.

Preferably, the first block is a right block of the two blocks and the second block is a right block of the two blocks, The predicted value of the right block is generated based on the reference samples of the upper neighboring block or the upper right neighboring block adjacent to the current block and the predicted value of the left block is generated symmetrically As shown in FIG.

Preferably, the first block is the left block of the two blocks and the second block is the left block of the two blocks, based on the mode information indicating the upper left direction and the IMP mode of the current block being the vertical direction mode. The predicted value of the left block is generated based on the reference samples of the upper neighbor block, the upper left neighbor block, or the left neighbor block adjacent to the current block, and the predicted value of the right block is generated based on the predicted value of the left block, Can be generated by symmetrically copying in units of units.

Preferably, the first block is the bottom block of the two blocks and the second block is the bottom block of the two blocks, based on the mode information indicating the left bottom direction and the IMP mode of the current block being the horizontal mode. And a predicted value of the lower block is generated based on a reference sample of a left neighboring block or a lower left neighboring block adjacent to the current block, and the predicted value of the upper block is a symmetric As shown in FIG.

Preferably, the first block is an upper block of the two blocks and the second block is an upper block of the two blocks, based on the mode information indicating a left upper direction and the IMP mode of the current block being a horizontal direction mode. And a predicted value of the upper block is generated based on a reference sample of a left neighbor block, a left upper neighbor block, or an upper neighbor block adjacent to the current block, and the predicted value of the lower block is a predicted value of the upper block, Can be generated by symmetrically copying in units of units.

Preferably, the first block is an upper block among the two blocks and the second block is an upper block and the second block is an upper block and a lower block, respectively, based on the mode information indicating a right upper direction and the IMP mode of the current block being a horizontal direction mode. And the predicted value of the upper block is generated based on the reference samples of the upper neighbor block or the upper right neighbor block adjacent to the current block and the predicted value of the lower block is a symmetric As shown in FIG.

Preferably, the first block is the left block of the two blocks and the second block is the leftmost block of the two blocks, based on the mode information indicating the lower left direction and the IMP mode of the current block being the vertical direction mode. And the predicted value of the left block is generated based on the reference samples of the left neighboring block or the left lower neighboring block adjacent to the current block and the predicted value of the right block is calculated symmetrically As shown in FIG.

The current block may be divided into a first block and a second block.

According to the present invention, a video signal can be efficiently processed.

In addition, according to the present invention, coding efficiency and prediction performance can be improved by performing intra prediction without dividing the current block into sub-blocks.

The effects obtained by the present invention are not limited to the above-mentioned effects, and other effects not mentioned can be clearly understood by those skilled in the art from the following description will be.

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of the specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

Figure 1 illustrates an encoder for encoding a video signal.

Figure 2 illustrates a decoder for decoding a video signal.

Figure 3 illustrates a segmented structure of a coding unit.

Figure 4 illustrates a quadtree-binary tree of the partitioning structure of a coding unit.

5 illustrates an intra prediction mode.

6 illustrates a signaling method for intra prediction mode.

Figure 7 illustrates an extended intra prediction mode.

FIG. 8 illustrates an MPM candidate for an extended intra-prediction mode.

9 and 10 illustrate the problem of the intra prediction method according to the prior art.

11 and 12 illustrate various cases in which the intra-mirror prediction mode of the present invention can be applied.

13 illustrates a comparison between the intra-mirror prediction mode of the present invention and the existing intra-prediction mode.

FIG. 14 illustrates an embodiment according to the intra-mirror prediction mode of the present invention.

15 illustrates an intra-mirror prediction method for the vertical mode of the present invention.

16 illustrates an intra-mirror prediction method for the horizontal direction mode of the present invention.

17 illustrates other examples of the intra-mirror prediction method of the present invention.

18 illustrates an intra-mirror prediction method and a method of reducing discontinuity.

19 illustrates a flowchart of a method for performing intra prediction according to the intra-mirror prediction mode of the present invention.

20 illustrates a vertical-horizontal merge method according to the present invention.

FIG. 21 illustrates an image processing apparatus to which the present invention can be applied.

The following description may be used in an image signal processing apparatus configured to encode and / or decode a video signal. Generally, a video signal refers to an image signal or a sequence of pictures that is perceptible to the eye, but the video signal is referred to herein as a sequence of bits representing a coded picture or a bit corresponding to a bit sequence Stream. A picture can refer to an array of samples and can be referred to as a frame, an image, or the like. More specifically, a picture may refer to a two-dimensional array of samples or a two-dimensional sample array. A sample may refer to the smallest unit of a picture and may be referred to as a pixel, a picture element, a pel, or the like. The sample may include a luminance (luma) component and / or a chrominance (chroma, color difference) component. In this specification, coding may be used to refer to encoding, or may be collectively referred to as encoding / decoding.

The picture may comprise at least one slice, and the slice may comprise at least one block. The slice may be configured to include an integer number of blocks for purposes such as parallel processing, resynchronization of decoding if the bitstream is corrupted due to data loss, etc., and each slice may be coded independently of each other. A block may include at least one sample and may refer to an array of samples. A block may have a size less than or equal to a picture. A block may be referred to as a unit. A picture that is currently coded (encoded or decoded) is referred to as a current picture, and a block that is currently coded (encoded or decoded) may be referred to as a current block. For example, in the case of the International Telecommunication Union Telecommunication Standardization Sector (ITU-T) H.265 standard (or HEVC (High Efficiency Video Coding) standard), various blocks constituting the picture may exist. (Or a coding tree unit CTU), a coding block CB (or a coding unit CU), a prediction block PB (or a prediction unit PU), a transform block TB Conversion unit TU) may exist.

The coding tree block refers to the most basic unit of a picture and can be divided into quad-tree type coding blocks to improve the coding efficiency according to the texture of the picture. The coding block may refer to a basic unit for performing coding, and intra coding or inter coding may be performed on a coding block basis. Intra-coding may refer to performing coding using intra-prediction, and intra-prediction may refer to performing prediction using samples included in the same picture or slice. Intercoding may refer to performing coding using inter prediction, and inter prediction may refer to performing prediction using samples included in pictures that are different from the current picture. Blocks coded using intra coding or blocks coded in intra prediction mode may be referred to as intra blocks, and blocks coded using inter coding or blocks coded in inter prediction mode may be referred to as inter blocks. In addition, a coding mode using intra prediction can be referred to as an intra mode, and a coding mode using inter prediction can be referred to as an inter mode.

The prediction block may refer to a basic unit for performing prediction. The same prediction can be applied to one prediction block. For example, in the case of inter prediction, the same motion vector can be applied to one prediction block. The conversion block may refer to a basic unit for performing the conversion. The conversion may refer to the operation of converting the samples of the pixel domain (or of the spatial domain or of the time domain) into the transform coefficients of the frequency domain (or transform coefficient domain), or vice versa. In particular, the operation of converting the transform coefficients of the frequency domain (or transform coefficient domain) into samples of the pixel domain (or spatial domain or time domain) may be referred to as inverse transform. For example, the transform may include discrete cosine transform (DCT), discrete cosine transform (DST), Fourier transform, and the like. The prediction block and / or the transform block may be set to the same size as the coding block, in which case prediction may be performed and / or conversion may be performed on a coding block basis.

In this specification, a coding tree block (CTB) may be mixed with a coding tree unit (CTU), a coding block (CB) may be mixed with a coding unit (CU), and a prediction block , And the conversion block PB can be mixed with the conversion unit PU.

1 shows a schematic block diagram of an encoder in which the encoding of a video signal is performed, in which the present invention is applied.

1, the encoder 100 includes an image divider 110, a transform unit 120, a quantization unit 130, an inverse quantization unit 140, an inverse transform unit 150, a filtering unit 160, A picture buffer (DPB) 170, an inter prediction unit 180, an intra prediction unit 185, and an entropy encoding unit 190.

The image divider 110 may divide an input image (or a picture, a frame) input to the encoder 100 into one or more processing units. For example, the processing unit may be a coding tree unit (CTU), a coding unit (CU), a prediction unit (PU), or a transform unit (TU).

The encoder 100 may generate a residual signal by subtracting the prediction signal output from the inter prediction unit 180 or the intra prediction unit 185 from the input image signal, 120.

The conversion unit 120 may perform a conversion based on the residual signal to generate a transform coefficient. For example, the transform technique may include at least one of Discrete Cosine Transform (DCT), Discrete Sine Transform (DST), Karhunen-Loeve Transform (KLT), Graph-Based Transform (GBT), or Conditionally Non-linear Transform . More specifically, an integer-based DCT having a size of 4 × 4 to 32 × 32 can be used, and 4 × 4, 8 × 8, 16 × 16, and 32 × 32 transforms can be used. GBT refers to the transformation obtained from this graph when graphing the relationship information between pixels. The CNT means a transform obtained by generating a prediction signal using all previously reconstructed pixels and obtaining based thereon. In addition, the conversion process may be applied to a pixel block having the same size of a square, or to a block having a variable size other than a square.

The quantization unit 130 quantizes the transform coefficients and transmits the quantized transform coefficients to the entropy encoding unit 190. The entropy encoding unit 190 entropy-codes the quantized signal and outputs the quantized signal as a bitstream. For example, entropy encoding may be performed based on fixed length coding (FLC), variable length coding (VLC), or arithmetic coding. More specifically, context-based adaptive binary arithmetic coding (CABAC) based on arithmetic coding, Exp-Golomb coding based on variable length coding, and fixed length coding can be applied.

The quantized signal output from the quantization unit 130 may be used to generate a prediction signal. For example, the quantized signal can be reconstructed by applying inverse quantization and inverse transformation through the inverse quantization unit 140 and the inverse transform unit 150 in the loop. The restored signal can be generated by adding the restored residual signal to the prediction signal output from the inter prediction unit 180 or the intra prediction unit 185. [

The filtering unit 160 applies filtering to the restored signal and outputs the restored signal to the playback apparatus or the decoded picture buffer 170. For filtering, for example, a deblocking filter, a sample adaptive offset (SAO) filter may be applied. The filtered signal transmitted to the decoded picture buffer 170 may be used as a reference picture in the inter-prediction unit 180. As described above, not only the picture quality but also the coding efficiency can be improved by using the filtered picture as a reference picture in the inter prediction mode.

The decoded picture buffer 170 may store the filtered picture for use as a reference picture in the inter prediction unit 180. [

The inter-prediction unit 180 performs temporal prediction and / or spatial prediction to remove temporal redundancy and / or spatial redundancy with reference to the reconstructed picture. At this time, in order to reduce the amount of motion information transmitted in the inter prediction mode, motion information can be predicted based on the correlation of motion information between the neighboring block and the current block.

The intra predictor 185 can predict the current block by referring to samples in the vicinity of the current block. The intraprediction unit 185 may perform the following procedure to perform intraprediction. First, a reference sample necessary for generating a prediction signal can be prepared. Then, a prediction signal can be generated using the prepared reference sample. Thereafter, the prediction mode is encoded. At this time, reference samples can be prepared through reference sample padding and / or reference sample filtering. Since the reference samples have undergone prediction and reconstruction processes, quantization errors may exist. Therefore, a reference sample filtering process can be performed for each prediction mode used for intraprediction to reduce such errors.

The prediction signal generated through the inter prediction unit 180 or the intra prediction unit 185 may be used to generate a reconstructed signal or may be used to generate a residual signal.

Fig. 2 shows a schematic block diagram of a decoder in which decoding of a video signal is performed, according to an embodiment to which the present invention is applied.

2, the decoder 200 includes an entropy decoding unit 210, an inverse quantization unit 220, an inverse transform unit 230, a filtering unit 240, a decoded picture buffer (DPB) 250 ), An inter-prediction unit 260, and an intra-prediction unit 265.

The reconstructed video signal output through the decoder 200 can be reproduced through the reproduction apparatus.

The decoder 200 may receive the signal output from the encoder 100 of FIG. 1, and the received signal may be entropy-decoded through the entropy decoding unit 210.

The inverse quantization unit 220 obtains a transform coefficient from the entropy-decoded signal using the quantization step size information.

The inverse transform unit 230 performs inverse transform based on the inversely quantized transform coefficients to obtain a residual signal.

The restored signal is generated by adding the obtained residual signal to the prediction signal output from the inter-prediction unit 260 or the intra-prediction unit 265.

The filtering unit 240 applies filtering to the restored signal and outputs the restored signal to the playback apparatus or the decoded picture buffer unit 250. The filtered signal transmitted to the decoding picture buffer unit 250 may be used as a reference picture in the inter prediction unit 260.

The embodiments described in the filtering unit 160, the inter-prediction unit 180 and the intra-prediction unit 185 of the encoder 100 respectively include the filtering unit 240 of the decoder, the inter-prediction unit 260, The same can be applied to the intra prediction unit 265.

3 is a diagram for explaining a division structure of a coding unit.

The encoder and the decoder can divide one image (or picture) into a rectangular Coding Tree Unit (CTU) unit and perform encoding and decoding in units of CTU.

One CTU can be partitioned based on a quadtree (QT) structure. For example, one CTU can be divided into four units, each of which has a square shape and whose length is reduced by half. This division of the QT structure can be performed recursively.

Referring to FIG. 3, the root node of the QT may be associated with a CTU. The QT may be divided until it reaches a leaf node, and the leaf node may be referred to as a coding unit (CU). The CTU corresponds to the root node and has the smallest depth (i.e., level 0) value. Depending on the characteristics of the input image, the CTU may not be divided. In this case, the CTU corresponds to the CU.

In the example of FIG. 3, the CTU can be divided into QT forms, with the result that lower nodes having a depth of level 1 can be created. A node that is not further divided in a child node having a depth of level 1 (i.e., a leaf node) corresponds to a CU. For example, CU (a), CU (b), and CU (j) corresponding to nodes a, b, and j in FIG. 3B are once partitioned in the CTU and have a level 1 depth.

For one CU, information indicating whether or not the CU is divided may be transmitted to the decoder. For example, the information may be defined as a segmentation flag, and may be expressed as syntax information "split_cu_flag ". The split flag may be included in all CUs except for the minimum size CU. For example, when the value of the division flag is '1', the corresponding CU is divided into four CUs again. If the value of the division flag is '0', the corresponding CU is not divided any more, .

In the example of FIG. 3, the CU division process is described as an example. However, the QT structure described above can be applied to the division process of a transform unit (TU) as a basic unit for performing a transformation.

The TU may be hierarchically partitioned from the CU to be coded into a QT structure. For example, a CU may correspond to a root node of a tree for a transformation unit (TU).

Since the TU is divided into QT structures, the TUs divided from the CUs can be further divided into smaller lower TUs. For example, the size of the TU may be set to any one of 32x32, 16x16, 8x8, and 4x4, but the present invention is not limited thereto. In case of a high resolution image, the size of the TU may be increased or varied.

For one TU, information indicating whether the TU is divided may be communicated to the decoder. For example, the information may be defined as a split conversion flag, and may be represented by syntax information "split_transform_flag ".

4 is a diagram for explaining a quadtree-binary tree among the division structure of the coding unit.

The encoder and the decoder can divide one image (or picture) into a rectangular Coding Tree Unit (CTU) unit and perform encoding and decoding in units of CTU.

Compared with the example of FIG. 3, in the example of FIG. 4, one CTU can be partitioned based on a quadtree and binary tree (BT) structure. For example, one CTU can be divided into four units, each of which has a square shape, the length of each side is reduced by half, or divided into two units, each of which has a rectangular shape and whose width or height length is reduced by half. This division of the QT BT structure can be performed recursively.

Referring to FIG. 4, the root node of the QT may be associated with a CTU. The QT can be partitioned until it reaches the QT leaf node, and the leaf node of the QT can be partitioned into BT and can be partitioned until it reaches the BT leaf node.

In the example of FIG. 4, the CTU corresponds to a root node and has the smallest depth (i.e., level 0) value. Depending on the characteristics of the input image, the CTU may not be divided. In this case, the CTU corresponds to the CU.

The CTU can be divided into QT types, and QT leaf nodes can be divided into BT types. As a result, child nodes having a depth of level n can be generated. A node that is no longer subdivided (ie, a leaf node) in a child node having a depth of level n corresponds to a CU.

For one CU, information indicating whether or not the CU is divided may be transmitted to the decoder. For example, the information may be defined as a segmentation flag, and may be expressed as syntax information "split_cu_flag ". Further, information indicating whether or not to be divided into BT at the QT leaf node may be transmitted to the decoder. For example, the information may be defined as a BT fragment flag, and may be represented by syntax information " bt_split_flag ". In addition, if BT is divided by bt_split_flag, the BT split shape can be transmitted to the decoder such that it is divided into a rectangular shape having a half-size width or a rectangular shape having a half-height height. For example, the information may be defined as a BT split mode and may be expressed as syntax information " bt_split_mode ".

5 illustrates an intra prediction mode.

The intra prediction mode indicates various prediction modes depending on the value. For example, the value of the intra-prediction mode may correspond to the intra-prediction mode as illustrated in Table 1. [

In Table 1, INTRA_PLANAR indicates an intraplanar prediction mode. The reconstructed sample of the upper neighboring block adjacent to the current block, the reconstructed sample of the left neighboring block, the lower left (lower) -left) restoration sample of the neighboring block, and a restoration sample of the right-upper neighboring block to obtain a predicted value of the current block. INTRA_DC indicates an intra DC (Direct Current) prediction mode, and indicates a mode for obtaining a predicted value of a current block by using an average of restoration samples of the left neighboring block and restoration samples of the upper neighboring block. INTRA_ANGULAR2 to INTRA_ANGULAR34 represent intra-angular prediction modes and indicate a mode for obtaining a predicted value of a current sample using a reconstructed sample of a neighboring block located in a direction of a specific angle with respect to a current sample in the current block (e.g., 5). If there is no actual sample in the direction of a specific angle, interpolation or padding may be performed on the neighboring reconstructed samples to generate a virtual sample for the corresponding direction to obtain a predicted sample.

Intra prediction mode can be derived for each coding block, but intra prediction can be performed on a coding block or a transform block basis. A reconstructed sample existing in a neighboring block of a current block in the current picture for intra prediction can be referred to, and a sample to be referred to for intra prediction can be referred to as a reference sample. In intra prediction, the prediction of the current block is performed based on the prediction mode induced. Depending on the prediction mode, the reference sample used for prediction and the concrete prediction method may be different.

The encoder and decoder can determine if neighboring samples of the current block can be used for prediction and construct reference samples to use for prediction. For example, if the size of the current block is nSxnS, the neighbor samples of the current block in intra prediction are nS reference samples adjacent to the left (or left boundary) of the current block, bottom-left , NS reference samples adjacent to the top (or top boundary) of the current block, nS reference samples adjacent to the top-right, top (left top) of the current block (top -left) can be used as a reference sample. If some of the surrounding samples of the current processing block are not available, the encoder and decoder may perform interpolation or padding based on samples not available with the available samples to construct reference samples for prediction have. Filtering of the reference samples may be performed based on the intra prediction mode.

The encoder and decoder may generate predicted values for the current block based on the intra prediction mode and the reference samples. Specifically, the encoder determines an intra-prediction mode using reference samples or filtered reference samples based on bit-rate-distortion (RD) optimization, and encodes the syntax information indicating the determined intra-prediction mode into a bitstream A prediction value for the current block is generated based on the determined intra prediction mode, and the current block is encoded using the generated prediction value. The decoder can restore the current block based on the intra prediction mode and the predicted value generated after generating the prediction value for the current block based on the reference samples. That is, the decoder can generate a prediction value for the current block based on the intra prediction mode derived in the intra prediction mode derivation step, the reference samples obtained through the reference sample construction step, and the reference sample filtering step.

6 illustrates a signaling method for intra prediction mode.

Generally, if an image is divided into blocks, a current block and a neighboring block to be coded may have similar image characteristics. In the intra prediction mode, there is a high probability that the current block and the neighboring block have the same or similar intra prediction modes. Thus, the encoder or decoder may use the prediction mode of the neighboring block to encode or derive the prediction mode of the current block.

First, if the neighboring block is intra-coded, the encoder can confirm or derive the prediction mode of the neighboring block (S610). For example, the prediction mode of the current block can be determined based on the prediction mode of the left neighboring block and the prediction mode of the upper neighboring block, and the prediction mode of the neighboring block can be determined as MPM (Most Probable Mode). The MPM may refer to a mode used for improving the coding efficiency in consideration of the similarity between the current block and the neighboring block in intra prediction mode coding. Deciding on an MPM may also be expressed as listing up the MPM (most probable modes) candidate (or MPM list).

The encoder can check whether the prediction mode of the left neighboring block and the prediction mode of the upper neighboring block are the same (S620).

If the prediction mode of the left neighbor block and the prediction mode of the upper neighbor block are not the same, the first MPM may be set to the prediction mode of the left neighbor block, and the second MPM may be set to the prediction mode of the upper neighbor block , And the third MPM may be set to any one of an intra-plane mode, an intra-DC mode, and an intra-vertical mode (S630).

If the prediction mode of the left neighboring block is equal to the prediction mode of the upper neighboring block, the encoder can check whether the prediction mode of the left neighboring block is smaller than 2 (S640).

If the prediction mode of the left neighbor block is smaller than 2, the first MPM can be set to the intra plane mode, the second MPM can be set to the intra DC mode, and the third MPM can be set to the intra vertical mode (S650).

If the prediction mode of the left neighbor block is not smaller than 2, the first MPM may be set to the prediction mode of the left neighbor block, the second MPM may be set to (prediction mode -1 of the left neighbor block) The third MPM may be set to (prediction mode of the left neighboring block +1) (S660).

The encoder can determine whether the optimal intra prediction mode to be applied to the current block belongs to the MPM candidate configured in advance. If the intra prediction mode of the current block belongs to the MPM candidate, the encoder can encode the MPM flag and the MPM index into a bit stream. The MPM flag may indicate whether the intra prediction mode of the current block is derived from the surrounding intra prediction block (i. E., The intra prediction mode of the current block falls within the MPM). The MPM index may indicate which MPM mode is applied as the intra prediction mode of the current block among the MPM candidates.

On the other hand, if the intra-prediction mode of the current block does not belong to the MPM candidate, the encoder can encode the intra-prediction mode of the current block into a bitstream.

Figure 7 illustrates an extended intra prediction mode.

It is possible to extend 35 intra prediction modes to 67 intra prediction modes for intra-coding and more accurate prediction of high-resolution images. The arrows indicated by dashed lines in Fig. 7 represent the newly added 32 intra-angle prediction modes, for example, the values of the extended intra-prediction modes can be matched with the intra-prediction modes as illustrated in Table 2. [

The intra-plane prediction mode (INTRA_PLANAR) and the intra-DC prediction mode (INTRA_DC) in the extended intra-prediction mode are the same as those of the intra-plane prediction mode and the intra-prediction mode. The newly added 32 directional modes can be applied to all block sizes and can be applied to intra coding of luminance components and chrominance components.

FIG. 8 illustrates an MPM candidate for an extended intra-prediction mode.

As the number of intra prediction modes is increased to 67, the number of MPM (Most Probable Mode) derived from neighboring blocks is increased from 3 to 6 and the method of constructing the MPM list can be changed for efficient coding of the intra prediction mode .

The method of constructing the MPM list based on the six MPM candidates proceeds largely in the following three steps.

- Using the surrounding intra prediction mode

- Use derived intra prediction mode

- Use default intra prediction mode

An MPM list containing six MPM candidates is first generated using the surrounding intra prediction mode. As illustrated in FIG. 8, the neighboring blocks (AL, A, AR, L, BL) of the current intraprediction block (i.e., current block) are searched to determine the intra prediction mode as six MPM candidate lists . At this time, duplicate inspection is performed to exclude the same intra prediction mode, and a new intra prediction mode is added to the six MPM candidate lists. For example, the search order of neighboring blocks can be proceeded in the order of L -> A -> intra-plane prediction mode -> intra DC prediction mode -> BL -> AR -> AL. If the six MPM candidate lists are completed using the surrounding intra prediction mode, the candidate list generation process is terminated.

If the neighboring intra-prediction mode is not used to complete the six MPM candidate lists, a candidate list is constructed using the derived intra-prediction mode. The derived intra-prediction mode is generated by adding -1 or +1 to the intra-prediction mode already existing in the candidate list, and the generated intra-prediction mode is added to the candidate list. At this time, duplicate checking is performed to exclude the same mode, and added to the candidate list when the mode is a new mode.

If the candidate list is still not completed, the candidate list is finally constructed using the default intra prediction mode. In the case of the default intra prediction mode, it may be one of {Vertical, Horizontal, Intra_Angular2, Diagonal} modes. In the extended intra-prediction mode, Vertical indicates an intra-vertical prediction mode (e.g., Intra_Angular50), Horizontal indicates an intra-horizontal prediction mode (e.g., Intra_Angular18), and Diagonal indicates a diagonal intra-prediction mode (e.g. Intra_Angular34 or Intra_Angular66) . Perform duplicate checking in order and add to the candidate list.

9 and 10 illustrate the problem of the intra prediction method according to the prior art.

As described with reference to FIG. 5, in the conventional technique (e.g., the HEVC standard), 33 directional prediction methods (e.g., intra-angle prediction mode) and two non-directional prediction methods Mode, and intra-DC prediction mode), a predictor (predictor or predicted block) of the current block is generated using a total of 35 prediction methods. At this time, prediction using the neighboring reference samples (e.g., upper neighbor reference samples, upper right neighbor reference samples, upper left neighbor sample, left neighbor reference samples, lower left reference samples) to predict the current block After the sample is generated, a predicted value of the current block is generated by copying the predicted sample generated along the predicted direction.

Referring to FIG. 9, as described above, the intraprediction value generation method copies a predicted sample generated according to a prediction direction, and thus has a shape as shown in FIG. 9 (a) It is difficult to accurately predict the block. Therefore, as illustrated in FIG. 9 (b), in the case of a block having a shape that advances in one direction and is bent in another direction in the middle, the block is divided into smaller blocks, and prediction is performed. In this case, overhead bits such as mode information for each block may be generated, resulting in a problem of poor coding efficiency.

For example, as described with reference to FIG. 6, for each block, information indicating whether the intra-prediction mode is derived from the surrounding intra-predicted block (e.g., MPM flag), the intra- (E.g., MPM index) indicating the intra-prediction mode of the corresponding block in the MPM candidate list if it is derived from the neighboring block, and an intra-prediction mode indicating the intra-prediction mode of the corresponding block if the intra- Information can be signaled through the bitstream. Therefore, when the current block is divided into four blocks, syntax information for signaling the intra-prediction mode of each block is needed more. Therefore, in the example of FIG. 11, the same / similar prediction performance is achieved without dividing the current block The coding efficiency for the current block can be improved at least four times.

In the method 1 of the present invention, Intra Mirror Prediction (IMP) method is proposed in which predicted values can be efficiently generated without performing block division on a block having a shape that is turned in one direction illustrated in FIG. 9 and is bent in another direction do.

Figure 10 shows another problem of intra prediction according to the prior art.

Referring to FIG. 10, when different prediction directions exist in one block, it is difficult to generate accurate prediction blocks in the existing intra prediction mode. For example, as illustrated in FIG. 10, when different intra prediction mode components (e.g., an intra-vertical prediction mode and an intra-horizontal prediction mode) exist in each block (blocks 1 & cir & In the conventional intraprediction method, a block is divided into smaller blocks, and prediction is performed. In this case as well, as described with reference to FIG. 9, overhead bits such as mode information for each of the divided blocks may be generated, which may result in a problem of poor coding efficiency.

In the method 2 of the present invention, when different intra prediction mode components (for example, an intra prediction mode or an intra prediction mode) exist in one block illustrated in FIG. 10, a vertical- We propose a Vertical-Horizontal Merge Prediction (VHMP) method.

Method 1: Intra  Mirror Prediction Intra  Mirror Prediction, IMP) method

The method 1 of the present invention proposes a method of generating a predicted value by using a reference sample when encoding / decoding is performed in the intra-mirror prediction mode. The proposed intra-mirror prediction method can efficiently encode / decode a block whose prediction direction is reversed in the middle as in the example of FIG. 9 (a), but it can be used limitedly as shown in FIG.

11 illustrates various cases in which the intra-mirror prediction mode of the present invention can be applied.

Referring to FIG. 11, prediction values can be generated using the intra-mirror prediction method in the case of the example of FIG. 11 (a) and the example of FIG. 11 (c) in which a broken block is attached to a usable reference sample. On the other hand, in the case of the example of FIG. 11 (b) and the example of FIG. 11 (d) where the broken blocks are separated from the usable reference samples, the effect of the intra-mirror prediction method is reduced, The intra-mirror prediction method may not be applied.

The example of FIG. 9 can be specifically shown as shown in FIG. In Fig. 12, one rectangle represents one pixel. That is, FIG. 12 corresponds to a case where an intermediate-folded block is attached to the upper reference sample as shown in FIG. 11 (a). For convenience of explanation, a method of generating prediction values by applying intra-mirror prediction based on the example of FIG. 12 is described, but the present invention can be applied to other examples which are not illustrated in FIG. 12, in the same way. Also, for convenience of description herein, it is assumed that the current block is 8x8 block as illustrated in Fig. 12, but the present invention can be applied without limitation to the size of the current block.

13 illustrates a comparison between the intra-mirror prediction mode of the present invention and the existing intra-prediction mode.

Referring to FIG. 13A, it is assumed that the current block performs intra prediction in the direction of the arrow, but the present invention can be applied to the same or similar case even when another intra-angle prediction mode is applied. 13 (b) illustrates predicted values obtained by applying existing intra-prediction based on the intra-angle prediction mode corresponding to the arrow direction shown in FIG. 13 (a), and FIG. 13 (c) The predicted values obtained by applying the intra-mirror prediction method are illustrated.

Referring to FIG. 13 (b), prediction is performed as shown in FIG. 13 (b) by performing the intra prediction in the direction of the arrow with respect to the intermediate folded block as in the example of FIG. However, the predicted value generated in Fig. 13 (b) is much different from the block in Fig. 13 (a). That is, if the existing intra prediction method is applied to the example of FIG. 12, prediction can not be performed accurately. In this case, as described with reference to FIG. 9 (b), in the conventional intraprediction encoding / decoding, the current block is divided into smaller blocks and intra prediction is performed again.

13 (c), prediction values are generated by applying the intra-mirror prediction method according to the method 1 of the present invention. The prediction values generated using the intra-mirror prediction method are the same when they are compared with the block of FIG. 13 (a). 11A and 11C (when the intra-mirror prediction method can be used), if the intra-mirror prediction method is used, (Or a reduction in coding efficiency) while increasing prediction accuracy.

FIG. 14 illustrates an embodiment according to the intra-mirror prediction mode of the present invention.

Referring to FIG. 14 (a), the current block may have a size of 2Nx2N. In the example of FIG. 14A, it is assumed that N is 4 and the intra-angle prediction mode in the arrow direction is applied, but the present invention is not limited thereto. (Eg, I, J, K, L, and H) of the top neighboring block adjacent to the current block and the reconstructed samples (eg, E, F, M, N, O, P) may be used as a reference sample. (E.g., A, B, C, and D) of the upper neighbor block adjacent to the current block, restoration samples (e.g., Q) of the upper left neighbor block, It will be appreciated that a reconstructed sample (e.g., R, S, T, U, V, W, X, Y) of a left lower neighbor block may be used as a reference sample.

When the intra-mirror prediction mode of the present invention is applied, the current block may be divided vertically or horizontally. For example, in the case where the current block is divided in the vertical direction, the current block may be bisected by rectangular blocks having a larger vertical size. When the block is divided in the horizontal direction, the current block is divided into rectangular blocks, . In this specification, a case in which the current block is divided in the vertical direction to apply the intra-mirror prediction mode is referred to as a vertical direction mode, and a case in which the current block is divided in the horizontal direction to apply the intra-mirror prediction mode is referred to as a horizontal direction mode . Therefore, the vertical direction mode of the present invention is different from the conventional intra vertical prediction mode (e.g., the intra prediction mode 26 of FIG. 5, INTRA_ANGULAR 26), and the horizontal direction mode of the present invention is an intra- (E.g., the 10th intra prediction mode in FIG. 5, INTRA_ANGULAR10).

Syntax information indicating whether the intra-mirror prediction mode is applied to the current block can be signaled through the bitstream. And / or syntax information indicating whether the intra-mirror prediction mode of the current block is the vertical direction mode or the horizontal direction mode can be signaled through the bit stream.

In the example of FIG. 14, it is assumed that the vertical direction mode is applied in the intra-mirror prediction mode, but the same / similar principle can be applied even when the horizontal direction mode is applied.

14B, when the prediction mode of the current block is the vertical direction mode, the current block (e.g., 2Nx2N) is divided into two Nx2N (vertically bisected), and the right block (e.g., Nx2N) can be generated using the existing intra prediction method. More specifically, when the prediction mode of the current block is the vertical direction mode and the intra-angle prediction mode in the upper right diagonal direction is applied to the right block of the current block, the prediction value for the right block of the current block is . ≪ / RTI >

Referring to FIG. 14C, a predicted value can be generated by symmetrically copying a predicted value of a right block (for example, Nx2N) already generated for a left block (e.g., Nx2N) of the current block. The method of symmetrically copying the predicted value of the right block will be described in detail with reference to FIG. 15 and FIG.

15 illustrates an intra-mirror prediction method for the vertical mode of the present invention.

Referring to FIG. 15, the intra-mirror prediction for the vertical direction mode copies symmetrically the predicted sample values of the predicted right block to the left block. More specifically, in the case of the vertical direction mode, predicted values obtained on the basis of the existing intraprediction method are copied symmetrically to different blocks in units of columns to generate predicted values of the other partial blocks. For example, as illustrated in FIG. 15, first, the sample value of the first column of the right block is copied to the first column of the left block. Copy the sample value for column 2 in the right block to column 2 in the left block. In this way, the sample values of columns ③ and ④ of the left block are copied to columns ③ and ④ of the right column to generate predicted values of the left block. In the example of FIG. 15, the intra-mirror prediction method for the vertical direction mode is described using the 8x8 block. Intra-mirror prediction can also be performed by applying the same to block sizes other than 8x8.

16 illustrates an intra-mirror prediction method for the horizontal direction mode of the present invention.

16, when a horizontal mode is applied, the current block (e.g., 2Nx2N) may be divided into two 2NxN (bisectors), and the lower block of the current block (e.g., 2NxN) The prediction value can be obtained by applying the existing intraprediction method based on the intra-angle prediction mode. More specifically, when the prediction mode of the current block is the horizontal direction mode and the intra-angle prediction mode in the lower left diagonal direction is applied to the lower block of the current block, the prediction value for the lower block of the current block is . ≪ / RTI >

The intra-mirror prediction for the horizontal direction mode generates a prediction block by symmetrically copying the prediction sample value of the lower-stage block in which the prediction has been performed, to the upper block. More specifically, in the case of the horizontal direction mode, predicted values obtained on the basis of the existing intraprediction method are copied symmetrically to another block in units of rows to generate predicted values of other block. For example, as illustrated in FIG. 16, the sample value for the row 1 in the lower block is copied to the row 1 in the upper block. Copy the sample value for row 2 in the lower block to row 2 in the upper block. In this way, the sample values of row 3 and row 4 of the lower block are copied to rows 3 and 4 of the upper block to generate a prediction block.

In FIGS. 15 and 16, the intra-mirror prediction method for the horizontal direction mode and the vertical direction mode has been described. The intra-mirror prediction method of the present invention can also be applied to various other direction modes.

17 illustrates other examples of the intra-mirror prediction method.

17, a straight arrow indicates a prediction direction for applying an existing intra-prediction mode, a shaded block indicates a prediction value generated earlier by using an existing intra-prediction method for intra-mirror prediction, Represents an operation of symmetrically copying the predictive value generated first. In the case of FIGS. 17A and 17B, since the mode is the vertical direction, block division occurs in the vertical direction, and prediction value copying is performed in units of columns. In the case of FIGS. 17 (c) and 17 (d), since the mode is the horizontal direction, block division occurs in the horizontal direction and prediction value copying is performed row by row. As illustrated in Figs. 11 (b) and 11 (d), both the vertical direction mode and the horizontal direction mode can be performed for the same intra-prediction direction (or intra-angle prediction mode) It is possible to perform intra-mirror prediction by determining or applying an appropriate direction mode from the vertical direction mode and the horizontal direction mode.

The intra-mirror prediction method generates prediction values for one block in a current block using an existing intra prediction method, and generates predicted values of blocks in the opposite block in the current block using the prediction values. Therefore, if another block is generated by copying the generated block first, discontinuity may occur between the sample of another block and the reference sample of the neighboring block adjacent thereto. In order to eliminate such discontinuity, when the sample value of the reference block and the adjacent block are copied, blending or weighted sum may be applied to reduce the discontinuity.

18 shows an intra-mirror prediction method and a method for reducing discontinuity. In FIG. 18, the thick straight arrow indicates the direction to which the blending or weighted sum is applied in each reference sample direction, the shaded block indicates the predicted value generated earlier by using the existing intraprediction method for intra-mirror prediction, The arrow indicates the direction of symmetrically copying the predicted value generated first.

In the example of FIG. 18, the intra-mirror prediction method described with reference to FIG. 17 is applied, and the predicted value of the dotted block based on the predicted value of the reference sample of the adjacent neighboring block and the shaded block Blending or weighted sum. For example, for boundary samples adjacent to a reconstructed block in a block filled with a point, a prediction sample is generated using a weighted sum or average based on the predicted value of the shaded block and a sample of the neighboring block adjacent to the boundary sample And a predicted sample can be generated by copying the predicted value of the shaded block to another sample in the block filled with the point.

19 illustrates a flowchart of a method for performing intra prediction according to the intra-mirror prediction mode of the present invention. The method of FIG. 19 can be performed by the encoder 100 and the decoder 200 (e.g., intra prediction unit 185 of FIG. 1, intra prediction unit 265 of FIG. 2) The decoder may be referred to as a device.

In step S1910, the device can divide the current block into at least two blocks. If the intra-mirror prediction mode is the vertical direction mode, the current block can be divided into two blocks in the vertical direction, and if the intra-mirror prediction mode is the horizontal direction mode, the current block can be divided into two blocks in the horizontal direction (See, for example, Figs. 14, 15, 17 (a), 17 (b), 18 (a), 18 (b) and related description).

In the case of the encoder, the intra-mirror prediction mode of the current block can be determined as a mode showing the best performance in the bit rate-distortion (RD) among the vertical direction mode and the horizontal direction mode. The encoder may entropy encode information (e.g., flag information) indicating a determined direction mode among the vertical direction mode and the horizontal direction mode into a bit stream. On the other hand, in the case of a decoder, information indicating a vertical direction mode and a horizontal direction mode (e.g., flag information) is acquired through a bitstream, and an intra-mirror prediction mode for the current block is determined based on a direction mode indicated by the information Can be applied.

In step S1920, the apparatus can obtain the intra prediction mode information on the current block. Specifically, the apparatus can acquire an intra-angle prediction mode for applying an existing intra-prediction method in the intra-mirror prediction mode. When the intra-mirror prediction mode is applied to the current block, the current block includes at least two blocks, but the intra prediction mode information can be signaled only to the current block, rather than being signaled for each of the two blocks. One intra prediction mode is obtained for the current block and the intra prediction mode is not obtained for each of the two blocks.

For example, in the case of an encoder, intra prediction is performed on a current block by applying all possible intra prediction directions (e.g., FIGS. 5 and 7), and then, based on the RD optimization, (E.g., an intra-angle prediction mode) corresponding to the determined intra-prediction direction. In addition, the encoder may encode the acquired intra prediction mode information into a bitstream through entropy encoding (e.g., see FIG. 6 and related description). On the other hand, in the case of a decoder, it is possible to obtain the intra prediction mode information on the current block from the bitstream and determine the intra prediction direction for the current block. For example, based on the method described with reference to FIG. 6, the decoder may obtain intra prediction mode information.

In step S1930, the apparatus can obtain the predicted value of the first block among the two blocks included in the current block, based on the intra prediction mode information acquired in step S1920 and the reference sample of the neighboring block adjacent to the current block. For example, as described with reference to FIG. 14 (b), the apparatus compares the reference samples in the intra-prediction direction corresponding to the intra-prediction modes acquired in step S1920 and the reference samples in the intra- The predicted value of the first block can be obtained.

In step S1940, the apparatus may obtain the predicted value of the second block in the current block based on the predicted value of the first block obtained in step S1930. More specifically, the predicted value of the second block may be generated by symmetrically copying the predicted value of the first block obtained in step S1930 (e.g., see FIG. 14 (c), FIG. 15 to FIG.

18, the predicted value of the second block may be generated using the weighted sum based on the predicted value of the first block obtained in step S1930 and the reference sample of the neighboring block adjacent to the current block .

For example, in steps S1930 and S1940, the intra-prediction mode information of the current block indicates the upper right direction and the intra-mirror prediction mode of the current block is the vertical direction mode, the first block is selected from two blocks The second block may be determined as the left block among the two blocks in the current block and the predicted value of the right block is generated based on the reference samples of the upper neighbor block or the upper right neighbor block adjacent to the current block, The prediction value is generated by symmetrically copying the predicted value of the right block in units of columns (for example, see Figs. 15 and 17 (a)), or the prediction value of the right block and the predicted value of the upper neighbor block and / May be generated using a weighted sum based on the reference samples (e.g., see FIG. 18 (a)).

In another example, in steps S1930 and S1940, the intra-prediction mode information of the current block indicates the upper left direction and the intra-mirror prediction mode of the current block is the vertical direction mode, Block and the second block can be determined as a right block among the two blocks in the current block and the predicted value of the left block is generated based on the reference samples of the upper neighbor block or the upper left neighbor block or the left neighbor block adjacent to the current block , The predicted value of the right block is generated by symmetrically copying the predicted value of the left block in column units (for example, see Fig. 17 (b)), or the predicted value of the right block is calculated by weighting based on the predicted value of the right block and the reference sample of the upper neighbor block adjacent to the current block (For example, see Fig. 18 (b)).

In another example, in steps S1930 and S1940, the intra-prediction mode information of the current block indicates the lower left direction and the intra-mirror prediction mode of the current block is the horizontal direction mode, The second block may be determined as the upper block among the two blocks in the current block and the predicted value of the lower block is generated based on the reference samples of the left neighboring block or the lower left neighboring block adjacent to the current block, 16 and 17 (c)). Alternatively, the prediction value may be generated by symmetrically copying the predicted value of the lower block in units of rows (e.g., see FIGS. 16 and 17 (E. G., Fig. 18 (c)).

As another example, in steps S1930 and S1940, the intra-prediction mode information of the current block indicates the upper left direction and the intra-mirror prediction mode of the current block is the horizontal direction mode, the first block is selected from two blocks And the second block may be determined as the lower block among the two blocks in the current block and the predicted value of the upper block is generated based on the reference samples of the left neighbor block or the upper left neighbor block or the upper neighbor block adjacent to the current block The predicted value of the lower block is generated by symmetrically copying the predicted value of the upper block in a row unit basis (for example, see Fig. 17 (d)), or the predicted value of the lower block may be generated by copying the predicted value of the lower block to the left neighboring block or the lower left neighboring block (E. G., Fig. 18 (d)) based on the reference samples.

As another example, in steps S1930 and S1940, the intra-prediction mode information of the current block indicates the upper right direction and the intra-mirror prediction mode of the current block is the horizontal direction mode, the first block is selected from two blocks And the second block may be determined as the lower block among the two blocks in the current block and the predicted value of the upper block is generated based on the reference samples of the upper neighbor block or the upper right neighbor block adjacent to the current block, The prediction value may be generated by symmetrically copying the predicted value of the upper block row by row or by using a weighted sum based on the predicted value of the upper block and the reference samples of the left neighboring block or the lower left neighboring block adjacent to the current block.

In another example, in steps S1930 and S1940, the intra-prediction mode information of the current block indicates the lower left direction and the intra-mirror prediction mode of the current block is the vertical direction mode, And the second block may be determined as the right block among the two blocks in the current block and the predicted value of the left block is generated based on the reference samples of the left neighboring block or the lower left neighboring block adjacent to the current block, The predictive value may be generated by symmetrically copying the predicted value of the left block in column units or by using a weighted sum based on the predicted value of the left block and the reference sample of the upper neighbor block adjacent to the current block.

After obtaining the predicted value for the current block according to the method of FIG. 19, the encoder 100 encodes the current block (e.g., residual signal acquisition, transformation 120, quantization 130 ), Entropy encoding 190). Also, as described in relation to steps S1910 and S1920, the encoder 100 can encode information (e.g., flag information) indicating a determined direction mode in the vertical direction mode and the horizontal direction mode into a bit stream, Prediction mode information can be encoded into a bitstream. The decoder 200 can restore the current block based on the predicted value of the current block acquired according to the method of FIG. 19 (e.g., see FIG. 2 and related description).

Method 2: Vertical-Horizontal Merge Prediction Method

As described with reference to FIG. 10, even when different directional components for intraprediction coexist in one block, the coding efficiency of intra prediction may be degraded. Method 2 of the present invention proposes a Vertical-Horizontal Merge Prediction method as a method for solving this technical problem, and uses a reference sample when encoding / decoding is performed using the proposed method We propose a concrete method of generating a prediction block. The proposed vertical-horizontal merge prediction method is a method which can efficiently perform encoding when two different prediction direction components (a vertical component and a horizontal component) coexist in one block as in the example of FIG.

20 illustrates a vertical-horizontal merge method according to the present invention.

Referring to FIG. 20, one block is divided into three prediction regions (region 1, region 2, region 3). The samples in each prediction region (sample a in the region 1, sample b in the region 2, and sample c in the region 3) are generated in the following manner.

- Sample a in No. 1 area: The predicted value in the No. 1 area is always copied in the same way as in the vertical mode to generate a predicted value. That is, the predicted value of the a sample in the (1) region is generated by copying the value of the upper reference sample E. All predicted values in (1) are generated in the same way.

- Sample 2 of the area b: The prediction value of the area ② is always copied in the same way as in the horizontal mode to generate the predicted value by copying the left reference sample. That is, the predicted value of the b sample in the (2) area is generated by copying the value of the leftmost reference sample O. All predicted values in the 2 nd area are generated in the same way.

- Sample ③ c: The predicted value of ③ area is always generated as average value of vertical mode and horizontal mode. Therefore, the average value of the upper reference sample, which is a reference sample in the vertical mode, and the left reference sample, which is a reference sample in the horizontal mode, are generated as prediction values. That is, the predicted value of the c sample in the area (3) is generated by calculating the average value of the left reference sample N and the upper reference sample D. All predicted values in the ③ area are generated in the same way as the average values of the left reference sample and the upper reference sample of the current sample position.

The method illustrated in FIG. 20 can be performed by the encoder 100 and the decoder 200 (see, for example, the intra prediction unit 185 of FIG. 1, the intra prediction unit 265 of FIG. 2).

After obtaining the predicted value for the current block according to the method of FIG. 20, the encoder 100 encodes the current block (e.g., residual signal acquisition, transformation 120, quantization 130 ), Entropy encoding 190). The decoder 200 can restore the current block based on the predicted value of the current block acquired according to the method of FIG. 20 (e.g., see FIG. 2 and related description).

FIG. 21 illustrates a block diagram of an image processing apparatus to which the present invention can be applied. The image processing apparatus may include a video signal encoding apparatus and / or a decoding apparatus. For example, the image processing apparatus to which the present invention may be applied may include a mobile terminal such as a smart phone, a portable device such as a laptop computer, a home appliance such as a digital TV, a digital video player, and the like.

The memory 12 may store a program for processing and controlling the processor 11, and may store an encoded bit stream, a decoded image, control information, and the like. Also, the memory 12 can be utilized as a buffer for various video signals. The memory 12 may be a ROM (Read Only Memory), a RAM (Random Access Memory), an EPROM (Erasable Programmable Read Only Memory), an EEPROM (Electrically Erasable Programmable Read Only Memory), a flash memory, , A hard disk drive (HDD), a solid state drive (SSD), or the like.

The processor 11 controls the operation of each module in the image processing apparatus. In particular, the processor 11 may perform various control functions for performing encoding / decoding according to the present invention. The processor 11 may also be referred to as a controller, a microcontroller, a microprocessor, a microcomputer, or the like. The processor 11 may be implemented by hardware or firmware, software, or a combination thereof. When the present invention is implemented using hardware, an application specific integrated circuit (ASIC) or a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD) field programmable gate array) may be provided in the processor 11. Meanwhile, when the present invention is implemented using firmware or software, firmware or software may be configured to include a module, a procedure, or a function for performing the functions or operations of the present invention. The firmware or software may be contained within the processor 11 or may be stored in the memory 12 and driven by the processor 11.

Meanwhile, the encoding method and / or decoding method according to the present invention can be stored in a computer-readable recording medium in the form of a computer program configured to perform the methods described herein when being executed by a processor. The bitstream generated by the encoding method according to the present invention can be stored in a computer-readable recording medium in the form of multimedia data. A computer-readable recording medium includes all kinds of storage devices in which data that can be read by a computer system is stored. Examples of the computer-readable recording medium include a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, a hard disk, a solid state disk (SSD) can do.

In addition, the device 10 may optionally include a network interface module (NIM) 13. The network interface module 13 is operatively connected to the processor 11 and the processor 11 controls the network interface module 13 to transmit information and / or data, signals, messages And / or the like. The network interface module 13 supports various communication standards such as IEEE 802 series, 3GPP LTE (-A), Wi-Fi, Advanced Television System Committee (ATSC), Digital Video Broadcasting It is possible to transmit / receive a video signal such as a bit stream encoded according to the control information and / or the encoding method of the present invention according to the standard. The network interface module 13 may not be included in the apparatus if necessary.

In addition, the device 10 may optionally include an input / output interface 14. The input / output interface 14 is operatively connected to the processor 11 and the processor 11 controls the input / output interface 14 to receive and output control signals and / or data signals. The input / output module 14 may be connected to an input device such as a keyboard, a mouse, a touch pad, a camera, etc., and an output device such as a display, Interface, DVI (Digital Visual Interface), HDMI (High Definition Multimedia Interface), and the like.

The methods and embodiments described above are those in which the elements and features of the present invention are combined in a predetermined form. Each component or feature shall be considered optional unless otherwise expressly stated. Each component or feature may be implemented in a form that is not combined with other components or features. It is also possible to construct embodiments of the present invention by combining some of the elements and / or features. The order of the operations described in the embodiments of the present invention may be changed. Some configurations or features of certain embodiments may be included in other embodiments, or may be replaced with corresponding configurations or features of other embodiments. It is clear that the claims that are not expressly cited in the claims may be combined to form an embodiment or be included in a new claim by an amendment after the application.

The methods and embodiments of the present invention may be implemented by various means, for example, hardware, firmware, software, or a combination thereof. In the case of hardware implementation, an embodiment of the present invention may include one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs) field programmable gate array), a processor, a controller, a microcontroller, a microprocessor, or the like.

In the case of firmware or software implementation, the present invention may be implemented in software code or instructions, including modules, procedures, functions, and the like, to perform the functions or acts described above. The software code or instructions may be stored on a computer readable medium and may be executed by a processor and may perform operations in accordance with the present invention when it is being executed by a processor. The computer-readable medium may be connected to the processor via a network, or may be located within or external to the processor, and may exchange data with the processor.

It will be apparent to those skilled in the art that the present invention may be embodied in other specific forms without departing from the spirit of the invention. Accordingly, the above description should not be construed in a limiting sense in all respects and should be considered illustrative. The scope of the present invention should be determined by rational interpretation of the appended claims, and all changes within the scope of equivalents of the present invention are included in the scope of the present invention.

The present invention can be applied to an image processing apparatus such as a decoding apparatus and an encoding apparatus.

Claims (11)

1. A method for decoding a video signal based on Intra Mirror Prediction (IMP) mode,

   Dividing the current block into two blocks,

   Wherein the current block is divided into a vertical direction based on the IMP mode of the current block being a vertical direction mode and the current block is divided into a horizontal direction based on an IMP mode of the current block being a horizontal direction mode;

   Obtaining mode information indicating an intra prediction direction of the current block;

   Obtaining a predicted value of a first block among the two blocks based on the acquired mode information and a reference sample of a neighboring block adjacent to the current block;

   Obtaining a predicted value of a second block among the two blocks based on a predicted value of the specific block; And

   And restoring the current block based on the obtained predicted value.
2. The method according to claim 1,

   Wherein the predicted value of the second block is generated by symmetrically copying the predicted value of the first block.
3. The method according to claim 1,

   Wherein a predicted value of the second block is generated using a weighted sum based on a predicted value of the first block and a reference sample of a neighboring block adjacent to the current block.
4. The method according to claim 1,

   Wherein the first block is a right block and the second block is a left block of the two blocks based on the mode information indicating a right upper direction and the IMP mode of the current block being a vertical direction mode , The predicted value of the right block is generated based on the reference sample of the upper neighbor block or the upper right neighbor block adjacent to the current block and the predicted value of the left block is generated by symmetrically copying the predicted value of the right block in column units How.
5. The method according to claim 1,

   The first block is the left block of the two blocks and the second block is the right block of the two blocks based on the mode information indicating the upper left direction and the IMP mode of the current block being the vertical direction mode The predicted value of the left block is generated based on the reference sample of the upper neighbor block, the upper left neighbor block, or the left neighbor block adjacent to the current block, and the predicted value of the right block is generated symmetrically . ≪ / RTI >
6. The method according to claim 1,

   The first block is a lower block among the two blocks and the second block is an upper block among the two blocks based on the mode information indicating a lower left direction and the IMP mode of the current block being a horizontal direction mode , The predicted value of the lower block is generated based on a reference sample of a left neighboring block or a lower left neighboring block adjacent to the current block, and the predicted value of the upper block is generated by symmetrically copying the predicted value of the lower block, How.
7. The method according to claim 1,

   The first block is the upper block and the second block is the lower block of the two blocks based on the mode information indicating the upper left direction and the IMP mode of the current block being the horizontal mode A predicted value of the upper block is generated based on a reference sample of a left neighbor block, a left upper neighbor block, or an upper neighbor block adjacent to the current block, and the predicted value of the lower block is generated symmetrically . ≪ / RTI >
8. The method according to claim 1,

   The first block is the upper block and the second block is the lower block of the two blocks based on the mode information indicating the upper right direction and the IMP mode of the current block being the horizontal mode , The predicted value of the upper block is generated based on the reference samples of the upper neighbor block or the upper right neighbor block adjacent to the current block, and the predicted value of the lower block is generated by symmetrically copying the predicted value of the upper block, How.
9. The method according to claim 1,

   The first block is the left block of the two blocks and the second block is the right block of the two blocks based on the mode information indicating the lower left direction and the IMP mode of the current block being the vertical direction mode , The predicted value of the left block is generated based on a reference sample of a left neighboring block or a left lower neighboring block adjacent to the current block, and the predicted value of the right block is generated by symmetrically copying the predicted value of the left block, How.
10. The method according to claim 1,

    Wherein the current block is bisected by the two blocks.
11. An apparatus for decoding a video signal based on an Intra Mirror Prediction (IMP) mode,

    Memory; And

    A processor operatively coupled to the memory, the processor comprising:

    The current block is divided into two blocks,

    Wherein the current block is divided into a vertical direction based on the IMP mode of the current block being a vertical direction mode and the current block is divided into a horizontal direction based on an IMP mode of the current block being a horizontal direction mode;

    Acquiring mode information indicating an intra prediction direction of the current block;

    Obtain a predicted value of a first block among the two blocks based on the acquired mode information and a reference sample of a neighboring block adjacent to the current block;

    Obtain a predicted value of a second block among the two blocks based on the predicted value of the specific block; And

    And to restore the current block based on the obtained predicted value.

Images