KR20130002242A

KR20130002242A - Method for encoding and decoding video information

Info

Publication number: KR20130002242A
Application number: KR1020110109959A
Authority: KR
Inventors: 권재철
Original assignee: 주식회사 케이티
Priority date: 2011-06-28
Filing date: 2011-10-26
Publication date: 2013-01-07

Abstract

The present invention is an image information encoding / decoding method using inter prediction. According to an embodiment of the present invention, an encoding method according to the present invention is an image information encoding method. And performing quantization and entropy encoding, wherein the performing inter prediction includes selecting a candidate available for each spatio-temporal adjacent region of the current block, selecting a final prediction candidate from among the adjacent region candidates, and performing the final prediction. And performing prediction on the current block by using motion information of a prediction candidate. According to the present invention, the merge mode and the AMVP mode may be integrated to effectively perform inter prediction on the current block.

Description

Method for encoding and decoding video information {Method For Encoding And Decoding Video Information}

The present invention relates to an image compression technique, and more particularly, to an encoding and decoding technique of image information using inter prediction.

Recently, the demand for high resolution and high quality images such as high definition (HD) images and ultra high definition (UHD) images is increasing in various fields.

In order to provide a high resolution and high quality image, the data amount of the image data is increased. Therefore, the transmission cost and the storage cost of the image data for providing a high resolution and high quality image are increased as compared with the conventional image data processing method. High-efficiency image compression techniques can be utilized to solve such problems as image data becomes high-resolution and high-quality.

A technique for compressing image data. An inter prediction method for predicting a pixel value included in a current picture from another picture, and an information of another pixel of the current picture using the pixel value of the current picture. Various techniques are used, such as intra prediction (Intra Prediction) method, which is predicted by using a signal, and an entropy encoding method that performs encoding by assigning a shorter code to a signal having a higher frequency of occurrence or occurrence. .

Various techniques are used to reduce the amount of transmission data in the process of compressing and processing image data. For example, in the case of inter prediction, a merge mode that predicts pixel values of a current prediction block using information of neighboring blocks. And AMVP (Advanced Motion Vector Prediction) are used.

An object of the present invention is to provide an encoding / decoding method and apparatus for improving compression transmission efficiency.

It is a technical object of the present invention to provide a method and apparatus that can increase encoding / decoding efficiency through effective prediction.

It is a technical object of the present invention to provide a method and apparatus capable of increasing encoding / decoding efficiency by integrating intra prediction modes.

An object of the present invention is to provide a method and apparatus for performing inter prediction on a current block by integrating a merge mode and an AMVP mode.

(1) An embodiment of the present invention is an image information encoding method, comprising: performing inter prediction of a current block and quantizing and entropy encoding an inter predicted value, and performing the inter prediction, Selecting a candidate available for each spatio-temporal adjacent region of the current block, selecting a final prediction candidate from among the adjacent region candidates, and performing prediction on the current block using motion information of the final prediction candidate. Include.

(2) In (1), the candidate for each region may be a block having the largest size among the blocks in the region.

(3) In (2), in a neighboring area in which a plurality of blocks having the largest size among the adjacent areas exist, the neighboring area may be scanned in a predetermined direction to select a block having the largest size detected first.

(4) In (1), the candidate for each region may be a block having a median vector value among the blocks in the region.

(5) In (4), in a neighboring region in which a plurality of blocks having a median vector value exist in the neighboring region, candidates for the first detected block among the blocks having a median vector value are scanned by scanning the adjacent region in a predetermined direction. Can be selected.

(6) In (1), the adjacent region is a left region, an upper region, a corner region, and a temporal region, and in the prediction candidate selection step, one prediction block may be selected for each region.

(7) In (1), it may be a candidate having the same motion information as the current block.

(8) In (7), when there is no candidate having the same motion information as the current block, a candidate for minimizing the bit cost among the candidates for each region may be selected as the final prediction candidate.

(9) In (1), the reference picture index of the same position block that is a temporal adjacent region among the adjacent regions may use the reference picture index of the current block.

(10) In (1), the reference picture index of the same location block, which is a temporal adjacent area among the adjacent areas, may use a reference picture index having the smallest value among the reference picture indexes for the candidate for each adjacent area.

(11) Another embodiment of the present invention is a method of decoding image information, comprising: entropy decoding a bit stream, performing inter prediction on a current block based on an entropy decoded signal, and reproducing based on the inter prediction And generating a signal, wherein performing the inter prediction, extracting a prediction candidate for each spatio-temporal adjacent region of the current block, constructing a list of the extracted prediction candidates, and based on the list. Predicting motion information of the current block, In the step of predicting the motion information of the current block,

If there is a first candidate having the same motion information as the motion information of the current block among the prediction candidates of the list, the motion information of the current block is predicted based on the motion information of the first candidate, and the first candidate is If not, the motion information of the current block may be predicted using the second candidate used for prediction of the current block among the candidates on the list and the residual signal of the current block.

(12) In (11), the prediction candidate for each neighboring region may be a block having the largest size among blocks in each neighboring region of the current block.

(13) In (11), the prediction candidate for each neighboring region may be a block having a median vector value among blocks in each neighboring region of the current block.

(14) In (11), the reference picture index of the same position block that is a temporal adjacent region among the adjacent regions may use the reference picture index of the current block.

(15) In (11), the reference picture index of the same position block that is the temporal neighboring region among the adjacent regions may use the reference picture index having the smallest value among the reference picture indexes for the candidate for each neighboring region.

(16) In (11), the entropy decoded information includes information indicating that the first candidate exists and information indicating the first candidate from the list, and in the step of predicting motion information of the current block, The indicated motion information of the first candidate may be used as the motion information of the current block.

(17) In (11), the entropy decoded information includes information indicating that the first candidate does not exist and information indicating a second candidate used in the prediction of the current block in the list, wherein In the predicting the motion information, the motion information of the current block may be generated based on the indicated second candidate and the residual signal of the current block.

(18) In (17), the motion vector of the current block may be the sum of the motion vector of the second candidate and the residual signal.

According to the present invention, the encoding / decoding efficiency can be increased through effective prediction.

According to the present invention, the encoding / decoding efficiency can be increased by integrating intra prediction modes.

According to the present invention, the merge mode and the AMVP mode may be integrated to effectively perform inter prediction on the current block.

FIG. 1 is a block diagram illustrating a configuration of a video encoder according to an embodiment of the present invention. Referring to FIG.
2 is a block diagram illustrating a configuration of a video decoder according to an embodiment.
3 is a diagram schematically illustrating a method of determining a merge candidate of a current PU.
4 is a diagram schematically illustrating a method of selecting an AMVP candidate of a current PU.
5 is a diagram schematically illustrating an example of a method for selecting a candidate group (left, top, corner, temporal).
FIG. 6 is a diagram schematically illustrating an example of a method for selecting a candidate group (left, top, temporal).
7 is a diagram schematically illustrating an example of a method for selecting a candidate group {left, top, corner1, corner2, and time}.
FIG. 8 schematically illustrates an example of PU partitioning of a temporal co-located region in the example of FIGS. 5 to 7.
9 is a flowchart schematically illustrating the operation of an encoder in a system to which the present invention is applied.
10 is a flowchart schematically illustrating an operation of a decoder in a system to which the present invention is applied.

Each of the components in the drawings described herein are shown independently for the convenience of description regarding different characteristic functions in the image encoder / decoder, and it is understood that each of the components is implemented in separate hardware or separate software. It does not mean. For example, two or more of each configuration may be combined to form one configuration, or one configuration may be divided into a plurality of configurations. Embodiments in which each configuration is integrated and / or separated are also included in the scope of the present invention unless they depart from the essence of the present invention.

Hereinafter, with reference to the accompanying drawings, it will be described in detail a preferred embodiment of the present invention. The same reference numerals are used for the same components in the drawings, and duplicate descriptions of the same components are omitted.

FIG. 1 is a block diagram illustrating a configuration of a video encoder according to an embodiment of the present invention. Referring to FIG. Referring to FIG. 1, the video encoder includes a picture splitter 110, an inter predictor 120, an intra predictor 125, a transformer 130, a quantizer 135, an inverse quantizer 140, An inverse transform unit 145, a filter unit 150, a memory 155, a reordering unit 160, and an entropy encoding unit 165 are included.

The picture division unit 110 may divide the inputted current picture into one or more coding units. A CU (CU) is a unit which is encoded by a video encoder. The CU is hierarchically divided by a depth information based on a quad tree structure, .

CUs can have various sizes such as 8 × 8, 16 × 16, 32 × 32, and 64 × 64. The largest sized CU is called a Large Coding Unit (LCU), and the smallest sized CU is called a Smallest Coding Unit (SCU). The picture division unit 110 may divide the CU and generate a prediction unit (PU) and a transform unit (TU).

In the inter prediction mode, the inter prediction unit 120 may perform motion estimation (ME) and motion compensation (MC). The inter prediction unit 120 generates a prediction block based on at least one picture information of a previous picture or a subsequent picture of the current picture.

The inter prediction unit 120 performs motion estimation based on the divided prediction target block and at least one reference block stored in the memory unit 155. The inter prediction unit 120 generates motion information including a motion vector (MV), a reference block index, a prediction mode, and the like as a result of the motion estimation.

The inter-prediction unit 120 performs motion compensation using the motion information and the reference block. In this case, the inter-prediction unit 120 generates and outputs a prediction block corresponding to the input block from the reference block.

In the intra prediction mode, the intra prediction unit 125 may generate a prediction block based on pixel information in the current picture. In the intra prediction mode, the intra predictor 125 may perform prediction on the current block based on the prediction target block and the reconstructed block that is previously transformed and quantized and then reconstructed. The reconstruction block may be a reconstructed image before passing through the deblocking filter unit.

The residual block is generated by the difference between the prediction target block and the prediction block generated in the inter or intra prediction mode.

The transform unit 130 performs transform on the residual block for each TU to generate transform coefficients.

The TU may have a tree structure within the range of the maximum size and the minimum size. A flag may indicate whether a current block is divided into sub-blocks for each TU. The transform unit 130 may perform transformation using a discrete cosine transform (DCT) and / or a discrete sine transform (DST).

The quantizer 135 may quantize the values converted by the transformer 130. Depending on the block or the importance of the image, the quantization factor may change. The quantized transform coefficient values may be provided to the reordering unit 160 and the inverse quantization unit 140.

The reordering unit 160 may align the transform coefficients of the quantized 2D block type into the transform coefficients of the 1D vector form through a scan in order to increase the efficiency of entropy encoding. At this time, the reordering unit 160 may increase the entropy encoding efficiency by changing the scan order based on the probabilistic statistics.

The entropy encoder 165 entropy encodes the values obtained by the reordering unit 160. The coded information forms a compressed bit stream and is transmitted or stored through a network abstraction layer (NAL).

The inverse quantizer 140 inverse quantizes the transform coefficients quantized by the quantizer 135, and the inverse transformer 145 inverse transforms the inverse quantized transform coefficients to generate a reconstructed residual block. The reconstructed residual block may be combined with the predicted block generated by the inter predictor 120 or the intra predictor 125 to generate a reconstructed block. The reconstruction block is provided to the intra predictor 125 and the filter 150.

The filter unit 150 may apply a deblocking filter, an adaptive loop filter (ALF), a sample adaptive offset (SAO), and the like to the restored residual block. The deblocking filter filters the reconstructed blocks to remove distortions between block boundaries occurring in the encoding and decoding processes.

The ALF performs filtering to minimize the error between the predicted block and the last reconstructed block. The ALF performs filtering based on a value obtained by comparing the reconstructed block filtered through the deblocking filter with the current block to be predicted, and the filter coefficient information of the ALF is carried in a slice header and transmitted from the encoder to the decoder. have.

SAO is a loop filter process for restoring an offset difference from an original image on a pixel-by-pixel basis for a residual block to which a deblocking filter is applied. Offsets applied through SAO include a band offset and an edge offset. The band offset divides the pixel into 32 bands according to intensity, and applies the offset by dividing the 32 bands into two band groups of 16 bands at the edge and 16 bands at the center. The edge offset applies an offset by classifying the direction of the edge and the intensity of the surrounding pixels for each pixel.

The memory 155 may store a final reconstructed block that has passed through the filter unit 150, and the stored final reconstructed block may be provided to the inter predictor 120 that performs inter prediction.

2 is a block diagram illustrating a configuration of a video decoder according to an embodiment. Referring to FIG. 2, the video decoder includes an entropy decoder 210, a reordering unit 215, an inverse quantizer 220, an inverse transform unit 225, an inter predictor 230, an intra predictor 235, and a filter. The unit 240 and the memory 245 are included.

The entropy decoder 210 receives the compressed bit stream from the NAL. The entropy decoding unit 210 entropy decodes the received bit stream, and entropy decodes the prediction mode, motion vector information, and the like when the bit stream is included in the bit stream.

The entropy decoded transform coefficient or residual signal is provided to the reordering unit 215. The reordering unit 215 inverse scans the decoded transform coefficients or the residual signal to generate transform coefficients in the form of a two-dimensional block.

The inverse quantization unit 220 inverse quantizes the rearranged transform coefficients. The inverse transform unit 225 inversely transforms the inverse quantized transform coefficients to generate a residual block.

The residual block may be combined with the prediction block generated by the inter predictor 230 or the intra predictor 235 to generate a reconstructed block. The reconstruction block is provided to the intra predictor 235 and the filter 240. The operations of the inter predictor 230 and the intra predictor 235 may be the same as the operations of the inter predictor 120 and the intra predictor 125 in the video encoder, respectively.

The filter unit 240 may apply a deblocking filter, ALF, SAO, etc. to the reconstruction block. The deblocking filter filters the reconstructed blocks to remove distortions between block boundaries occurring in the encoding and decoding processes. The ALF performs filtering on the deblocking filtered reconstructed block to minimize the error between the predicted block and the last reconstructed block. In addition, SAO may be applied to the deblocking filtered reconstructed block on a pixel basis to reduce a difference from the original image.

The memory 245 may store a final reconstructed block obtained through the filter unit 240, and the stored final reconstructed block may be provided to the inter predictor 230 that performs inter prediction.

On the other hand, as the inter prediction method, there are a merge mode and an AMVP mode as described above.

The merge mode obtains merge candidates for the current PU from the motion information of blocks adjacent to the current PU in time and space.

3 is a diagram schematically illustrating a method of determining a merge candidate of a current PU. Referring to FIG. 3, the merge candidates for the current PU (P) are co-located blocks of the neighboring blocks A, B, C, and D of the current PU (P) and the current PU (P). COL). The five candidates A, B, C, D, and COL are searched for available blocks in the order A → B → COL → C → D and adopted as merge candidates of the current PU (P).

For example, as a result of the encoder searching in the above order, if a candidate having the same motion information as the current PU (P) among the merge candidates (A, B, COL, C, D) exists, the prediction mode for the current PU (P) Merge mode is available. The encoder may transmit information indicating that the merge mode is used, for example, to which merge candidate the current PU (P) merges among the merge candidates with the merge flag (Merge_flag), for example, a merge index (Merge_index) to the decoder. have.

In the AMVP mode, the motion vector (MV) is predicted using a block adjacent to the current PU in space and time, and then the current block is selected by selecting the adjacent block having the smallest prediction error, that is, the neighboring block having the motion vector closest to the motion vector of the current vector. And a method of encoding a motion vector difference of the selected neighboring block.

4 is a diagram schematically illustrating a method of selecting an AMVP candidate of a current PU. Referring to FIG. 4, in the AMVP mode, candidate lists having three directions A, B, and T _BR are selected, and candidates having optimal prediction values among these candidates are selected from the viewpoint of Rate-Distortion Optimization (RDO).

In the AMVP, as shown in FIG. 4, the left block, the right block, and the co-locater blocks of the current PU (P) become motion vector (MV) candidates. Among the neighboring blocks, the first block determined to be available as a result of searching in the scan order as shown in FIG. 1 may be selected as the representative block.

The merge mode and the AMVP mode described above were used for different purposes in the prior art, and the positions of prediction candidates (candidates to be encoded) were also different. However, similarities exist between merge mode and AMVP mode. For example, both the merge mode and the AMVP mode predict information of the current block by using neighboring block information of the current PU. Also, by making MODE_SKIP as much as possible or by setting the Motion Vector Difference (MVD) to 0, the motion vector difference (MVD) is also changed in the AMVP mode, as in the merge mode proposed to increase coding efficiency. Skip mode may occur where there is zero or no residual signal to be transmitted after motion compensation.

Therefore, considering the overall efficiency of compression techniques such as coding efficiency and complexity, a method of integrating and applying merge mode and AMVP mode can be considered.

In the motion prediction incorporating the merge mode and the AMVP mode (hereinafter referred to as 'integrated motion prediction' for convenience of description), the method includes: among available neighboring blocks (including co-located blocks) of the current PU. Determine the candidate. As a method for determining a candidate, a candidate may be determined by giving priority to a block having the largest size, or a candidate may be determined by giving priority to a median block.

In this case, available blocks include (1) a motion vector (MV), (2) a reference picture index that is the same as the current PU, and (3) a reference direction index that is the same as the current PU. Branches are blocks. For example, the reference direction index may be an index of the reference list, and blocks having the same reference direction index may use the same reference list L0 or L1.

How to choose candidates for integrated motion prediction

(1) {Left, Top, Corner, Temporal} candidate group selection method

According to the method, the spatio-temporal neighboring block of the current PU is divided into four neighboring blocks: a left neighboring block, a top neighboring block, a corner neighboring block, and a temporal neighboring block. For each region, one block among adjacent blocks of the region may be adopted as a candidate block of the region for the current PU.

5 is a diagram schematically illustrating an example of a method for selecting a candidate group (left, top, corner, temporal). Candidate groups can be set as shown in FIGS. 5A and 5B according to whether the upper left corner block of the current PU is the adjacent block of the corner region or the adjacent block of the upper region.

As shown in Fig. 5A, when the upper left corner block is the adjacent block of the corner region, the candidate block of the left region is selected from the adjacent blocks L0, L1, and L2 of the left region, and adjacent to the upper region. The candidate block of the upper region is selected from the blocks T0, T1, and T2, and the candidate block of the corner region is selected from the adjacent blocks C0, C1, and C2 of the corner region, and the candidate for the time region COL. You can select a block.

As shown in Fig. 5B, when the upper left corner block is the adjacent block of the upper region, the candidate block of the left region is selected from the adjacent blocks L0, L1, and L2 of the left region, and adjacent to the upper region. The candidate block of the upper region is selected from the blocks T0, T1, T2, and T3, the candidate block of the corner region is selected from the adjacent blocks C0 and C1 of the corner region, and the candidate for the time region COL. You can select a block.

(2) {left, top, temporal} candidate group selection method

According to the method, the spatio-temporal neighboring blocks of the current PU are divided into three regions (left region, upper region, and time region), and one block among neighboring blocks of the region for each region is a candidate of the corresponding region for the current PU. Can be adopted as a block.

FIG. 6 is a diagram schematically illustrating an example of a method for selecting a candidate group (left, top, temporal). Referring to FIG. 6, candidate blocks of the left region are selected from adjacent blocks L0, L1, L2, and L3 of the left region with respect to the current PU (P), and adjacent blocks T0, T1, and T2 of the upper region are selected. , Candidate blocks in the upper region may be selected from among T3 and T4, and candidate blocks may be selected in the time domain COL.

(3) {left, top, corner1 (corner1), corner2 (corner2), time (temport)} candidate group selection method

According to the method, the spatio-temporal neighboring blocks of the current PU are divided into four regions, and the top corner and the left corner are independently divided, and one block among the neighboring blocks of the corresponding region for each region. May be adopted as a candidate block of the corresponding region for the current PU.

7 is a diagram schematically illustrating an example of a method for selecting a candidate group {left, top, corner1, corner2, and time}. Referring to FIG. 7, the candidate block of the left region is selected from the adjacent blocks L0, L1, and L2 of the left region with respect to the current PU (P), and the adjacent blocks T0, T1, T2, and T3 of the upper region are selected. ) Select a candidate block of the upper region, select block (C) of the corner 1 as a candidate block of the corner 1 region, select block (D) of the corner 2 as a candidate block of the corner 2 region, Candidate block can be selected for COL).

5 to 7 illustrate a predetermined number of blocks and a block size for each region as an example, but this is only an example for describing the present invention, and the number and size of adjacent blocks in each region are not limited to a specific example and vary. can do.

Hereinafter, a method of determining a prediction block for each region will be described.

Region Prediction Block Determination Method

Spatio-temporal neighboring blocks that may be included in each region candidate group should be available blocks. Here, the available blocks include (1) a motion vector (MV), (2) the same reference picture index as the current PU, and (3) the same reference direction index as the current PU. A block that has

Accordingly, in the method for selecting the candidate for the integrated motion prediction, neighboring blocks for each region may be configured among available blocks, and in the process of selecting candidate blocks after constructing neighboring blocks for each region, the (1) to The availability according to (3) may be first determined.

Candidate blocks for each region may be determined according to priorities among available adjacent blocks. As the priority for selecting candidate blocks, (1) the size of the block, (2) whether or not the median vector has a median value, or the like can be used.

(1) Priority of Candidate Block Determination-Block Size

In this case, the candidate block may be determined by giving a high priority to the block having the largest size among the available neighboring blocks. When considering the size of a block preferentially, candidate blocks can be selected as shown in the following (a) to (b).

(a) Of the neighboring blocks for each region existing in the spatiotemporal neighboring region (eg, left region, upper region, corner region, and time region) of the current PU, the highest priority is given to the largest block, It selects as a candidate block.

(b) When there are a plurality of blocks having the largest size in each region, candidate blocks of the region may be determined according to the following (i) to (iii). If a block corresponding to a higher priority does not exist, a block corresponding to a next priority may be selected as a candidate block of the corresponding region.

(Iii) In the case of the upper region, the block having the largest size to be scanned first is determined as a candidate block while scanning from right to left.

(Ii) In the case of the left region, the block having the largest size to be scanned first is determined as a candidate block while scanning from the bottom to the top.

(I) In the case of a corner area, the largest sized block scanned first is determined as a candidate block while scanning in the order of right-top, left-below, and left-top.

(Iii) Even in the temporal co-located region, the block having the largest size scanned first is determined as a candidate block while scanning in the order of right to left and bottom to top.

In the case where the method of selecting the candidate blocks for each region by giving priority to the size of the blocks is applied to the example of FIGS. 5 to 7, the candidate blocks for each region to be selected are as follows. FIG. 8 schematically illustrates an example of PU partitioning of a temporal co-located region in the example of FIGS. 5 to 7.

5 and 8, the candidate block in the left region is L0, the candidate block in the upper region is T1, the candidate block in the corner region is C2, and the candidate block in the time region is Co6.

6 and 8, the candidate block in the left region is L1, the candidate block in the upper region is T2, and the candidate block in the time region is Co6.

7 and 8, the candidate block in the left region is L0, the candidate block in the upper region is T1, the candidate block in the corner 1 region is C, the candidate block in the corner 2 region is D, and the candidate block in the time region is It becomes Co6.

(2) Priority of candidate block determination-presence of median vector value

Instead of giving the highest priority to the block of the largest size, a method of giving the highest priority to the block having the median vector value may be used.

In the case where the median vector value is considered first, candidate blocks for each region can be selected as shown in (a) to (b) below.

(a) A block having a median vector value is selected as a candidate block in each region.

(b) When there are a plurality of blocks having a median vector value in the region, candidate blocks of the region may be determined according to the following (i) to (iii). If a block corresponding to a higher priority does not exist, a block corresponding to a next priority may be selected as a candidate block of the corresponding region.

(Iii) In the upper region, a block to be scanned first is selected as a candidate block among blocks having a median vector value while scanning from right to left.

(Ii) In the left region, a block to be scanned first is selected as a candidate block among blocks having a median vector value while scanning from the bottom to the top.

(Iii) In the case of a corner area, among the blocks having the median vector value while scanning in the order of right-top, left-below, and left-top, the first scanned block is selected as a candidate block. Decide

(Iii) In the case of the temporal co-located region, the first scanned block is determined as a candidate block among blocks having a median vector value while scanning from right to left and bottom to top.

Whether the block having the largest size is selected as a candidate block of a corresponding region or a block having a median vector value as a candidate block of the corresponding region may be selected in units of sequence levels, in units of picture levels, or CU. It may be selected in units.

How to determine the reference picture index of the same position block for the current PU

Meanwhile, in the above-described methods, the motion information of the spatial neighboring block may be copied by the motion information of the spatial previous block at the corresponding position. In contrast, the motion information of a temporal neighboring block must determine which picture of the co-located block the motion information of the picture is to be obtained. That is, the reference picture index refIdxLX of the same block must be determined.

As a method of determining the reference picture index of the same position block, the following two methods may be used.

(1) How to copy and use the reference picture index value of the current PU

The reference picture index of the current PU, which is a motion prediction target, may be used as a reference picture index of a candidate block of a temporal region as it is.

(2) Method for Computing Using Reference Picture Index of Spatial Adjacent Block

A reference picture index of the same position block may be calculated from the reference picture indexes of spatial neighboring blocks (blocks of left, top, and corners) of the current PU.

For example, if all candidate blocks are available, the reference picture index refIdxLX of the same location block may be determined as follows.

In the example of FIG. 5, refIdxLX = min (refIdxLX (left), refIdxLX (upper), refIdxLX (corner)).

In the example of FIG. 6, refIdxLX = min (refIdxLX (left), refIdxLX (upper)).

In the example of FIG. 7, refIdxLX = min (refIdxLX (left), refIdxLX (top), refIdxLX (corner 1), refIdxLX (corner 2)).

Here, refIdxLX (R) represents a reference picture index of a candidate block of the region R.

9 is a flowchart schematically illustrating the operation of an encoder in a system to which the present invention is applied.

Referring to FIG. 9, the encoder performs motion prediction on the current PU (S910).

When a new CU of the current frame is input, the encoder may perform inter prediction on the current PU to calculate a motion vector predictor (MVP) of the current PU which is an inter mode.

Specifically, first, a new CU of the current frame is input. When the input CU is a CU of inter prediction mode (hereinafter, referred to as 'inter CU'), one inter CU may be configured of several inter PUs, and two prediction modes (PredMode) and skipping It may have one of a mode (MODE_SKIP, hereinafter referred to as 'MODE_SKIP') and an inter mode (MODE_INTER, hereinafter referred to as 'MODE_INTER'). A CU having MODE_SKIP is not divided into smaller PUs, and motion information of a PU having a partition mode (PartMode) of PART_2Nx2N is allocated.

A CU having MODE_INTER can exist in four types of PU partitions, and information indicating whether the prediction mode is MODE_INTER (PredMode == MODE_INTER) and the partition type is PART_2Nx2N, PART_2NxN, PART_Nx2N, or PART_NxN in the CU-level syntax. PartMode == PART_2Nx2N, PartMode == PART_2NxN, PartMode == PART_Nx2N, or PartMode == PART_NxN) may be delivered to the decoder.

Subsequently, motion prediction is performed for the current PU in inter mode. When a CU is partitioned into multiple PUs, a current PU (PU to be currently encoded) is input. The motion prediction for the current PU may be performed using the previous frame, the next frame, or the previous and subsequent frames of the current frame. Through motion prediction, motion information (motion vector, reference picture index, prediction direction index) for the current PU can be obtained.

Subsequently, the motion prediction value MVP of the current PU in the inter mode is calculated. The motion information of the current PU is not sent to the decoder as it is, and the difference between the prediction value obtained from the space-time adjacent blocks and the current PU motion information is transmitted to the decoder in order to improve compression efficiency. In order to calculate the motion prediction value, an integrated prediction candidate list is created. The method of selecting a candidate block in the integrated motion prediction method is as described above.

If there are a plurality of candidate blocks having the same motion vector and the reference picture index among candidate blocks on the integrated prediction candidate list including the candidate blocks for each region selected by using the integrated prediction method, remaining candidates except the candidate block having the highest priority You can delete blocks from the list.

If there is a candidate block having the same motion information as the current PU among candidate blocks on the integrated prediction candidate list, the candidate block is selected as the final prediction candidate block for the current PU. If a candidate block having the same motion information as the current PU is not in the integrated prediction candidate list, a motion vector of a best predictor is selected as a final prediction candidate among candidate blocks in the integrated prediction candidate list. That is, the candidate block having the best predictor is selected as the final prediction candidate block.

The best predictor is a motion block of a candidate block (RD) that minimizes a Rate Distortion (RD) cost function (e.g., J _{Mot SAD} considering the bit cost and sum of absolute difference).

The encoder calculates motion information of the current PU. If a candidate block having the same motion information as the current PU exists in the unified prediction candidate list, information indicating this may be configured and encoded together with the index of the final prediction candidate block. For example, a value of a flag (eg, Motion_Flag) indicating that there is a candidate block having the same motion information as the current PU may be set to 1, and the index (Mption_Idx) of the final prediction candidate block may be encoded.

If there is no candidate block having the same motion information as the current PU, the residual signal may be obtained after motion compensation using the motion information difference value MVD and the best predictor. The difference (MVD) between the motion vector of the current PU and the motion vector of the best predictor is then entropy encoded.

Through the motion compensation, a residual signal may be obtained by obtaining a pixel unit difference between the pixel value of the current block and the pixel value of the final prediction candidate block (S920).

The residual signal is encoded through a transform. 2X2, 4X4, 8X8, 16X16, 32X32, 64X64, etc. may be used as the transform encoding kernel, and the kernel used for the conversion may be limited in advance. In this case, the transform coefficient C for the nxn block may be calculated as in Equation 1.

Here, T (n, n) is a transformation matrix for nxn blocks, and B (n, n) represents nxn residual blocks. The transform coefficients are then quantized.

Subsequently, it is determined based on the ROD which one of the residual signal and the changed coefficient of the residual signal is transmitted (S930). If the prediction is performed properly, the residual signal can still be transmitted without performing transform encoding. At this time, the cost function can be selected by comparing the cost functions before and after the transform encoding.

Then, signal the type of signal to be transmitted for the current block. For example, when the cost when transmitting the residual signal is lower, the residual signal is signaled for the current block, and when the cost when transmitting the transform coefficient is lower, the transform coefficient is signaled for the current block.

Subsequently, entropy encoding is performed on the transmission target information (S940). For example, when the transform coefficients are transmitted, the transform coefficients are scanned, and then information about the scanned transform coefficients and the inter prediction mode is entropy encoded.

10 is a flowchart schematically illustrating an operation of a decoder in a system to which the present invention is applied.

The decoder first entropy decodes the received bitstream (S1010). The decoder can identify the block type from a variable length coding (VLC) table and know the prediction mode of the current block. In addition, the decoder may check whether the information transmitted for the current block is a residual signal or a transform coefficient. According to the confirmed result, the residual signal or the transform coefficient for the current block can be obtained.

The decoder inversely scans the entropy decoded residual signal or transform coefficients (S1020). The decoder may perform inverse scan to generate a two-dimensional block. If the received signal is a residual signal, the residual block may be generated. If the received block is a transform coefficient, a transform block can be generated. The decoder may dequantize and inverse transform the transform block to obtain a residual block. The process of obtaining the residual block through the inverse transform of the transform block is shown in Equation 2.

Subsequently, the decoder performs inter prediction on the current block (S1030). First, the decoder creates a unified prediction candidate list. Specifically, a prediction candidate is extracted from neighboring PU partitions in the spatial direction. In addition, it is possible to extract a temporal prediction candidate for the current PU. If a reference picture index (refIdxLX) is needed for this purpose, a reference picture index can be obtained as described above. The available temporal motion vector prediction value (MVP) can be obtained using the reference picture index.

An integrated prediction candidate list (MotionCandList) may be generated as spatial and temporal prediction candidates. If there are candidates having the same reference picture index among the prediction candidates, the remaining candidates except the candidate having the highest priority among the candidates having the same reference picture index may be deleted from the list.

Then, a motion vector is obtained and motion compensation is performed. For example, if the above-described value of Motion_Flag is 1, the motion vector mvLX and the reference picture index refIdxLX of the candidate indicated by the index Motion_Idx of the last prediction candidate can be extracted and used for motion compensation. If the value of Motion_Flag is not 1, the motion vector indicated by the received index Motion_Idx is assigned to mvpLX. Subsequently, the motion vector mvLX may be calculated using Equation 3 and the mvpLX value.

Here, mvLX [0], mvdLX [0] and mvpLX [0] are values in the x direction, and mvLX [1], mvdLX [1] and mvpLX [1] are values in the y direction.

Subsequently, the decoder generates a reproduction signal (S1040). For example, the decoder may add a residual signal and a signal of a previous frame to generate a reproduction signal. The play signal may be generated by adding the motion compensated prediction signal in the previous frame and the residual signal of the decoded current PU using the calculated motion vector.

Although the configuration of the present invention has been described in detail with reference to the preferred embodiments and the accompanying drawings, this is merely an example of the present invention and various modifications are possible within the scope without departing from the spirit of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined not only by the scope of the following claims, but also by the equivalents of the claims.

Claims

Performing inter prediction of the current block; And
Quantizing and entropy encoding the inter predicted value,
The performing of the inter prediction may include:
Selecting an available candidate for each spatio-temporal adjacent region of the current block;
Selecting a final prediction candidate from among the adjacent region candidates; And
And predicting the current block by using motion information of the final prediction candidate.

The method of claim 1, wherein the candidate for each region is a block having the largest size among the blocks in the region.

The method according to claim 2, wherein in the adjacent area where a plurality of blocks having the largest size among the adjacent areas exist, the adjacent area is scanned in a predetermined direction to select a block having the largest size detected first as a candidate. Image information encoding method.

The image information encoding method of claim 1, wherein the candidate for each region is a block having a median vector value among the blocks in the region.

The method of claim 4, wherein in a neighboring region having a plurality of blocks having a median vector value among the adjacent regions, the neighboring region is scanned in a predetermined direction to select a first detected block among the blocks having a median vector value as a candidate. Video information encoding method.

The image information encoding method of claim 1, wherein the adjacent region is a left region, an upper region, a corner region, and a temporal region, and in the prediction candidate selection step, one prediction block is selected for each region.

The video information encoding method of claim 1, wherein the video information encoding method is a candidate having the same motion information as the current block.

The method of claim 7, wherein when there is no candidate having the same motion information as the current block, a candidate for minimizing a bit cost among the candidates for each region is selected as a final prediction candidate.

The image information encoding method of claim 1, wherein the reference picture index of the same location block that is a temporal adjacent area among the adjacent areas uses a reference picture index of the current block.

The image information encoding according to claim 1, wherein the reference picture index of the same position block which is a temporal adjacent region among the adjacent regions uses a reference picture index having the smallest value among the reference picture indexes for the candidate for each adjacent region. Way.

Entropy decoding the bit stream;
Performing inter prediction on the current block based on the entropy decoded signal; And
Generating a reproduction signal based on the inter prediction;
In the performing of the inter prediction,
Extracting a prediction candidate for each spatio-temporal adjacent region of the current block;
Constructing a list of the extracted prediction candidates; And
Predicting motion information of the current block based on the list;
In the predicting motion information of the current block,
If there is a first candidate having the same motion information as the motion information of the current block among the prediction candidates of the list, the motion information of the current block is predicted based on the motion information of the first candidate, and the first candidate is And if there is none, predicting motion information of the current block by using a second candidate used for prediction of the current block among the candidates on the list and the residual signal of the current block.

12. The method of claim 11, wherein the prediction candidate for each neighboring region is a block having the largest size among blocks in each neighboring region of the current block.

12. The method of claim 11, wherein the prediction candidate for each neighboring region is a block having a median vector value among blocks in each neighboring region of the current block.

12. The method of claim 11, wherein the reference picture index of the same location block that is a temporal adjacent area among the adjacent areas uses the reference picture index of the current block.

12. The image information decoding of claim 11, wherein the reference picture index of the same location block that is a temporal adjacent area among the adjacent areas uses a reference picture index having the smallest value among the reference picture indexes for the candidate for each adjacent area. Way.

The method of claim 11, wherein the entropy decoded information includes information indicating that the first candidate exists and information indicating the first candidate from the list.
And predicting motion information of the current block comprises using the indicated motion information of the first candidate as motion information of the current block.

The method of claim 11, wherein the entropy decoded information includes information indicating that the first candidate does not exist and information indicating a second candidate used in the prediction of the current block in the list.
And predicting motion information of the current block generates motion information of the current block based on the indicated second candidate and the residual signal of the current block.

The method of claim 17, wherein the motion vector of the current block is a sum of the motion vector of the second candidate and the residual signal.