KR102114421B1 - Method and apparatus for encoding/decoding image by using motion vector of previous block as motion vector of current block - Google Patents

Abstract

The motion vector of the current block is determined based on the motion vector of at least one block encoded / decoded before the current block, and the current block is based on one of first, second, and bi-directional predictions according to the determined motion vector Disclosed is a method and apparatus for predictive encoding / decoding.

Description

Method and apparatus for encoding / decoding image by using motion vector of previous block as motion vector of current block}

The present invention relates to a method and apparatus for encoding and decoding an image, and more particularly, to a method and apparatus for encoding and decoding an image based on inter prediction.

In video compression schemes such as MPEG-1, MPEG-2, and MPEG-4 H.264 / MPEG-4 Advanced Video Coding (AVC), the video is divided into blocks of a predetermined size in order to encode it. Then, each block is predictively coded using inter prediction or intra prediction.

For inter prediction, a block identical or similar to the current block is searched in at least one reference frame, motion is estimated, and a motion vector generated as a result of motion estimation is encoded together with a pixel value and inserted into a bitstream.

Technical problem to be solved by the present invention is to provide a method and apparatus for determining a motion vector of a current block based on a motion vector of at least one previous block, and encoding and decoding an image based on the determined motion vector. Disclosed is a computer readable recording medium recording a program for executing a method.

A motion vector of a current block is determined based on a motion vector of at least one block decoded before the current block according to an embodiment of the present invention for solving the above technical problem, and the current block is decoded using the determined motion vector The method comprises: decoding information about a prediction direction used for decoding of the current block and information about pixel values of the current block among the first direction, the second direction, and both directions; Determining a prediction direction of the current block based on information on the decoded prediction direction, and determining at least one motion vector for performing prediction of the determined prediction direction; And restoring the current block based on at least one motion vector for prediction of the determined prediction direction and information on the decoded pixel values, wherein the first direction is from the current picture to a preceding picture. Direction, and the second direction is a direction from the current picture to a succeeding picture.

In order to solve the above technical problem, a motion vector of a current block is determined based on a motion vector of at least one block encoded before the current block according to an embodiment of the present invention, and the current block is encoded using the determined motion vector The method includes determining a motion vector in a first direction and a motion vector in a second direction of the current block based on a motion vector of at least one block encoded before the current block; Determining a prediction used for encoding the current block among prediction in the first direction, prediction in the second direction, and prediction in both directions based on the motion vector in the first direction and the motion vector in the second direction; And encoding information on the prediction direction and information on pixel values of the current block generated based on the determined prediction, wherein the first direction is a direction from the current picture to a preceding picture, and the The second direction is characterized in that it is a direction from the current picture to the trailing picture.

A motion vector of a current block is determined based on a motion vector of at least one block decoded before the current block according to an embodiment of the present invention for solving the above technical problem, and the current block is decoded using the determined motion vector The apparatus includes: a decoding unit decoding information on a prediction direction used for decoding of the current block and information on pixel values of the current block among first, second, and bidirectional directions; A motion vector determiner determining a prediction direction of the current block based on information on the decoded prediction direction and determining at least one motion vector for performing prediction of the determined prediction direction; And a reconstruction unit reconstructing the current block based on at least one motion vector for prediction of the determined prediction direction and information on the decoded pixel values, wherein the first direction is from the current picture to a preceding picture. Direction, and the second direction is a direction from the current picture to a succeeding picture.

In order to solve the above technical problem, a motion vector of a current block is determined based on a motion vector of at least one block encoded before the current block according to an embodiment of the present invention, and the current block is encoded using the determined motion vector The apparatus comprises: a motion vector determiner configured to determine a motion vector in a first direction and a motion vector in a second direction of a current block based on a motion vector of at least one block encoded before the current block; The prediction used for encoding the current block among the prediction in the first direction, the prediction in the second direction, and the prediction in both directions based on the motion vector in the first direction and the motion vector in the second direction is determined, and the prediction direction And an encoder for encoding information on pixel values of the current block generated based on the determined information and the determined prediction, wherein the first direction is a direction from the current picture to a preceding picture, and the second direction is It is characterized in that it is a direction from the current picture to the succeeding picture.

In order to solve the above technical problem, the present invention provides a computer-readable recording medium for executing the image encoding method and / or image decoding method.

According to the present invention, an occurrence probability of an encoding mode in which a motion vector of at least one previous block is encoded as a motion vector of a current block is increased, and accuracy of prediction is also increased, so that an image can be encoded and decoded at a higher compression rate.

1 shows an image encoding apparatus according to an embodiment of the present invention.
2 shows an image decoding apparatus according to an embodiment of the present invention.
3 shows a hierarchical coding unit according to an embodiment of the present invention.
4 illustrates an image encoder based on coding units according to an embodiment of the present invention.
5 illustrates an image decoder based on coding units according to an embodiment of the present invention.
6 illustrates a maximum coding unit, a sub coding unit, and a prediction unit according to an embodiment of the present invention.
7 illustrates a coding unit and a transformation unit, according to an embodiment of the present invention.
8A and 8B illustrate a split form of a coding unit, a prediction unit, and a frequency transform unit, according to an embodiment of the present invention.
9 shows a video encoding apparatus according to another embodiment of the present invention.
10 illustrates a method of predicting a block included in a bidirectional predictive picture according to an embodiment of the present invention.
11A and 11B illustrate a method of determining a motion vector of a current block based on a motion vector of a previously coded block according to an embodiment of the present invention.
12A and 12B illustrate a method of determining a motion vector of a current block based on a motion vector of a previously coded block according to another embodiment of the present invention.
13A and 13B illustrate a method of determining a motion vector of a current block based on a motion vector of a previously coded block according to another embodiment of the present invention.
14 shows a video decoding apparatus according to another embodiment of the present invention.
15 is a flowchart illustrating an image encoding method according to an embodiment of the present invention.
16 is a flowchart illustrating an image decoding method according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

1 shows an image encoding apparatus according to an embodiment of the present invention.

Referring to FIG. 1, an image encoding apparatus 100 according to an embodiment of the present invention includes a maximum coding unit splitter 110, an encoding depth determining unit 120, an image data encoding unit 130, and an encoding information encoding unit. 140.

The largest coding unit splitter 110 may split the current frame or the current slice based on the largest coding unit that is the largest coding unit. The current frame or current slice may be divided into at least one largest coding unit.

According to an embodiment of the present invention, a coding unit may be expressed using a maximum coding unit and depth. As described above, the largest coding unit represents the largest coding unit among the coding units of the current frame, and the depth indicates the degree to which the coding unit is hierarchically reduced. As the depth increases, a coding unit may be reduced from a maximum coding unit to a minimum coding unit, a depth of the maximum coding unit may be defined as a minimum depth, and a depth of the minimum coding unit may be defined as a maximum depth. Since the maximum coding unit decreases in size according to depths as the depth increases, a sub-coding unit of a k depth may include a plurality of sub-coding units having a depth greater than k.

As the size of an encoded frame increases, when an image is encoded in a larger unit, an image may be encoded with a higher image compression rate. However, if the coding unit is enlarged and the size is fixed, it is impossible to efficiently encode the image by reflecting the characteristics of the image that is continuously changing.

For example, when encoding a flat region for the sea or the sky, the larger the coding unit may be, the higher the compression rate may be, but when coding a complex region for people or buildings, the smaller the coding unit, the higher the compression rate.

To this end, an embodiment of the present invention sets a maximum image coding unit having a different size for each frame or slice, and sets a maximum depth. Since the maximum depth means the maximum number of times that a coding unit can be reduced, it is possible to variably set the size of the minimum coding unit included in the maximum video coding unit according to the maximum depth.

The coded depth determining unit 120 determines a maximum depth. The maximum depth may be determined based on R-D cost (Rate-Distortion Cost) calculation. The maximum depth may be determined differently for each frame or slice, or may be determined differently for each largest coding unit. The determined maximum depth is output to the encoding information encoder 140, and image data for each largest coding unit is output to the image data encoder 130.

The maximum depth means the smallest coding unit that can be included in the largest coding unit, that is, the smallest coding unit. In other words, the largest coding unit may be divided into sub-coding units of different sizes according to different depths. This will be described later in detail with reference to FIGS. 8A and 8B. Also, sub-coding units of different sizes included in the largest coding unit may be predicted or transformed based on processing units of different sizes. The transform may be a discrete cosine transform (KLT) or a Karhunen Loever Transform (KLT) as transforming pixel values in the spatial domain into coefficients in the frequency domain.

In other words, the image encoding apparatus 100 may perform a plurality of processing steps for image encoding based on various sizes and various types of processing units. In order to encode the image data, processing steps such as prediction, transformation, and entropy encoding are performed, and processing units of the same size may be used in all stages, and processing units of different sizes may be used in steps.

For example, the image encoding apparatus 100 may select a processing unit different from the coding unit in order to predict a predetermined coding unit.

When the size of the coding unit is 2Nx2N (where N is a positive integer), the processing unit for prediction may be 2Nx2N, 2NxN, Nx2N, NxN, or the like. In other words, motion prediction may be performed based on a processing unit in which at least one of the height or width of the coding unit is halved. Hereinafter, a processing unit that is the basis of prediction is referred to as a 'prediction unit'.

The prediction mode may be at least one of an intra mode, an inter mode, and a skip mode, and a specific prediction mode may be performed only on prediction units having a specific size or shape. For example, the intra mode can be performed only on a prediction unit having a square size of 2Nx2N and NxN. In addition, the skip mode can be performed only on prediction units having a size of 2Nx2N. If there are a plurality of prediction units in the coding unit, prediction may be performed on each prediction unit to select a prediction mode having the smallest coding error.

Also, the image encoding apparatus 100 may convert image data based on a processing unit having a different size from the encoding unit. In order to transform the coding unit, transformation may be performed based on a data unit having a size equal to or smaller than the coding unit. Hereinafter, the processing unit that is the basis of the conversion is referred to as a 'conversion unit'.

The encoding depth determiner 120 may determine sub-coding units included in the largest coding unit using a Lagrangian Multiplier-based rate-distortion optimization technique. In other words, it is possible to determine what type of maximum coding unit is divided into a plurality of sub-coding units, where the plurality of sub-coding units have different sizes according to depths. Then, the image data encoding unit 130 encodes the largest coding unit based on the division type determined by the encoding depth determination unit 120 and outputs a bitstream.

The encoding information encoding unit 140 encodes information about an encoding mode of a maximum coding unit determined by the encoding depth determination unit 120. A bitstream is output by encoding information on a split type of a largest coding unit, information on a maximum depth, and information on a coding mode of a sub-coding unit according to depths. Information about a coding mode of a sub coding unit may include information about a prediction unit of a sub coding unit, prediction mode information for each prediction unit, and information about a transformation unit of a sub coding unit.

The information on the split type of the largest coding unit may be information indicating whether to split each coding unit. For example, when the maximum coding unit is split and coded, information indicating whether or not the maximum coding unit is split is coded, and sub-coding units generated by splitting the maximum coding unit are split and coded again, respectively. Information indicating whether to split is coded for the sub-coding unit. Information indicating whether to divide or not may be flag information indicating whether to divide.

Since sub-coding units having different sizes exist for each largest coding unit, and information about a coding mode must be determined for each sub-coding unit, information about at least one coding mode may be determined for one largest coding unit.

As the depth increases, the image encoding apparatus 100 may generate a sub-coding unit by dividing the maximum coding unit in half by height and width. That is, if the size of the coding unit of the k depth is 2Nx2N, the size of the coding unit of the k + 1 depth is NxN.

Accordingly, the image encoding apparatus 100 according to an embodiment may determine an optimal division type for each maximum coding unit based on the size and maximum depth of the maximum coding unit in consideration of the characteristics of the image. It is possible to more efficiently encode images of various resolutions by variably adjusting the size of the largest coding unit in consideration of image characteristics, and splitting the largest coding unit into sub-coding units of different depths to encode the images.

2 shows an image decoding apparatus according to an embodiment of the present invention.

Referring to FIG. 2, the image decoding apparatus 200 according to an embodiment of the present invention includes an image data acquisition unit 210, an encoding information extraction unit 220, and an image data decoding unit 230.

The image-related data acquisition unit 210 parses the bitstream received by the image decoding apparatus 200, acquires image data for each largest coding unit, and outputs the image data to the image data decoding unit 230. The image data acquisition unit 210 may extract information on the maximum coding unit of the current frame or slice from the header for the current frame or slice. In other words, the bitstream is divided into the largest coding unit, and the image data decoder 230 decodes the image data for each largest coding unit.

The encoding information extraction unit 220 parses the bit stream received by the image decoding apparatus 200, and extracts a maximum encoding unit, a maximum depth, and a maximum encoding unit split form and a sub-coding unit encoding mode from a header for a current frame. Extract information about. Information on the split mode and the encoding mode is output to the image data decoder 230.

The information on the split type of the largest coding unit may include information on sub-coding units of different sizes according to depths included in the largest coding unit. As described above, the information on the split type may be information (eg, flag information) indicating whether or not to be coded for each coding unit. Information about a coding mode may include information about a prediction unit for each sub-coding unit, information about a prediction mode, and information about a transformation unit.

The image data decoding unit 230 restores the current frame by decoding the image data of each largest coding unit based on the information extracted by the encoding information extraction unit.

Based on the information on the split type of the largest coding unit, the image data decoder 230 may decode the sub coding unit included in the largest coding unit. The decoding process may include an inter prediction process including intra prediction and motion compensation, and an inverse transform process.

The image data decoder 230 may perform intra prediction or inter prediction based on information on a prediction unit for each sub coding unit and information on a prediction mode, for prediction of a sub coding unit. In addition, the image data decoder 230 may perform inverse transformation for each sub-coding unit based on information on the transform unit of the sub-coding unit.

3 shows a hierarchical coding unit according to an embodiment of the present invention.

Referring to FIG. 3, the hierarchical coding unit according to the present invention may include 32x32, 16x16, 8x8, and 4x4 from coding units having a width x height of 64x64. In addition to the square coding units, there may be coding units having a width x height of 64x32, 32x64, 32x16, 16x32, 16x8, 8x16, 8x4, 4x8.

Referring to FIG. 3, for video data 310 having a resolution of 1920x1080, a maximum coding unit size is set to 64x64 and a maximum depth of 2.

For the video data 320 having another resolution of 1920x1080, the size of the largest coding unit is 64x64 and the maximum depth is set to 4. For video data 330 having a resolution of 352x288, the size of the largest coding unit is set to 16x16 and the maximum depth is 2.

When the resolution is high or the amount of data is large, it is preferable that the maximum size of the encoding size is relatively large in order to accurately reflect image characteristics as well as to improve the compression rate. Accordingly, as compared to the image data 330, the image data 310 and 320 having high resolution may be selected to have a size of 64x64 for the largest coding unit.

The maximum depth represents the total number of layers in the hierarchical coding unit. Since the maximum depth of the image data 310 is 2, the coding unit 315 of the image data 310 is from the largest coding unit having a long axis size of 64, and sub-coding units having a long axis size of 32 or 16 as the depth increases. Can include up to.

On the other hand, since the maximum depth of the image data 330 is 2, the coding unit 335 of the image data 330 is encoded from the largest coding units having a long axis size of 16, and the long axis sizes of 8 and 4 are increased as the depth increases. Units may also be included.

Since the maximum depth of the image data 320 is 4, the coding unit 325 of the video data 320 is from the largest coding unit having a long axis size of 64, and as the depth increases, the long axis size is 32, 16, 8, or 4 It may include sub-coding units. As the depth increases, an image is encoded based on a smaller sub-coding unit, and thus it is suitable for encoding an image including a more detailed scene.

4 illustrates an image encoder based on coding units, according to an embodiment of the present invention.

The intra prediction unit 410 performs intra prediction on the prediction unit of the intra mode among the current frames 405, and the motion estimation unit 420 and the motion compensation unit 425 present the current frame ( 405) and the reference frame 495 to perform inter prediction and motion compensation.

Residual values are generated based on prediction units output from the intra prediction unit 410, the motion estimation unit 420, and the motion compensation unit 425, and the generated residual values are transform units 430 and quantization units ( 440) and output as a quantized transform coefficient.

The quantized transform coefficients are restored to residual values through the inverse quantization unit 460 and the inverse transform unit 470, and the restored residual values are post-processed through the deblocking unit 480 and the loop filtering unit 490. Is output to the reference frame 495. The quantized transform coefficient may be output to the bitstream 455 through the entropy encoder 450.

In order to encode according to an image encoding method according to an embodiment of the present invention, an intra prediction unit 410, a motion estimation unit 420, a motion compensation unit 425, and a transformation unit that are components of the image encoding unit 400 The 430, the quantization unit 440, the entropy encoding unit 450, the inverse quantization unit 460, the inverse transformation unit 470, the deblocking unit 480, and the loop filtering unit 490 are all the largest coding units and depths. The image encoding processes are processed based on a sub-coding unit, a prediction unit, and a transformation unit according to.

5 illustrates an image decoder based on coding units according to an embodiment of the present invention.

The bitstream 505 passes through the parsing unit 510, and the encoded image data to be decoded and encoding information necessary for decoding are parsed. The encoded image data is output as inverse-quantized data through an entropy decoding unit 520 and an inverse quantization unit 530, and is restored to residual values through an inverse transformation unit 540. The residual values are added to the result of the intra prediction of the intra prediction unit 550 or the result of the motion compensation of the motion compensation unit 560 and reconstructed for each coding unit. The reconstructed coding unit is used for prediction of a next coding unit or a next frame through a deblocking unit 570 and a loop filtering unit 580.

The parsing unit 510, the entropy decoding unit 520, the inverse quantization unit 530, and the inverse transform unit 540, which are components of the image decoding unit 400, for decoding according to the image decoding method according to an embodiment of the present invention ), The intra prediction unit 550, the motion compensation unit 560, the deblocking unit 570, and the loop filtering unit 580 are all based on the largest coding unit, sub-coding units according to depths, prediction units, and transformation units. Processes video decoding processes.

In particular, the intra prediction unit 550 and the motion compensation unit 560 determine the prediction unit and prediction mode in the sub coding unit in consideration of the maximum coding unit and depth, and the inverse transform unit 540 takes into account the size of the transformation unit Inverse transformation is performed.

6 illustrates a maximum coding unit, a sub coding unit, and a prediction unit according to an embodiment of the present invention.

The image encoding apparatus 100 and the image decoding apparatus 200 according to an embodiment of the present invention use hierarchical coding units to perform encoding and decoding in consideration of image characteristics. The maximum coding unit and the maximum depth may be adaptively set according to characteristics of an image, or may be variously set according to a user's request.

The hierarchical structure 600 of the coding unit according to an embodiment of the present invention shows a case where the maximum coding unit 610 has a height and a width of 64 and a maximum depth of 4. Depth increases along the vertical axis of the hierarchical structure 600 of the coding unit, and widths and heights of the sub coding units 620 to 650 decrease as the depth increases. Also, along the horizontal axis of the hierarchical structure 600 of coding units, prediction units of the largest coding unit 610 and the sub coding units 620 to 650 are illustrated.

The maximum coding unit 610 has a depth of 0, and the size of the coding unit, that is, width and height, is 64x64. Depth increases along the vertical axis, sub-coding unit 620 of depth 1 of size 32x32, sub-coding unit 630 of depth 2 of size 16x16, sub-coding unit 640 of depth 3 of size 8x8, size 4x4 There is a sub-coding unit 650 of depth 4. The sub-coding unit 650 having a size of 4x4 and having a depth of 4 is a minimum coding unit.

Referring to FIG. 6, examples of prediction units along a horizontal axis for each depth are illustrated. That is, the prediction unit of the largest coding unit 610 of depth 0 is the same as or smaller than the coding unit 610 of size 64x64, the prediction unit 610 of size 64x64, the prediction unit 612 of size 64x32, size 32x64 The prediction unit 614 may be a prediction unit 616 having a size of 32x32.

The prediction unit of the coding unit 620 of the size 32x32 of the depth 1 is the prediction unit 620 of the size 32x32 which is the same or smaller than the coding unit 620 of the size 32x32, the prediction unit 622 of the size 32x16, the size 16x32 The prediction unit 624 may be a prediction unit 626 having a size of 16x16.

The prediction unit of the size 16x16 coding unit 630 of the depth 2 is the size 16x16 prediction unit 630, the size 16x8 prediction unit 632, the size 8x16, which is the same or smaller than the size 16x16 coding unit 630. The prediction unit 634 of, may be a prediction unit 636 of size 8x8.

The prediction unit of the coding unit 640 of the size 8x8 of the depth 3 is the prediction unit 640 of the size 8x8, the prediction unit 642 of the size 8x4, the size 4x8 of the prediction unit 640 of the size equal to or smaller than the coding unit 640 of the size 8x8 The prediction unit 644 may be a prediction unit 646 having a size of 4x4.

Lastly, a coding unit 650 having a depth of 4 and a size of 4x4 is a coding unit having a maximum depth, and a prediction unit is a prediction unit 650 having a size of 4x4. However, the coding units of the maximum depth do not necessarily have the same size of the coding unit and the prediction unit, and like other coding units 610 to 650, prediction may be performed by dividing into prediction units having a size smaller than the coding unit. have.

7 illustrates a coding unit and a transformation unit, according to an embodiment of the present invention.

The video encoding apparatus 100 and the video decoding apparatus 200 according to an embodiment of the present invention encode the same as the maximum coding unit, or split and encode the maximum coding unit into sub-coding units smaller than or equal to the maximum coding unit. The size of the transform unit for transformation during the encoding process may also be selected as the size for the highest compression rate regardless of the coding unit and the prediction unit. For example, when the current coding unit 710 is 64x64 in size, transformation may be performed using a 32x32 sized transformation unit 720.

8A and 8B show a split form of a coding unit, a prediction unit, and a transformation unit, according to an embodiment of the present invention.

8A illustrates coding units and prediction units according to an embodiment of the present invention.

The left side of FIG. 8A shows a split form selected by the image encoding apparatus 100 according to an embodiment of the present invention to encode the largest coding unit 810. The video encoding apparatus 100 divides the largest coding unit 810 into various forms, encodes, and compares the encoding results of the various types of splits based on the R-D cost to select an optimal split form. When it is optimal to encode the maximum coding unit 810 as it is, the maximum coding unit 800 may be encoded without dividing the maximum coding unit 810 as illustrated in FIGS. 8A and 8B.

Referring to the left side of FIG. 8A, the largest coding unit 810 having a depth of 0 is split into sub-coding units having a depth of 1 or greater, and then encoded. The maximum coding unit 810 is divided into four sub-coding units of depth 1, and then all or part of the sub-coding units of depth 1 is further divided into sub-coding units of depth 2.

Among the sub-coding units of depth 1, a sub-coding unit shouting at the upper right and a sub-coding unit located at the lower left are divided into sub-coding units having a depth of 2 or higher. Some of the sub coding units having a depth of 2 or higher may be divided into sub coding units having a depth of 3 or higher.

The right side of FIG. 8B shows a split form of a prediction unit with respect to the largest coding unit 810.

Referring to the right side of FIG. 8A, the prediction unit 860 for the largest coding unit may be divided differently from the largest coding unit 810. In other words, the prediction unit for each of the sub coding units may be smaller than the sub coding unit.

For example, a prediction unit for a sub-coding unit 854 that is placed in the lower right of the sub-coding units of depth 1 may be smaller than the sub-coding unit 854. Prediction units for some sub-coding units 815, 816, 850, and 852 of the sub-coding units 814, 816, 818, 828, 850, and 852 of depth 2 may be smaller than the sub-coding units.

Also, a prediction unit for sub-coding units 822, 832, and 848 of depth 3 may be smaller than a sub-coding unit. The prediction unit may have a form in which each sub-coding unit is divided in a height or width direction, or may be in a form of a quadrant in the height and width direction.

8B illustrates prediction units and transformation units according to an embodiment of the present invention.

The left side of FIG. 8B shows a split form of a prediction unit for the largest coding unit 810 illustrated on the right side of FIG. 8A, and the right side of FIG. 8B shows a split form of a transform unit of the maximum coding unit 810.

Referring to the right side of FIG. 8B, the division form of the transformation unit 870 may be set differently from the prediction unit 860.

For example, even if the prediction unit for the coding unit 854 of the depth 1 is selected in a form in which the height is halved, the transform unit may be selected to have the same size as the size of the coding unit 854 of the depth 1. Likewise, even if the prediction unit for the coding units 814 and 850 of depth 2 is selected in a form in which the heights of the coding units 814 and 850 of depth 2 are selected in half, the transformation unit is the coding unit 814 and 850 of depth 2 It can be selected in the same size as the original size.

A transform unit may be selected with a size smaller than the prediction unit. For example, when the prediction unit for the coding unit 852 of depth 2 is selected in a form in which the width is halved, the transformation unit may be selected in a form in which the height and width, which are smaller than the prediction unit, are halved.

9 shows a video encoding apparatus according to another embodiment of the present invention. The image encoding apparatus 900 of FIG. 9 may be included in the image encoding apparatus 100 of FIG. 1 and the image encoding apparatus 400 of FIG. 4 and may be an apparatus for encoding a current block based on inter prediction.

Referring to FIG. 9, the image encoding apparatus 900 includes a motion vector determination unit 910 and an encoding unit 920.

In some encoding modes, a motion vector of a current block is determined based on a motion vector of at least one block encoded before the current block, and a current block is encoded based on the determined motion vector. Direct prediction mode and skip mode are examples of these modes. Both the direct prediction mode and the skip mode are coding modes that do not separately encode a motion vector as information about the current block because the motion vector of the current block is determined based on previously coded information.

However, in the direct prediction mode, the residual block generated by subtracting the prediction block generated according to the motion vector from the current block is encoded as information about the pixel value, whereas in the skip mode, the prediction block is regarded as the same as the current block and is skipped. It is different to encode only the flag information indicating that the mode has been encoded as information about the pixel value.

Although both the direct prediction mode and the skip mode are encoding based on inter prediction, since the motion vector is not separately encoded, the compression rate is greatly affected. However, the direct prediction mode and the skip mode are disadvantageous in that the probability of occurrence is low and the probability of incorrect prediction is high because the current block is encoded by performing only prediction in a specific prediction direction. Therefore, when encoding the current block included in the bidirectional prediction picture in the direct prediction mode or the skip mode, the image encoding apparatus 900 according to an embodiment of the present invention predicts and encodes various prediction directions. This will be described in detail with reference.

10 illustrates a method of predicting a block included in a bidirectional predictive picture according to an embodiment of the present invention.

Referring to FIG. 10, when encoding a current block 1010 included in a current picture 1000 that is a bidirectional predictive picture, the current block is one of a picture 1020 at a time before the current picture and a picture 1030 at a time after the current picture. It is predictively coded with reference to at least one. Search for at least one block 1022 and 1032 corresponding to the current block 1010 from at least one of a previous time picture 1020 and a later time picture 1030, and predictively encode the current block based on the searched block do.

In other words, when the direction from the current picture 1000 to the previous picture 1020 is referred to as a first direction (ie, L0 direction), the current block 1010 is based on the corresponding block 1022 retrieved from the picture in the first direction. Can be predicted. (Hereinafter referred to as 'first direction prediction'.) Similarly, when the direction from the current picture 1000 to the subsequent picture 1030 is referred to as a second direction (ie, L1 direction), the current block 1010 is the second direction It can be predicted based on the corresponding block 1032 retrieved from the picture of. (Hereinafter, it is referred to as “second direction prediction.”) Also, the current block 1010 may be predicted based on both first direction prediction and second direction prediction, that is, bidirectional prediction.

However, the direct prediction mode according to the related art can be applied only when bidirectional prediction is performed on the current block 1010. In other words, when the current block is predictively coded by performing only the first direction prediction, it cannot be encoded according to the direct prediction mode, and even when the current block is predictively encoded by performing only the second direction prediction, encoding is performed according to the direct prediction mode. Can't. The same is true for the skip mode according to the prior art. According to the skip mode according to the prior art, the motion vector of the current block 1010 determined by the motion vector of the previously coded block adjacent to the current block 1010 should be the motion vector (mv_L0) in the first direction, and the first The prediction block generated based on the direction prediction is regarded as the same as the current block 1010, and flag information indicating that it is encoded in the skip mode is encoded.

In short, according to the prior art, although the prediction that can be performed on a block included in the bidirectional prediction picture is the first direction prediction, the second direction prediction, and the bidirectional prediction, the direct prediction mode and the skip mode predict prediction in a specific direction. Since it can be applied only when performing, there is a disadvantage that the direct prediction mode and the skip mode, which greatly affect the compression rate, are limitedly applied.

To solve this disadvantage, the video encoding apparatus 900 according to the present invention determines the motion vector of the current block 1010 based on previously encoded information, without encoding the motion vector, and prediction is performed in all directions Encoding is performed according to an image encoding method that can be performed for.

The motion vector determiner 910 determines a first direction motion vector and a second direction motion vector of the current block 1010 to perform first direction prediction, second direction prediction, and bidirectional prediction. The first direction motion vector and the second direction motion vector are determined based on the motion vector of at least one block encoded before the current block.

The first direction motion vector and the second direction motion vector may be determined according to a conventional motion vector determination method. For example, when the current block 1010 is encoded according to a direct prediction mode, that is, without encoding information about a motion vector, and only encoding information about a residual block, the H.264 / AVC direct prediction mode is used. A motion vector may be determined according to a method for determining a motion vector. When the current block 1010 is encoded in a skip mode, that is, in a mode that does not encode all information about a motion vector and information about a residual block, according to a method of determining a motion vector in an H.264 / AVC skip mode The motion vector can be determined.

The H.264 / AVC direct prediction mode includes a temporal direct prediction mode and a spatial direct prediction mode. Accordingly, the method for determining the motion vector of the current block in the temporal direct prediction mode of H.264 / AVC and the method for determining the motion vector of the current block in the spatial direct prediction mode are the motion vector determining unit 910 moving in the first direction. It can be used to determine a vector and a second direction motion vector. This will be described in detail with reference to FIGS. 11A and 11B.

11A illustrates a method of determining a motion vector of a current block based on a motion vector of a previously coded block according to an embodiment of the present invention.

Referring to FIG. 11A, when the current block 1010 is encoded in a time direct prediction mode, the motion vector of the current block 1010 of the current picture 1110 is colocated with respect to the temporally trailing picture 1114. It can be generated using the motion vector of the block 1120 of.

In other words, when the temporally trailing picture 1114 is called an anchor picture, the first direction movement of the current block based on the motion vector of the block 1120 at the same position of the current block in the anchor picture 1114 The vector and the second direction motion vector can be determined. The motion vector of the block 1120 in the same position is based on the temporal distance between the current picture 1110 and the preceding picture 1112 and the time distance between the anchor picture 1114 and the preceding picture 1112 The first motion vector and the second motion vector of the current block 1010 may be determined by scaling the motion vector of 1120.

For example, if the motion vector mv_colA of the block 1120 at the same position as the current block 1010 is generated for the searched block 1122 of the picture 1112 at a time before the current picture 1010, the current block 1110 The first direction motion vectors and the second direction motion vectors of mv_L0A and mv_L1A may be generated as follows.

mv_L0A = (t1 / t2) * mv_colA

mv_L1A = mv_L0A-mv_colA

Here, mv_L0A means the first direction motion vector of the current block 1010, and mv_L1A means the second direction motion vector of the current block 1010.

11B illustrates a method of determining a motion vector of a current block based on a motion vector of a previously coded block according to an embodiment of the present invention.

Referring to FIG. 11B, the first direction motion vector and the second direction motion vector of the current block 1010 are based on the motion vectors mv_A, mv_B, and mv_C of blocks 1130 to 1134 adjacent to the current block 1100. Can be determined. The first direction motion vector and the second direction motion vector are determined based on the motion vectors of the blocks located on the left, upper, and upper right sides of the current block 1010.

When the current block is encoded in the spatial direct prediction mode, the median of motion vectors in the first direction among mv_A, mv_B, and mv_C is determined as a first direction motion vector of the current block 1010, and a second of mv_A, mv_B, and mv_C The median of the motion vectors in the direction is determined as the second motion vector of the current block 1010. When determining the first direction motion vector, the motion vector that is the second direction of mv_A, mv_B, and mv_C is set to '0' to calculate the median value, and when determining the second direction motion vector, the first of mv_A, mv_B, and mv_C The direction motion vector is set to '0' and the median is calculated.

A picture corresponding to the smallest number among first direction reference numbers according to mv_A, mv_B, and mv_C is determined as a first direction reference picture, and a picture corresponding to the smallest number of second direction reference numbers according to mv_A, mv_B, and mv_C Is determined as the second direction reference picture.

When the current block 1010 is encoded in the skip mode, as shown in FIG. 11B, the median of the first direction motion vectors among the motion vectors mv_A, mv_B, and mv_C of blocks 1130 to 1134 adjacent to the current block 1010 Can be determined as the motion vector of the current block.

When the video encoding apparatus 900 determines the motion vector according to FIGS. 11A and 11B and encodes the current block, the decoding side may determine and reconstruct the motion vector of the current block using the same method as FIGS. 11A and 11B. . In other words, the side that encodes the image and the side that decodes the image can implicitly share the same motion vector determination method, and determine the motion vector of the current block according to the shared motion vector determination method.

However, according to another embodiment of the present invention, the encoding side generates a plurality of motion vector candidates based on motion vectors of a plurality of blocks encoded before the current block, and the current block from the generated plurality of motion vector candidates The information for specifying the motion vector of is explicitly encoded and inserted into the Beaststream, and the decoding side can determine the motion vector of the current block based on the information that is explicitly encoded and inserted. It will be described in detail below with reference to FIGS. 12A and 12D.

12A and 12B illustrate a method of determining a motion vector of a current block based on a motion vector of a previously coded block according to another embodiment of the present invention.

Referring to FIG. 12A, the motion vector determiner 910 may generate a plurality of motion vector candidates based on motion vectors of previously coded blocks adjacent to the current block. Among the blocks adjacent to the top of the current block, the motion vectors of the leftmost a0 block, the uppermost b0 block, the upper right c block, the upper left d block, and the lower right e block motion vectors It may be a motion vector candidate.

One of the motion vectors in the first direction among motion vectors of the a0 block, the b0 block, the c block, and the d block is determined as the motion vector in the first direction of the current block, and the a0 block, b0 block, c block, and d block One of the motion vectors in the second direction among the motion vectors may be determined as a motion vector in the second direction of the current block. For example, if the motion vectors of the a0 block and the b0 block among the motion vectors of the a0 block, the b0 block, the c block, and the d block are motion vectors in the first direction, one of the motion vector of the a0 block and the motion vector of the b0 block The first direction motion vector of the current block may be determined, and one of the motion vectors of the remaining c blocks and d blocks may be determined as the second direction motion vector of the current block.

Although there is no limitation in the method of determining one of the plurality of motion vector candidates as the motion vector of the current block, it is generally possible to select a motion vector that can more accurately predict the current block. In the above example, when prediction is performed according to the motion vector of the a0 block among the motion vector of the a0 block and the motion vector of the b0 block, which are motion vectors in the first direction, if the current block is predicted more accurately, the motion of the a0 block The vector may be determined as the first direction motion vector of the current block. The prediction block generated as a result of prediction according to each motion vector is subtracted from the current block to calculate the sum of absolute difference (SAD) of the residual block, and a motion vector having a small SAD can be determined as the motion vector of the current block. have.

Referring to FIG. 12B, motion vectors of all blocks adjacent to the current block may be used as a motion vector candidate. In other words, a motion vector of not only the leftmost a0 block among blocks adjacent to the top but also all blocks adjacent to the top can be used as a motion vector candidate, and the topmost b0 block among blocks adjacent to the left side is adjacent to the left side as well. The motion vector of all blocks can be used as a motion vector candidate.

As described above with reference to FIG. 12A, in the motion vector candidate of all blocks, one of the motion vectors in the first direction is determined as the first direction motion vector of the current block, and one of the motion vectors in the second direction is the current block. It can be determined as a second direction motion vector of.

In summary, according to FIGS. 11A, 11B, 12A, and 12B, the motion vector determiner 910 may determine one of a plurality of motion vector candidates as a motion vector of the current block as follows.

C = {median (mv_a0, mv_b0, mv_c), mv_a0, mv_a1 ..., mv_aN, mv_b0, mv_b1, ..., mv_bN, mv_c, mv_d, mv_e, mv_L0A, mv_L1A}

mv_L0A and mv_L1A may be motion vectors according to FIG. 11A, and median (mv_a0, mv_b0, mv_c) may be motion vectors according to FIG. 11B. The remaining motion vectors are described above with respect to FIGS. 12A and 12B.

A first direction motion vector of the current block may be determined from among motion vectors of the first direction included in the C set, and a second direction motion vector of the current block may be determined from among motion vectors of the second direction included in the C set. .

In addition, according to another embodiment of the present invention, the number of candidates may be reduced, and among them, a motion vector of the current block may be determined.

C = {median (mv_a ', mv_b', mv_c '), mv_a', mv_b ', mv_c', mv_L0A, mv_L1A}

Here, mv_x means the motion vector of the x block, median () means the median value, and mv_a 'means the first valid motion vector among mv_a0, mv_a1 ..., and mv_aN. For example, when the a0 block is coded using intra prediction, mv_a0, which is a motion vector of a0, is invalid, so mv_a '= mv_a1, and when the motion vector of the a1 block is also invalid, mv_a' = mv_a2.

Similarly, mv_b 'denotes the first valid motion vector among mv_b0, mv_b1, ..., mv_bN, and mv_c' denotes the first valid motion vector among mv_c, mv_d, and mv_e.

mv_L0A and mv_L1A may be motion vectors according to FIG. 11A. Further, according to another embodiment of the present invention, other motion vectors generated using time distances between adjacent pictures may be included in candidates instead of mv_L0A and mvL1A. This will be described in detail with reference to FIGS. 13A and 13B.

13A and 13B illustrate a method of determining a motion vector of a current block based on a motion vector of a previously coded block according to another embodiment of the present invention.

Referring to FIG. 13A, a motion vector candidate determined based on a time distance may be determined in a different method from a time direct prediction mode according to the prior art.

The motion vector candidate of the current block 1010 of the current picture 1110 may be generated using the motion vector of the block 1320 at the same position of the preceding picture 1112. For example, if the motion vector mv_colB of the block 1320 at the same position as the current block 1010 is generated for the searched block 1322 of the trailing picture 1114, the motion vector candidates of the current block 1010 are mv_L0B and mv_L1B may be generated as follows.

mv_L1B = (t3 / t4) * mv_colB

mv_L0B = mv_L1B-mv_colB

Here, mv_L0B means a first direction motion vector candidate of the current block 1010, and mv_L1B means a second direction motion vector candidate of the current block 1010.

Referring to FIG. 13B, a motion vector candidate of the current block 1010 of the current picture 1110 may be determined using the motion vector of the block 1330 at the same position of the preceding picture 1112. For example, if the motion vector mv_colC of the block 1330 at the same position as the current block 1010 is generated for the searched block 1332 of another preceding picture 1116, the motion vector candidate of the current block 1010 is mv_L0C may be generated as follows.

mv_L0C = (t6 / t5) x mv_colC

13A and 13B, the aforementioned mv_L0B, mv_L1B, and mv_L0B may be included in the C set of candidates of the motion vector of the current block 1010 described above. Instead of mv_L0A and mv_L1A, they may be included in the C set, or may be additionally included in the C set.

Referring back to FIG. 9, when the motion vector determiner 910 determines the motion vector of the current block based on the motion vectors of blocks encoded before the current block, the encoder 920 encodes information about the current block. do.

When the first direction motion vector and the second direction motion vector of the current block are determined, the encoder 920 predicts the first direction and the second direction based on the determined first direction motion vector and the second direction motion vector. And bidirectional prediction. Then, the prediction direction for predicting the current block most accurately is determined. The prediction direction in the first direction, the prediction result in the second direction, and the bidirectional prediction result may be compared with the current block to determine a prediction direction with the lowest SAD of the residual block. Prediction in the first direction, prediction in the second direction, and bidirectional prediction are all performed, and the coded results may be compared based on an RD cost (rate distortion cost) to determine a prediction direction.

When the prediction direction is determined, information about the determined prediction direction and pixel values is encoded. Information about the prediction direction may be encoded as a sequence parameter, slice parameter, or block parameter. A predetermined binary number is allocated to the first direction, the second direction, and both directions, and the binary number corresponding to the prediction direction used for encoding of the current block can be entropy-encoded and inserted into the bitstream.

As information on pixel values, information on a residual block, such as a direct prediction mode, may be separately coded. The prediction block generated as a result of prediction according to the determined prediction direction is subtracted from the current block to generate a residual block, transform the residual block, generate coefficients in the frequency domain, and quantize and entropy the generated coefficients. . The transform can be a discrete cosine transform or a KLT transform.

Also, as information on pixel values, only the skip mode according to the related art and the flag information indicating that the prediction block and the current block are the same may be encoded. The prediction block generated as a result of prediction according to the determined prediction direction is regarded as the same as the current block, and only flag information indicating that the current block and the prediction block are the same can be encoded.

A picture referenced for prediction may be inferred based on a reference index of previously coded blocks. As described above with reference to FIG. 11B, a reference number of a current block is determined based on a reference number of previously coded blocks, and a reference picture corresponding thereto may be determined. Also, a reference picture corresponding to the motion vector determined by the motion vector determining unit 910 may be determined as a reference picture of the current block. For example, as described above with reference to FIG. 11A, if mv_L0A is determined as the first direction motion vector of the current block, a preceding picture 1112 according to mv_L0A may be automatically determined as a reference picture of the current block.

When the motion vector determination unit 910 explicitly determines the first direction motion vector and the second direction motion vector among a plurality of motion vector candidates as shown in FIGS. 12A, 12B, 13A, and 13B, the encoder 920 may include a plurality of motion vector. The information indicating the at least one of the first direction motion vector and the second direction motion vector of the current block may be encoded in the candidate vector candidate.

When it is determined that prediction is performed only in the first direction as a result of determining the prediction direction, only information indicating the first direction motion vector is encoded by a plurality of motion vector candidates. Even when prediction is performed only in the second direction, only information indicating the second direction motion vector may be encoded. However, in the case of performing bidirectional prediction, information indicating a first direction motion vector and a second direction motion vector may be encoded from a plurality of motion vector candidates.

Binary numbers may be allocated to each of the motion vectors included in the above-described C set, and among them, binary numbers assigned to the first direction motion vector and / or the second direction motion vector determined as the motion vector of the current block may be encoded. However, as described above, the first direction motion vector of the current block is determined as one of the first direction motion vectors included in the C set, and is sequentially assigned to the first direction motion vectors included in the C set. Among binary numbers, a binary number allocated to a motion vector determined as a first direction motion vector of a current block may be encoded as information indicating the first direction motion vector. Likewise, in the second direction motion vector, the second direction motion vector is the binary number assigned to the motion vector determined as the second direction motion vector of the current block among the binary numbers sequentially assigned to the second direction motion vectors included in the C set. It can be encoded as indicated information.

14 shows a video decoding apparatus according to another embodiment of the present invention. The image decoding apparatus 1400 of FIG. 14 may be included in the image decoding apparatus 200 of FIG. 2 and the image encoding apparatus 500 of FIG. 5 to perform inter prediction to decode a current block.

Referring to FIG. 14, an image decoding apparatus 1400 according to an embodiment of the present invention includes a decoding unit 1410, a motion vector determining unit 1420, and a restoration unit 1430.

The decoder 1410 receives a bitstream including data for a current block, and decodes information about a prediction direction. The image decoding apparatus 1400 determines a motion vector of the current block based on a motion vector of at least one block decoded before the current block, and performs decoding according to a decoding mode that restores the current block based on the determined motion vector Therefore, before determining the motion vector of the current block, information on the prediction direction of the current block is decoded. As a sequence parameter, slice parameter, or block parameter, information about a prediction direction included in a bitstream is decoded.

An encoding / decoding mode for determining a motion vector of a current block based on a motion vector of a previously coded block, such as a direct prediction mode or a skip mode, encodes / decodes a current block using only prediction in a specific direction. According to the encoding / decoding method, the current block may be encoded / decoded in a direct prediction mode or a skip mode based on both prediction in the first direction, prediction in the second direction, and bidirectional prediction. To this end, the above-described image encoding apparatus 900 encodes information about a prediction direction and inserts it into a bitstream. The decoder 1410 decodes information about a prediction direction inserted into the Beaststream by the video encoding apparatus 900.

The decoder 1410 may decode information indicating a motion vector of a current block from a plurality of motion vector candidates along with information on a prediction direction. When the motion vector of the current block is implicitly determined according to a method of determining a motion vector shared between the encoding side and the decoding side as described above, the motion vector of the current block is determined as one according to a predetermined method. There is no need to decrypt the indicated information. However, as described above, a plurality of motion vector candidates are generated based on motion vectors of a plurality of previously coded blocks, and when a motion vector of the current block is determined, a motion vector of the current block from a plurality of motion vector candidates Information for specifying can be decoded.

In addition, the decoder 1410 decodes information on pixel values of the current block. The information on the residual block of the current block is decoded, or flag information indicating that the current block is the same as the prediction block is decoded. The decoder 1410 may restore the residual block by entropy decoding, inverse quantizing, and inverse transforming information about the residual block.

The motion vector determiner 1420 determines a motion vector of a current block for performing prediction in a predetermined direction based on information about a prediction direction decoded by the decoder 1410. As a result of referring to the decoded prediction direction information, if it is determined to perform prediction in the first direction, the first direction motion vector is determined, and if it is determined to perform prediction in the second direction, the second direction motion vector is determined. Decide. If it is determined to perform bidirectional prediction, the first direction motion vector and the second direction motion vector are determined.

When the motion vector of the current block is implicitly determined according to the shared motion vector determination method between the encoding side and the decoding side, the motion vector of the current block is determined according to the shared motion vector determination method. The method for implicitly determining the motion vector of the current block has been described above with reference to FIGS. 11A and 11B.

However, when determining a motion vector of a current block among a plurality of motion vector candidates, the motion vector of the current block is determined based on information indicating the current block from the plurality of motion vector candidates decoded by the decoder 1410. A plurality of motion vector candidates are generated using the motion vectors of at least one block decoded before the current block, and a motion vector of the current block is determined from the generated plurality of motion vector candidates.

When the current block uses only prediction in the first direction, the decoder 1410 determines a first direction motion vector among a plurality of motion vector candidates by referring to information indicating the decoded first direction motion vector, and the current block When only the prediction of the second direction is used, the second direction motion vector is determined from among a plurality of motion vector candidates by referring to information indicating the decoded second direction motion vector by the decoder 1410. When the current block uses bidirectional prediction, the first direction motion among the plurality of motion vector candidates is referred to by referring to information indicating the first direction motion vector and the second direction motion vector decoded by the decoder 1410. The vector and the second direction motion vector can be determined.

The reconstruction unit 1430 reconstructs the current block based on information about pixel values decoded by the decoder 1410 and a motion vector of the current block determined by the motion vector determination unit 1420.

The prediction of the current block is generated by performing prediction in the first direction, prediction in the second direction, or prediction in both directions according to the motion vector determined by the motion vector determining unit 1410. As a result of decoding the information on the pixel values by the decoder 1410, when the residual block is restored, the current block is reconstructed by adding the restored residual block and the prediction block. When the flag information indicating that the prediction block is the same as the current block is decoded by the decoder 1410, the prediction block is used as the restored current block.

A picture referenced for prediction may be inferred based on a reference index of previously decoded blocks. As described above with reference to FIG. 11B, the reference number of the current block is determined based on the reference number of previously decoded blocks, and a reference picture corresponding thereto may be determined. Also, a reference picture corresponding to a motion vector determined by the motion vector determining unit 1410 may be determined as a reference picture of the current block. For example, as described above with reference to FIG. 11A, if mv_L0A is determined as the first direction motion vector of the current block, a preceding picture 1112 according to mv_L0A may be automatically determined as a reference picture of the current block.

15 is a flowchart illustrating an image encoding method according to an embodiment of the present invention.

15, in step 1510, the image encoding apparatus determines a motion vector of the current block. The first direction motion vector and the second direction motion vector of the current block included in the current picture, which is a bidirectional prediction picture, are determined. 11A and 11B, the motion vector of the current block may be implicitly determined according to a shared motion vector determination method between the encoding side and the decoding side as described above, and refer to FIGS. 12A, 12B, 13A, and 13B. As described above, a plurality of motion vector candidates may be generated, and a first direction motion vector and a second direction motion vector may be determined from among the generated plurality of motion vector candidates.

In operation 1520, the video encoding apparatus determines prediction used for encoding the current block among first direction prediction, second direction prediction, and bidirectional prediction. A prediction used for encoding of a current block is determined among first direction prediction according to a first direction motion vector, second direction prediction according to a second direction motion vector, bidirectional prediction according to a first direction motion vector and a second direction motion vector do.

Among the first direction prediction, the second direction prediction, and the bidirectional prediction, a prediction that can predict the current block most accurately can be determined as a prediction used to encode the current block. As described above, the longest prediction between the prediction block and the SAD of the current block may be determined as a prediction used to encode the current block.

In operation 1530, the image encoding apparatus encodes information about a prediction direction of prediction determined in operation 1530. Binary numbers may be allocated to the first direction, the second direction, and both directions, and binary numbers corresponding to the prediction direction used for encoding the current block may be encoded. The encoded binary number may be entropy encoded by a predetermined entropy encoding method. Information about the prediction direction may be encoded as a sequence parameter, slice parameter, or block parameter.

Information about pixel values of the current block is also encoded, and information about residual values of the residual block generated by subtracting the prediction block from the current block may be encoded. The residual block can be transformed, quantized, and entropy-encoded. Also, as described above, only the flag information indicating that the prediction block is the same as the current block may be encoded.

Information for specifying a motion vector used for encoding the current block from a plurality of motion vector candidates may also be encoded. A plurality of motion vectors are generated when a plurality of motion vector candidates are generated based on the motion vectors of at least one block encoded before the current block, and one of them is selected as a motion vector of the current block by selecting one of them. Information indicating a motion vector selected from candidates may be encoded. If the first direction is determined in the prediction direction of the current block in step 1520, only information about the first direction motion vector is encoded, and in step 1530, if the second direction is determined in the prediction direction of the current block, in the second direction motion vector Only information about can be encoded. If the bidirectional prediction is determined in the prediction direction of the current block in step 1520, both information about the first direction motion vector and information about the second direction motion vector are encoded.

16 is a flowchart illustrating an image decoding method according to an embodiment of the present invention.

Referring to FIG. 16, in step 1610, the image decoding apparatus decodes information about a prediction direction used for decoding a current block among first, second, and bidirectional directions. The information about the prediction direction included in the bitstream is decoded. Binary numbers assigned to each of the first direction, the second direction, and both directions decode binary numbers corresponding to the prediction direction used for decoding the current block. As a sequence parameter, slice parameter, or block parameter, information about a prediction direction included in a bitstream is decoded.

In addition, the image decoding apparatus decodes information about pixel values of the current block in step 1610. Information about residual values included in the residual block of the current block may be decoded. The residual block is reconstructed by entropy decoding, inverse quantization, and inverse transforming data on the residual block. Information indicating that the current block is the same as the prediction block may be decoded.

When the motion vector of the current block is determined from the plurality of motion vector candidates described above, information for specifying a motion vector used for decoding the current block from the plurality of motion vector candidates may be decoded. As a result of decoding the information about the prediction direction, if the prediction direction of the current block is the first direction, information indicating a first direction motion vector of the current block from a plurality of motion vector candidates may be decoded, and the prediction direction of the current block If this is the second direction, information indicating the second direction motion vector of the current block may be decoded from a plurality of motion vector candidates. If the prediction direction of the current block is bidirectional, information indicating the first direction motion vector and the second direction motion vector from a plurality of motion vector candidates may be decoded.

In step 1620, the image decoding apparatus determines a prediction direction of the current block based on the information about the prediction direction decoded in step 1610. In step 1610, the prediction direction of the current block is determined based on the binary number corresponding to the decoded prediction direction. When determined in the prediction direction, at least one motion vector for performing prediction of the determined prediction direction is determined.

As described above, a motion vector of a current block may be implicitly determined according to a method of determining a motion vector shared between a side that encodes and a side that decodes. In addition, when a motion vector of a current block is explicitly selected from a plurality of motion vector candidates, a plurality of motion vector candidates are generated based on the motion vectors of at least one block decoded before the current block, and the generated plurality of motions From the vector candidate, the motion vector of the current block can be determined. The motion vector of the current block may be determined based on information indicating the motion vector of the current block from the plurality of motion vector candidates decoded in step 1610. At least one of a first direction motion vector and a second direction motion vector may be determined based on the information decoded in step 1610.

In step 1630, the video decoding apparatus restores the current block. A prediction block is generated based on the motion vector of the current block determined in step 1620, and the current block is reconstructed based on the generated prediction block.

The prediction block of the current block is generated by performing one of the first direction prediction, the second direction prediction, and the bi-directional prediction based on the first direction motion vector and / or the second direction motion vector of the current block determined in step 1620.

When the residual block is restored in step 1610, the current block is restored by adding the generated prediction block and the restored residual block. In step 1610, when information indicating that the prediction block is the same as the current block and the current block is decoded, the prediction block becomes the current block.

As described above, although the present invention has been described by limited embodiments and drawings, the present invention is not limited to the above embodiments, and various modifications and modifications from these descriptions will be made by those skilled in the art to which the present invention pertains. Deformation is possible. Accordingly, the spirit of the present invention should be understood only by the claims set forth below, and all equivalent or equivalent modifications thereof will fall within the scope of the spirit of the present invention. In addition, the system according to the present invention can be embodied as computer readable codes on a computer readable recording medium.

For example, an image encoding apparatus, an image decoding apparatus, a motion vector encoding apparatus, and a motion vector decoding apparatus according to an exemplary embodiment of the present invention may include the apparatuses as shown in FIGS. 1, 2, 4, 5, 9, and 14. A bus coupled to each unit, and at least one processor coupled to the bus may be included. It may also include a memory coupled to the bus to store instructions, received messages or generated messages, coupled to at least one processor for performing the instructions as described above.

In addition, the computer-readable recording medium includes all kinds of recording devices in which data readable by a computer system is stored. Examples of the recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, and optical data storage device. The computer-readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

Claims (2)

In the video decoding method,
Hierarchically dividing the largest coding unit into at least one coding unit according to information indicating whether split is obtained from the bitstream;
Determining a current block included in a current coding unit among the at least one coding unit;
Obtaining information about a prediction direction of the current block indicating one of the L0 direction, the L1 direction, and the bidirectional direction from the bitstream;
Generating a motion vector candidate of the current block based on a motion vector of at least one block previously decoded of the current block; And
According to the information on the prediction direction, at least one motion vector of the current block is used using at least one of a motion vector candidate in the L0 direction and a motion vector candidate in the L1 direction included in the motion vector candidate. Including determining
The at least one block is included in a lower left block of the current block, a block located above the left lower block, and a temporal reference picture located in the L0 direction or the L1 direction, and at least one of blocks having the same position as the current block Including,
When the motion vector candidate of the current block is determined using a motion vector of a block at the same position as the current block included in the temporal reference picture located in the L0 direction or the L1 direction, the temporal reference including the block at the same position Based on a first temporal distance between a picture and a reference picture referenced by the block at the same position and a second temporal distance between a current picture containing the current block and a reference picture referenced by the current block, at the same position By scaling a motion vector of a block, a motion vector candidate of the current block is determined,
The image is divided into the largest coding unit having the largest size, the largest coding unit is hierarchically divided into the at least one coding unit having a depth, and the coding unit of the current depth is split from the coding unit of the higher depth. One of the coding units of the square, the coding unit of the current depth is divided into coding units of the lower depth independently of the neighboring coding unit,
When the coding unit of the current depth is split into at least one prediction unit for prediction decoding, the current block is at least one prediction unit that splits the coding unit of the current depth,
At least one transform unit is determined from a coding unit of the current depth, and an inverse transform is performed on the coding unit using the transform unit. A device for decoding an image,
A maximum coding unit is hierarchically divided into at least one coding unit according to information indicating whether split is obtained from a bitstream, a current block included in a current coding unit among the at least one coding unit is determined, and the A decoder for obtaining information on a prediction direction of the current block indicating one of a L0 direction, a L1 direction, and a bidirectional direction from a bitstream; And
A motion vector candidate of the current block is generated based on a motion vector of at least one block previously decoded in the current block, and a motion vector in the L0 direction included in the motion vector candidate is generated according to information on the prediction direction. And a motion vector determination unit for determining at least one motion vector of the current block using at least one motion vector candidate among a candidate and a motion vector candidate in the L1 direction,
The at least one block is included in a lower left block of the current block, a block located above the left lower block, and a temporal reference picture located in the L0 direction or the L1 direction, and at least one of blocks having the same position as the current block Including,
When the motion vector candidate of the current block is determined using a motion vector of a block at the same position as the current block included in the temporal reference picture located in the L0 direction or the L1 direction, the temporal reference including the block at the same position Based on a first temporal distance between a picture and a reference picture referenced by the block at the same position and a second temporal distance between a current picture containing the current block and a reference picture referenced by the current block, at the same position By scaling a motion vector of a block, a motion vector candidate of the current block is determined,
The image is divided into the largest coding unit having the largest size, the largest coding unit is hierarchically divided into the at least one coding unit having a depth, and the coding unit of the current depth is split from the coding unit of the higher depth. One of the coding units of the square, the coding unit of the current depth is divided into coding units of the lower depth independently of the neighboring coding unit,
When the coding unit of the current depth is split into at least one prediction unit for prediction decoding, the current block is at least one prediction unit that splits the coding unit of the current depth,
An image decoding apparatus characterized in that one or more transform units are determined from the coding units of the current depth, and inverse transform is performed on the coding units using the transform units.

Images