KR101857801B1 - Method and apparatus for encoding/decoding motion vector based on reduced motion vector predictor candidates - Google Patents

Abstract

A method and apparatus for encoding / decoding a motion vector of a current block based on candidates of a predicted motion vector generated by excluding at least one predicted motion vector candidate among all candidates of the predicted motion vector.

Description

[0001] The present invention relates to a method and an apparatus for encoding and decoding a motion vector based on candidates of a reduced predicted motion vector,

The present invention relates to a method and an apparatus for coding and decoding a motion vector, and more particularly, to a method and apparatus for predicting and decoding a motion vector of a current block.

In a codec such as MPEG-4 H.264 / MPEG-4 Advanced Video Coding (AVC), a motion vector of a previously coded block adjacent to a current block is used to predict a motion vector of the current block. The median of the motion vectors of the previously coded blocks adjacent to the upper left, upper and right upper portions of the current block is used as a motion vector predictor of the current block. The motion vector of the current block is not encoded as it is, but only the motion vector difference between the motion vector of the current block and the predicted motion vector is encoded.

According to an aspect of the present invention, there is provided a method and apparatus for predicting and decoding a motion vector, and a computer-readable recording medium having recorded thereon a program for executing the method.

According to another aspect of the present invention, there is provided a method of coding a motion vector of a current block, the method comprising: estimating a motion vector of a current block; Determining a motion vector of the current block as a predicted motion vector of the current block and generating information on the motion vector based on the motion vector of the current block and the predicted motion vector of the current block; Generating a virtual motion vector using the second predicted motion vector candidate and the difference vector of all the candidates, generating difference vectors between the virtual motion vector and each of all the candidates, And selectively excluding the second predicted motion vector candidate from all of the candidates; And encoding information on the motion vector and information on a predicted motion vector of the current block.

According to an aspect of the present invention, there is provided a method of decoding a motion vector of a current block, the method comprising: decoding information on a motion vector of a current block; Generating a hypothetical motion vector using information on a predetermined predicted motion vector candidate and the decoded motion vector among all candidates of the predicted motion vector, generating difference vectors between the hypothetical motion vector and each of all the candidates, Comparing the generated difference vectors with information on the decoded motion vector to selectively exclude the predetermined predicted motion vector from all candidates; And determining one of the candidates of the excluded prediction motion vector as a prediction motion vector of the current block and restoring the motion vector of the current block based on the determined prediction motion vector and information on the decoded motion vector .

According to an aspect of the present invention, there is provided an apparatus for encoding a motion vector, the apparatus comprising: a motion vector estimating unit for estimating a motion vector of a current block; A motion vector estimator for determining a predicted motion vector of the current block and generating information on the motion vector based on the motion vector of the current block and the predicted motion vector of the current block; Generating a virtual motion vector using the second predicted motion vector candidate and the difference vector of all the candidates, generating difference vectors between the virtual motion vector and each of all the candidates, A candidate determining unit for comparing the information on the first predicted motion vector candidate with the information on the second predicted motion vector candidate to selectively exclude the second predicted motion vector candidate from all the candidates; And a motion vector coding unit for coding information on the motion vector and information on a predicted motion vector of the current block.

According to an aspect of the present invention, there is provided an apparatus for decoding a motion vector, comprising: a motion vector decoding unit decoding motion vector information of a current block; Generating a hypothetical motion vector using information on a predetermined predicted motion vector candidate and the decoded motion vector among all candidates of the predicted motion vector, generating difference vectors between the hypothetical motion vector and each of all the candidates, A candidate decision unit for comparing the generated difference vectors with information about the decoded motion vector and selectively excluding the predetermined predicted motion vector from all the candidates; And a motion estimating unit that determines one of the candidates of the excluded prediction motion vector as the predicted motion vector of the current block and restores the motion vector of the current block based on the determined prediction motion vector and information on the decoded motion vector, And a vector restoration unit.

According to an aspect of the present invention, there is provided a computer-readable recording medium having recorded thereon a program for executing a method of coding and / or decoding a motion vector.

According to the present invention, even when the motion vectors are predictively and predictively decoded using the candidates of the predicted motion vector, the number of candidates of the predicted motion vector can be reduced to predict and decode the motion vector. Accordingly, it is possible to encode information necessary for specifying a predicted motion vector candidate used for predicting the motion vector of the current block among the candidates of the predicted motion vector to a minimum bit, thereby improving the compression rate of the motion vector encoding / decoding, The compression rate of image encoding / decoding can also be improved.

1 illustrates an image encoding apparatus according to an embodiment of the present invention.
FIG. 2 illustrates an image decoding apparatus according to an embodiment of the present invention.
FIG. 3 illustrates a hierarchical encoding unit according to an embodiment of the present invention.
FIG. 4 illustrates an image encoding unit based on an encoding unit according to an embodiment of the present invention.
FIG. 5 illustrates an image decoding unit based on an encoding unit according to an embodiment of the present invention.
FIG. 6 illustrates a maximum encoding unit, a sub-encoding unit, and a prediction unit according to an embodiment of the present invention.
7 shows an encoding unit and a conversion unit according to an embodiment of the present invention.
FIGS. 8A and 8B show a division form of an encoding unit, a prediction unit, and a frequency conversion unit according to an embodiment of the present invention.
9 illustrates an apparatus for encoding a motion vector according to an embodiment of the present invention.
10A and 10B illustrate candidates of a predicted motion vector according to an embodiment of the present invention.
Figures 10C-10E illustrate blocks of various sizes adjacent to a current block in accordance with an embodiment of the present invention.
Figures 11A through 11C illustrate candidates of a predicted motion vector according to another embodiment of the present invention.
12 illustrates a method of reducing candidates of a predicted motion vector according to an embodiment of the present invention.
FIGS. 13A to 13D show positions of current blocks included in a coding unit of a predetermined size according to an embodiment of the present invention.
FIG. 14 illustrates an apparatus for decoding a motion vector according to an embodiment of the present invention.
15 is a flowchart illustrating a method of coding a motion vector according to an embodiment of the present invention.
16 is a flowchart for explaining a method of decoding a motion vector 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 illustrates an image encoding apparatus according to an embodiment of the present invention.

1, an image encoding apparatus 100 according to an exemplary embodiment of the present invention includes a maximum encoding unit division unit 110, an encoding depth determination unit 120, an image data encoding unit 130, (140).

The maximum coding unit division unit 110 may divide the current frame or the current slice based on the maximum coding unit which is the coding unit of the maximum size. The current frame or the current slice can be divided into at least one maximum encoding unit.

According to an embodiment of the present invention, an encoding unit can be expressed using a maximum encoding unit and a depth. As described above, the maximum encoding unit indicates the largest encoding unit among the encoding units of the current frame, and the depth indicates the degree to which the encoding units are hierarchically reduced. As the depth increases, the encoding unit can be reduced from the maximum encoding unit to the minimum encoding unit, the depth of the maximum encoding unit can be defined as the minimum depth, and the depth of the minimum encoding unit can be defined as the maximum depth. As the depth of the maximum encoding unit increases, the size of the depth-dependent encoding unit decreases. Therefore, the sub-encoding unit of k depth may include a plurality of sub-encoding units of depth greater than k.

As the size of a frame to be encoded increases, if an image is encoded in a larger unit, an image can be encoded with a higher image compression rate. However, if the coding unit is enlarged and its size is fixed, the image can not be efficiently encoded reflecting the characteristics of the continuously changing image.

For example, when coding a flat area with respect to the sea or sky, the compression rate can be improved by increasing the coding unit. However, when coding a complex area for people or buildings, the compression ratio is improved as the coding unit is decreased.

To this end, one embodiment of the present invention sets a maximum image encoding unit of a different size for each frame or slice, and sets a maximum depth. Since the maximum depth means the maximum number of times the encoding unit can be reduced, the minimum encoding unit size included in the maximum image encoding unit can be variably set according to the maximum depth.

The encoding depth determination unit 120 determines the maximum depth. The maximum depth may be determined based on the Rate-Distortion Cost calculation. The maximum depth may be determined differently for each frame or slice, or may be determined differently for each maximum encoding unit. The determined maximum depth is output to the encoding information encoding unit 140, and the image data for each maximum encoding unit is output to the image data encoding unit 130.

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

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

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

If the size of the encoding unit is 2Nx2N (where N is a positive integer), the processing unit for prediction may be 2Nx2N, 2NxN, Nx2N, NxN, and the like. In other words, motion prediction may be performed based on a processing unit of a type that halves at least one of a height or a width of an encoding unit. Hereinafter, a processing unit serving as a 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 the specific prediction mode may be performed only for a prediction unit of a specific size or type. For example, the intra mode can be performed only for a 2Nx2N, NxN sized prediction unit, which is a square. In addition, the skip mode can be performed only for a prediction unit of 2Nx2N size. If there are a plurality of prediction units in an encoding unit, a prediction mode having the smallest coding error can be selected by performing prediction for each prediction unit.

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

The coding depth determiner 120 may determine sub-coding units included in the maximum coding unit using Rate-Distortion Optimization based on a Lagrangian Multiplier. In other words, it is possible to determine what type of sub-encoding unit the maximum encoding unit is divided into, wherein the plurality of sub-encoding units have different sizes depending on the depth. Then, the image data encoding unit 130 encodes the maximum encoding unit based on the division type determined by the encoding depth determination unit 120, and outputs the bit stream.

The encoding information encoding unit 140 encodes information on the encoding mode of the maximum encoding unit determined by the encoding depth determination unit 120. [ Information on the division type of the maximum encoding unit, information on the maximum depth, and information on the encoding mode of the sub-encoding unit by depth, and outputs the bit stream. The information on the encoding mode of the sub-encoding unit may include information on the prediction unit of the sub-encoding unit, prediction mode information per prediction unit, information on the conversion unit of the sub-encoding unit, and the like.

The information on the division type of the maximum encoding unit may be information indicating whether or not the division is performed for each encoding unit. For example, when the maximum encoding unit is divided and encoded, information indicating whether or not the maximum encoding unit is divided is encoded, and even when the sub-encoding unit generated by dividing the maximum encoding unit is divided and coded again, Information indicating whether or not the sub-encoding unit is divided is encoded. The information indicating whether or not to divide may be flag information indicating whether or not to divide.

There is a sub-encoding unit of a different size for each maximum encoding unit, and information on the encoding mode is determined for each sub-encoding unit, so that information on at least one encoding mode can be determined for one maximum encoding unit.

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

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

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

2, an image decoding apparatus 200 according to an exemplary embodiment of the present invention includes an image data obtaining unit 210, an encoding information extracting unit 220, and an image data decoding unit 230.

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

The encoding information extracting unit 220 parses the bit stream received by the video decoding apparatus 200 and extracts the maximum encoding unit, the maximum depth, the division type of the maximum encoding unit, and the encoding mode of the sub encoding unit from the header for the current frame . The information on the division type and the encoding mode is output to the video data decoding unit 230. [

The information on the division type of the maximum encoding unit may include information on the sub-encoding units of different sizes depending on the depth included in the maximum encoding unit. As described above, the information on the division type may be information (e.g., flag information) indicating whether or not the division is coded for each coding unit. The information on the encoding mode may include information on a prediction unit for each sub-encoding unit, information on a prediction mode, and information on a conversion unit.

The image data decoding unit 230 decodes the image data of each maximum encoding unit based on the information extracted by the encoding information extracting unit, and restores the current frame.

The image data decoding unit 230 can decode the sub-encoding unit included in the maximum encoding unit based on the information on the division type of the maximum encoding unit. The decoding process may include an inter prediction process including intra prediction and motion compensation, and an inverse conversion process.

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

FIG. 3 illustrates a hierarchical encoding 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 a coding unit having a width x height of 64x64. In addition to the square-shaped encoding units, there may be encoding units whose width x height is 64x32, 32x64, 32x16, 16x32, 16x8, 8x16, 8x4, 4x8.

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

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

It is desirable that the maximum size of the encoding size is relatively large in order to accurately reflect not only the compression ratio but also the image characteristic when the resolution is high or the data amount is large. Therefore, the size of the maximum encoding unit of the image data 310 and 320 having a higher resolution than the image data 330 can be selected to be 64x64.

The maximum depth indicates the total number of layers in the hierarchical encoding unit. Since the maximum depth of the image data 310 is 2, the coding unit 315 of the image data 310 is divided into sub-coding units 315 having a major axis size of 32 and 16 as the depth increases, .

On the other hand, since the maximum depth of the image data 330 is 2, the encoding unit 335 of the image data 330 is encoded from the maximum encoding units having the major axis size of 16 to the encoding units having the major axis size of 8 and 4 Units.

Since the maximum depth of the video data 320 is 4, the encoding unit 325 of the video data 320 has a length of 32, 16, 8, and 4 as the depth increases, Sub-encoding units. As the depth increases, the image is encoded based on a smaller sub-encoding unit, which makes it suitable for encoding an image including a finer scene.

FIG. 4 illustrates an image encoding unit based on an encoding unit according to an embodiment of the present invention.

The intra prediction unit 410 performs intra prediction on a prediction unit of the intra mode of the current frame 405 and the motion estimation unit 420 and the motion compensation unit 425 perform intra prediction on the prediction unit of the intra mode, 405 and a reference frame 495. The inter-prediction and the motion compensation are performed using the reference frame 495 and inter-prediction.

The residual values are generated based on the prediction unit output from the intra prediction unit 410, the motion estimation unit 420 and the motion compensation unit 425, and the generated residual values are output to the conversion unit 430 and the quantization unit 440) and output as quantized transform coefficients.

The quantized transform coefficients are restored to a residual value through an inverse quantization unit 460 and an inverse transform unit 470. The reconstructed residual values are passed through a deblocking unit 480 and a loop filtering unit 490, And output to the reference frame 495. The quantized transform coefficient may be output to the bitstream 455 via the entropy encoding unit 450.

In order to perform encoding according to the 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, The quantization unit 440, the entropy encoding unit 450, the inverse quantization unit 460, the inverse transform unit 470, the deblocking unit 480 and the loop filtering unit 490 all have the maximum encoding unit, The prediction unit, and the conversion unit according to the sub-encoding unit, the prediction unit, and the conversion unit.

FIG. 5 illustrates an image decoding unit based on an encoding unit according to an embodiment of the present invention.

The bitstream 505 passes through the parser 510 and the encoded image data to be decoded and the encoding information necessary for decoding are parsed. The encoded image data is outputted as inverse-quantized data through the entropy decoding unit 520 and the inverse quantization unit 530, and is restored to the residual values through the inverse transform unit 540. [ The residual values are added to the intraprediction result of the intraprediction unit 550 or the motion compensation result of the motion compensation unit 560, and restored for each encoding unit. The restored encoding unit is used for predicting the next encoding unit or the next frame through the deblocking unit 570 and the loop filtering unit 580.

A parsing unit 510, an entropy decoding unit 520, an inverse quantization unit 530, and an 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 intraprediction unit 550, the motion compensation unit 560, the deblocking unit 570 and the loop filtering unit 580 are all based on the maximum encoding unit, the depth-dependent sub-encoding unit, the prediction unit, And processes the image decoding processes.

In particular, the intra prediction unit 550 and the motion compensation unit 560 determine a prediction unit and a prediction mode in the sub-coding unit in consideration of the maximum coding unit and the depth, and the inverse transform unit 540 takes the size of the conversion unit into consideration Performs inverse transform.

FIG. 6 illustrates a maximum encoding unit, a sub-encoding 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 a hierarchical encoding unit to perform encoding and decoding in consideration of image characteristics. The maximum encoding unit and the maximum depth may be adaptively set according to the characteristics of the image, or may be variously set according to the demand of the user.

The hierarchical structure 600 of the encoding unit according to an embodiment of the present invention shows a case where the height and the width of the maximum encoding unit 610 are 64 and the maximum depth is 4. The depth increases along the vertical axis of the hierarchical structure 600 of the encoding unit and the width and height of the sub-encoding units 620 to 650 are reduced as the depth increases. Prediction units of the maximum encoding unit 610 and the sub-encoding units 620 to 650 are shown along the horizontal axis of the hierarchical structure 600 of the encoding units.

The maximum encoding unit 610 has a depth of 0, and the size of the encoding unit, i.e., the width and the height, is 64x64. A sub-encoding unit 620 having a depth 1 of 32x32 and a sub-encoding unit 630 having a depth 16x16 having a depth of 16x16, a sub-encoding unit 640 having a depth of 3x8 having a size 8x8, There is a sub-encoding unit 650 of depth 4. A sub-encoding unit 650 of depth 4 with a size of 4x4 is the minimum encoding unit.

Referring to FIG. 6, examples of prediction units along the horizontal axis are shown for each depth. That is, the prediction unit of the maximum coding unit 610 of depth 0 is a prediction unit 610 of size 64x64, a prediction unit 612 of size 64x32, and size 32x64, which are the same or smaller than the coding unit 610 of size 64x64, A prediction unit 614 of size 32x32, and a prediction unit 616 of size 32x32.

The prediction unit of the 32x32 coding unit 620 having the depth 1 has the prediction unit 620 of the size 32x32 which is equal to or smaller than the coding unit 620 of the size 32x32, the prediction unit 622 of the size 32x16, A prediction unit 624 of size 16x16, and a prediction unit 626 of size 16x16.

The prediction unit of the 16x16 encoding unit 630 of depth 2 has a prediction unit 630 of 16x16 size, a prediction unit 632 of size 16x8, and a size of 8x16, which are the same or smaller than the encoding unit 630 of size 16x16. A prediction unit 634 of size 8x8, and a prediction unit 636 of size 8x8.

The prediction unit of the 8x8 encoding unit 640 of depth 3 has a prediction unit 640 of size 8x8, a prediction unit 642 of size 8x4, and a size of 4x8, which are the same or smaller than the encoding unit 640 of size 8x8. A prediction unit 644 of size 4x4, and a prediction unit 646 of size 4x4.

Finally, a coding unit 650 with a depth 4 of size 4x4 is a coding unit with a maximum depth, and a prediction unit is a prediction unit 650 with a size 4x4. However, the encoding unit of the maximum depth does not necessarily have to be the same size as the encoding unit and the prediction unit, and the prediction unit may be divided into prediction units smaller than the encoding unit similarly to the other encoding units 610 to 650 have.

7 shows an encoding unit and a conversion 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 encode the maximum encoding unit as a unit or encode the maximum encoding unit in units of sub encoding units smaller than or equal to the maximum encoding unit. The size of the conversion unit for the conversion during the encoding process can be selected as the size for the highest compression ratio regardless of the encoding unit and the prediction unit. For example, when the current encoding unit 710 is 64x64, the conversion may be performed using the 32x32 conversion unit 720. [

FIGS. 8A and 8B show a division form of an encoding unit, a prediction unit, and a conversion unit according to an embodiment of the present invention.

8A shows an encoding unit and a prediction unit according to an embodiment of the present invention.

The left side of FIG. 8A shows a division type selected by the image encoding apparatus 100 according to an embodiment of the present invention to encode the maximum encoding unit 810. The image encoding apparatus 100 divides the maximum encoding unit 810 into various types, encodes the encoded image, and then selects an optimal division type by comparing the encoding results of various division types based on the R-D cost. When it is optimal to directly encode the maximum encoding unit 810, the maximum encoding unit 800 may be encoded without dividing the maximum encoding unit 810 as shown in FIGS. 8A and 8B.

Referring to the left side of FIG. 8A, a maximum encoding unit 810 having a depth of 0 is divided into sub-encoding units having a depth of 1 or more and encoded. The maximum encoding unit 810 is divided into sub-encoding units of four depths 1, and then all or a part of the sub-encoding units of depth 1 are divided into sub-encoding units of depth 2 again.

Among the sub-coding units of depth 1, sub-coding units coded at the upper right and sub-coding units located at the lower left are divided into sub-coding units of depth 2 or more. Some of the sub-encoding units having depths of 2 or more may be further divided into sub-encoding units having depths of 3 or more.

The right side of FIG. 8B shows the division type of the prediction unit for the maximum coding unit 810.

Referring to the right side of FIG. 8A, the prediction unit 860 for the maximum coding unit can be divided into the maximum coding unit 810 and the prediction unit 860 for the maximum coding unit. In other words, the prediction unit for each of the sub-encoding units may be smaller than the sub-encoding unit.

For example, a prediction unit for a sub-coding unit 854 coded right below the sub-coding unit at depth 1 may be smaller than the sub-coding unit 854. The prediction unit for some of the sub-encoding units 815, 816, 850, and 852 among the sub-encoding units 814, 816, 818, 828, 850, and 852 at the depth 2 may be smaller than the sub-

In addition, the prediction unit for the sub-coding units 822, 832, and 848 of depth 3 may be smaller than the sub-coding unit. The prediction unit may be a form in which each sub-coding unit is halved in the height or width direction, or may be divided into four in height and width directions.

FIG. 8B shows a prediction unit and a conversion unit according to an embodiment of the present invention.

The left side of FIG. 8B shows the division form of the prediction unit for the maximum coding unit 810 shown on the right side of FIG. 8A, and the right side of FIG. 8B shows the division form of the conversion unit of the maximum coding unit 810.

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

For example, even if the prediction unit for the depth 1 encoding unit 854 is selected to be half the height, the conversion unit may be selected to be the same size as the depth unit 854 encoding unit. Likewise, even if the prediction unit for the depth 2 coding units 814 and 850 is selected in such a manner that the height of the depth level 2 coding units 814 and 850 is halved, the level of the level 2 coding units 814 and 850 It can be selected to be the same size as the original size.

The conversion unit may be selected to be smaller than the prediction unit. For example, if the prediction unit for depth 2 coding unit 852 is chosen to be half the width, the conversion unit may be selected to be half the width and height, which is a smaller size than the prediction unit.

9 illustrates an apparatus for encoding a motion vector according to an embodiment of the present invention.

An apparatus for encoding a motion vector included in the image coding apparatus 100 of FIG. 1 or the image coding unit 400 of FIG. 4 is shown in detail in FIG. Referring to FIG. 9, a motion vector coding apparatus 900 according to an embodiment of the present invention includes a motion vector estimation unit 910, a candidate determination unit 920, and a motion vector coding unit 930.

In order to decode a block coded using inter prediction or temporal prediction, information on a motion vector indicating a relative positional difference between a current block and a similar block in a reference picture is needed. Therefore, when the image coding is performed, the information on the motion vector is encoded and inserted into the bitstream. If the information on the motion vector is directly encoded and inserted, an overhead for coding information on the motion vector increases And the compression rate of the video data is lowered.

Accordingly, in the image coding, only the motion vector difference between the motion vector predictor and the original motion vector generated as a result of the prediction is encoded, and the motion vector difference is inserted into the bitstream. Information about motion vectors is also compressed.

The predictive coding of the motion vector may include an explicit mode and an implicit mode.

In a codec such as MPEG-4 H.264 / MPEG-4 Advanced Video Coding (AVC), a motion vector of a previously coded block adjacent to a current block is used to predict a motion vector of the current block. The median of the motion vectors of the previous coded blocks adjacent to the upper left, upper and right upper portions of the current block is used as the predicted motion vector of the current block. Since motion vectors of all blocks encoded using inter prediction are predicted using the same method, information on the predicted motion vector of the current block need not be separately encoded. However, in order to more accurately predict a motion vector, the image encoding apparatus 100 or the image encoding unit 400 according to the present invention may use an implicit mode in which information on the predictive motion vector is not separately encoded, And the explicit mode in which information on the < RTI ID = 0.0 > The explicit mode refers to a mode in which information on a predictive motion vector used as a predictive motion vector of a current block among a plurality of predictive motion vector candidates is encoded and inserted as a sequence parameter, a slice parameter, or a block parameter in the bitstream.

FIG. 9 shows an apparatus for performing predictive coding when a motion vector is encoded according to this explicit mode.

The motion vector estimator 910 estimates a motion vector of the current block. A block similar to or identical to the current block is searched in at least one reference picture and a motion vector which is a relative position difference between the current block and the searched reference block is estimated based on the search result. A block similar to or identical to the current block is searched based on the calculation of the sum of absolute differences (SAD), and the motion vector of the current block can be estimated based on the search result.

In addition, the motion vector estimator 910 predicts a motion vector of a current block based on motion vectors of blocks included in a previously coded area adjacent to the current block. In other words, the motion vectors of the blocks included in the previously coded area adjacent to the current block are set as candidates of the motion vector predictor, and motion vectors of the current block estimated from among the candidates of the predicted motion vector And determines a predicted motion vector candidate most similar to the predicted motion vector candidate.

In a codec such as MPEG-4 H.264 / MPEG-4 Advanced Video Coding (AVC), a median of motion vectors of previously coded blocks adjacent to the left, top, Vector. All the blocks to be coded are predicted using the motion vectors of previously coded blocks, and since only one predicted motion vector is used, information on the predicted motion vector need not be separately coded. In other words, the predicted motion vector of a block coded using inter prediction is one.

However, if the motion vector of the current block is more accurately predicted, the motion vector can be encoded with a higher compression rate. To this end, one embodiment of the present invention selects one of the candidates of the plurality of predicted motion vectors, By using the predicted motion vector as a predicted motion vector, the motion vector of the current block is encoded with a higher compression ratio. Hereinafter, a method of coding a motion vector of a current block using a plurality of predictive motion vector candidates will be described in detail.

10A and 10B illustrate candidates of a predicted motion vector according to an embodiment of the present invention.

Referring to FIG. 10A, in the method of predicting a motion vector according to an embodiment of the present invention, one of motion vectors of previously coded blocks adjacent to a current block can be used as a predicted motion vector of a current block. The motion vectors of the leftmost a0 block, the uppermost block b0 adjacent to the left, the c block adjacent to the upper right, the d block adjacent to the upper left, and the e block adjacent to the lower left, among the blocks adjacent to the upper portion of the current block, And can be used as candidates of the predicted motion vector of the block.

The image encoding method and the decoding method according to the present invention perform image encoding and decoding based on coding units of various sizes classified by depth, and a motion vector of an e block adjacent to the lower left is also used as a prediction motion vector candidate. have.

8A, when the current block is an encoding unit 820, the upper, left upper, right upper, left and lower left encoding units 814, 816, 818, and 822 of the current block are moved to the current block As shown in FIG. Therefore, a motion vector of a block adjacent to the lower left of the current block can also be used as a predicted motion vector candidate.

Referring to FIG. 10B, motion vectors of all adjacent blocks of the current block may be used as candidates of a predicted motion vector. In other words, not only the leftmost a block of the blocks adjacent to the uppermost block but also the motion vectors of all the blocks ao to aN adjacent to the top of the block ao can be used as the predicted motion vector candidates. The motion vectors of all the blocks b0 to bN adjacent to the left side can be used as the predicted motion vector candidates.

In addition, a median value of motion vectors of adjacent blocks can be used as a predicted motion vector candidate. In other words, median (mv_a0, mv_b0, mv_c) can be used as a predicted motion vector candidate of the current block. Here, mv_a0 is a motion vector of the a0 block, mv_b0 is a motion vector of the b0 block, and mv_c is a motion vector of the c block.

However, the candidates of the predicted motion vector of the current block can be limited according to the size of the current block and the size of the adjacent blocks, which will be described in detail with reference to FIGS. 10C to 10E.

Figures 10C-10E illustrate blocks of various sizes adjacent to a current block in accordance with an embodiment of the present invention.

As described above, the image encoding method and the decoding method according to the present invention encode an image using various sizes of encoding units and prediction units determined according to the depth. Therefore, if the size of the current block and the size of some neighboring blocks are greatly different from each other, the motion vectors of some neighboring blocks having different sizes may not be used as the predicted motion vector candidates. have.

Referring to FIG. 10C, blocks 1014 to 1018 adjacent to the top of the current block 1010 are blocks smaller than the size of the current block 1010. The motion vector of the adjacent block 1012 having the same size as the current block 1010 may be the same as or similar to the motion vector of the current block 1010. Therefore, Only the motion vector of the adjacent block 1012 can be used as a predicted motion vector candidate.

Only the motion vectors of adjacent blocks of a predetermined size or more can be used as the predicted motion vector candidates. For example, only the motion vectors of the blocks 1012 and 1018 having a size equal to or larger than 1/4 of the size of the current block 1010 can be used as the predicted motion vector candidates.

Referring to FIG. 10D, the size of the block 1022 adjacent to the left side of the current block 1020 is 16 times that of the current block, and there is a noticeable difference in size. The possibility that the motion vector of the block 1022 adjacent to the left side is the same as or similar to the motion vector of the current block 1020 due to the remarkable difference in size may be low. Therefore, the motion vector of the block 1022 adjacent to the left side is not used as the predicted motion vector candidate of the current block 1020, and only the motion vectors of the block 1024 adjacent to the upper side and the block 1026 adjacent to the upper left side are predicted It is available as a vector candidate.

Referring to FIG. 10E, the size of the current block 1030 is larger than the size of all adjacent blocks 1031 to 1037. In this case, if all of the motion vectors of all the adjacent blocks 1031 to 1037 are used as the predicted motion vector candidates of the current block 1030, the number of candidates of the predicted motion vector of the current block 1030 may be too large. The larger the difference in size between the current block 1030 and the adjacent blocks 1031 to 1037, the more the number of candidates of the predicted motion vector becomes. Therefore, the motion vector estimation unit 910 according to the embodiment of the present invention does not use the motion vectors of some of the adjacent blocks as the predicted motion vector candidates of the current block 1030. [

For example, in the embodiment shown in FIG. 10E, the motion vectors of the block 1031 adjacent to the lower left and the block 1037 adjacent to the upper right may not be used as the predicted motion vector candidates of the current block 1030 .

If the size of the current block 1030 is greater than or equal to a predetermined size, motion vectors of adjacent blocks in a specific direction among adjacent blocks may not be used as the predicted motion vector candidates of the current block 1030. [

Figures 11A through 11C illustrate candidates of a predicted motion vector according to another embodiment of the present invention.

11A illustrates a method of determining a predicted motion vector candidate of a B-directional Predictive Picture according to an embodiment of the present invention. When the current picture including the current block is a B picture performing bidirectional prediction, a motion vector generated based on a temporal distance may be a predicted motion vector candidate.

The predicted motion vector candidate mv_temporal of the current block 1100 of the current picture 1110 can be determined using the motion vector of the block 1120 of the same colocated temporal preceding picture 1112. [ For example, if the motion vector mv_colA of the block 1120 in the same position as the current block 1100 is generated for the retrieved block 1122 of the temporally following picture 1114 of the current picture 1110, 1100), mv_L0A and mv_L1A, which are candidates of the predicted motion vector, can be determined as follows.

mv_L1A = (t1 / t2) * mv_colA

mv_L0A = mv_L1A - mv_colA

Here, mv_L0A denotes a predicted motion vector candidate of the current block 1110 with respect to the temporally preceding picture 1112, mv_L1A denotes a predicted motion vector candidate of the current block 1110 with respect to the temporally preceding picture 1114, .

In the embodiment shown in FIG. 11A, the current picture 1110 as a B picture exists between the temporally preceding picture 1112 and the temporally following picture 1114. At this time, when the motion vector mv_colA of the block 1120 in the same position is generated for the temporally following picture 1114 of the current picture 1110, the motion vector of the current block 1100 is predicted more accurately based on mv_L1A . In other words, when mv_colA is a motion vector opposite to the direction shown in Fig. 11A, that is, when mv_colA is generated for another picture preceding the temporally preceding picture 1112, The motion vector of the current block 1100 can be more accurately predicted.

Therefore, if the direction from the current block 1110 to the block 1120 at the same position is the List0 direction, the motion vector mv_colA of the block 1120 at the same position must be in the direction of List1. Is likely to exist between the preceding picture 1112 and the following picture 1114, so that the motion vector of the current block 1100 can be more accurately predicted based on mv_colA.

In addition, since the pictures 1110 to 1114 shown in FIG. 11A are arranged in time order, a predicted motion vector candidate (mv_temporal) of the current block can be generated based on a POC (Picture Order Count). The predicted motion vector candidate of the current block is generated based on the POC because the picture to which the current block refers may be a picture other than the neighboring pictures 1112 and 1114 shown in FIG. 11A.

For example, if the POC of the current picture is CurrPOC and the POC of the picture referred to by the current picture is CurrRefPOC, the predicted motion vector candidate of the current block can be generated as follows.

Scale = (CurrPOC-CurrRefPOC) / (ColPOC-ColRefPOC)

mv_temporal = Scale * mv_colA

Where ColPOC is the POC of the temporally preceding picture 1112 that contains the block 1120 of the same position and ColRefPOC is the temporal distance of the block 1122 containing the block 1122 referenced by the block 1120 of the same position POC of the picture 1114.

FIG. 11B shows a method of generating a predictive motion vector candidate of a B-directional predictive picture according to another embodiment of the present invention. Compared with the method shown in Fig. 11A, there is a difference in that a temporally following picture 1114 has a block at the same position as the current block 1100. Fig.

11B, the predicted motion vector candidate of the current block 1100 of the current picture 1110 may be generated using the motion vector of the block 1130 of the same position (colocated) of the temporally following picture 1114 have. For example, if the motion vector mv_colB of the block 1130 in the same position as the current block 1100 is generated for the retrieved block 1132 of the temporally preceding picture 1112 of the current picture 1110, 1100), mv_L0B and mv_L1B, which are candidates of the predicted motion vector, can be generated as follows.

mv_L0B = (t3 / t4) * mv_colB

mv_L1B = mv_L0B - mv_colB

Mv_L0B denotes a predicted motion vector of the current block 1110 with respect to the temporally preceding picture 1112 and mv_L1B denotes a predicted motion vector candidate of the current block 1100 with respect to the temporally preceding picture 1114 it means.

11A, the current picture 1110, which is a B picture, exists between the temporally preceding picture 1112 and the temporally following picture 1114 in the embodiment shown in FIG. 11B. Therefore, when the motion vector mv_colB of the block 1130 at the same position is generated for the temporally preceding picture 1112, the motion vector of the current block 1100 can be more accurately predicted based on mv_L0B. In other words, when mv_colB is a motion vector opposite to the direction shown in FIG. 11B, that is, when mv_colB is generated for another picture after the temporally following picture 1114, The motion vector of the current block 1100 can be more accurately predicted.

Therefore, if the direction from the current block 1110 to the block 1130 at the same position is the List 1 direction, the motion vector mv_colB of the block 1130 at the same position must be in the direction of List 0, Is more likely to exist between the preceding picture 1112 and the following picture 1114, it is possible to more accurately predict the motion vector of the current block 1100 based on mv_colB.

In addition, since the picture to which the current block refers may be a picture other than the neighboring pictures 1112 and 1114 shown in FIG. 11B, a predicted motion vector candidate of the current block is generated based on the POC.

For example, if the POC of the current picture is CurrPOC and the POC of the picture referred to by the current picture is CurrRefPOC, the predicted motion vector candidate of the current block can be generated as follows.

Scale = (CurrPOC-CurrRefPOC) / (ColPOC-ColRefPOC)

mv_temporal = Scale * mv_colB

Where ColPOC is the POC of the temporally trailing picture 1114 that contains the block 1130 of the same position and ColRefPOC is the temporal precedence of the block 1132 containing the block 1132 referenced by the block 1130 of the same position POC of the picture 1112.

FIG. 11C shows a predicted motion vector candidate of a P-picture (Predictive Picture) according to an embodiment of the present invention.

11C, the predicted motion vector candidate of the current block 1100 of the current picture 1110 may be determined using the motion vector of the block 1140 of the same colocated temporal preceding picture 1112 . For example, if the motion vector mv_colC of the block 1140 in the same position as the current block 1100 is generated for the retrieved block 1142 of another temporally preceding picture 1116, the prediction of the current block 1100 Mv_L0C, which is a motion vector candidate, can be determined as follows.

mv_L0C = (t6 / t5) * mv_colC

Mv_L0C may be determined based on the POC as described above with respect to Figures 11A and 11B. It is possible to determine mv_L0C based on the POC of the current picture 1110, the POC of the picture referred to by the current picture 1110, the POC of the temporally preceding picture 1112 and the POC of another temporally preceding picture 1116 have.

Since the current picture 1110 is a P-picture, only one candidate of the predicted motion vector of the current block 1100 is determined, unlike Figs. 11A and 11B.

11A and 11B, in order to use the predicted motion vector candidate generated based on the temporal distance to predict the motion vector of the current block, a predicted motion vector candidate is selected using any one of the blocks 1120 and 1130 in the same position Information to be generated must be coded together, and this information may be included in a slice header or a sequence header as a slice parameter or a sequence parameter.

In summary, according to FIGS. 10A and 10B and FIGS. 11A to 11C, the set C of the candidates of the predicted motion vector may be as follows.

Mv_b0, mv_b1, ..., mv_bN, mv_c, mv_d, mv_e, mv_temporal}, mv_a0, mv_a1,

Alternatively, the set C may be a set that reduces the number of candidates of the predicted motion vector.

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

Here, mv_x denotes a motion vector of x block, median () denotes a median value, and mv_temporal denotes predicted motion vector candidates generated using the temporal distance described above with reference to Figs. 11A to 11C. mv_a 'denotes a first motion vector valid among mv_a0, mv_a1, ..., mv_aN. For example, if the a0 block is coded using intraprediction or a different picture from the current block is referenced, mv_a0, which is a motion vector of a0, is not valid, so mv_a '= mv_a1, and a1 block motion vector is also invalid In the case of mv_a '= mv_a2. Likewise, mv_b 'denotes the first motion vector valid among mv_b0, mv_b1, ..., mv_bN, and mv_c' denotes the first motion vector valid among mv_c, mv_d, and mv_e.

A motion vector of a block that refers to a picture different from the current block among the motion vectors of blocks adjacent to the current block can not efficiently predict a motion vector of the current block. Therefore, the motion vector of a block that refers to a different picture from the current block in the set C of predicted motion vector candidates can be excluded.

When the motion vector coding apparatus 900 codes a motion vector according to the explicit mode, information indicating which prediction motion vector candidate in the C set is used for predicting the motion vector of the current block is also encoded. In other words, when the motion vector coding apparatus 900 encodes a motion vector, a binary number corresponding to each of the candidates of the C set, that is, the prediction motion vector, is allocated, and a prediction motion The binary number corresponding to the vector candidate is also encoded.

In order to specify one of the elements of the set C, a binary number corresponding to each of the predicted motion vector candidates is allocated and a binary number is output. Therefore, the smaller the number of elements of the C set, the smaller the number of elements of the C set can do.

Therefore, if there are overlapping predicted motion vector candidates in the C set, the redundant predicted motion vector candidates can be excluded from the C set and assigned a binary number. For example, when C set is C = {median (mv_a ', mv_b', mv_c '), mv_a', mv_b ', mv_c', mv_temporal}, mv_a ', mv_b' and mv_c ' If they are the same, the C set can be determined as two elements, such as C = {mv_a ', mv_temporal}, and binary numbers can be assigned. If the 5 elements of the C set can be specified using 3 bits before excluding the redundant predicted motion vector candidates, after excluding the redundant predicted motion vector candidates, two elements can be specified by using one bit.

Instead of excluding overlapping predicted motion vector candidates, a predetermined weight may be added to increase the probability that the redundant predicted motion vector candidate is determined as the predicted motion vector of the current block. Since mv_a ', mv_b' and mv_c 'are all the same in the above example, only mv_a' is included in the C set, a predetermined weight is added to mv_a 'so that the probability that mv_a' is used to predict the motion vector of the current block is .

If there is one prediction motion vector candidate, it is also possible not to encode a binary number for specifying one of the candidates of the prediction motion vector. For example, the C set is 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_temporal} If the aN block, the b0 to bN block, the c block, the d block, and the e block are all intraprediction blocks, then the C set is C = {mv_temporal}, so that it contains substantially only one element. Therefore, in this case, the motion vector coding apparatus 900 may not encode a binary number for specifying one of the predicted motion vector candidates.

It will be appreciated by those skilled in the art that other motion vectors than all the predicted motion vector candidates may be used as candidates for the predicted motion vector.

According to another embodiment of the present invention, the candidate determining unit 920 may reduce the number of candidates of the predicted motion vector.

As described above, in order to specify a predicted motion vector candidate used for predicting the motion vector of the current block among the candidates of the plurality of predicted motion vectors, additional information is encoded and included in the bitstream. Accordingly, as the number of elements of the C set is smaller, the information required for specifying the predicted motion vector candidates used for predicting the motion vector of the current block in the C set can be encoded with fewer bits. For this, the candidate determining unit 920 may selectively exclude a predetermined predicted motion vector candidate from all candidates of the predicted motion vector using a predetermined evaluation function. Will be described in detail with reference to FIG.

FIG. 12 illustrates a method of reducing candidates of a predicted motion vector according to an embodiment of the present invention.

12, it is assumed that MVP1, MVP2 and MVP3 are elements of the C set, and the motion vector of the current block is MV. Since the predicted motion vector candidate most similar to the motion vector of the current block is used to predict the motion vector of the current block, MVP3, which is most similar to MV, is used to predict the motion vector of the current block.

Accordingly, the difference vector between the motion vector of the current block, which is encoded as information on the motion vector in the motion vector coding apparatus 900, and the predicted motion vector candidate used to predict the motion vector of the current block (hereinafter, (motion vector difference) is (2, 0). Since MV is (5, 0) and MVP3 is (3, 0), the actual motion vector difference is (2, 0).

The candidate determining unit 920 selectively excludes at least one predicted motion vector candidate among all candidates of the predicted motion vector using the actual motion vector difference and the predetermined evaluation function. More specifically, a virtual motion vector is generated using an actual motion vector difference and a predetermined predicted motion vector candidate, motion vector differences between the generated virtual motion vector and all candidates (hereinafter referred to as 'virtual motion vector differences' ) For all candidates. A virtual motion vector is generated by adding an actual motion vector difference and a predetermined predicted motion vector candidate, and a difference between a generated virtual motion vector and a motion vector between all candidates is calculated. It is possible to selectively exclude a predetermined predicted motion vector candidate from all candidates of the predicted motion vector by comparing the actual motion vector difference and the calculated virtual motion vector difference with respect to all of the candidates.

12, the candidate determining unit 920 determines whether MVP1, which is one of the predicted motion vector candidates, is excluded from all candidates.

MVP1 can not be used to predict the motion vector of the current block if the difference between the virtual motion vector generated by subtracting the MVP1-based virtual motion vector and another predicted motion vector candidate is smaller than the actual motion vector difference. For example, if a virtual motion vector difference generated by subtracting MVP3 from a virtual motion vector generated by adding MVP1 and an actual motion vector difference is smaller than an actual motion vector difference, then MVP3 is more accurate than MVP1, And in this case, it is apparent that MVP1 can not be a predicted motion vector.

In Fig. 12, when MVP1 and the actual motion vector difference are added, the virtual motion vector based on MVP1 is (2, 0). Therefore, when a virtual motion vector is generated based on MVP1, the virtual motion vector difference for MVP2 is (2,0), and the virtual motion vector difference for MVP3 is (-1,0). In this case, MVP1 can not be a predicted motion vector of the current block because the magnitude of (-1,0), which is a virtual motion vector difference for MVP3, is smaller than (2,0), which is the magnitude of the actual motion vector difference. Therefore, MVP1 can be excluded from all candidates of the predicted motion vector. In other words, the predicted motion vector candidate corresponding to MVP1 in the above-described C set can be excluded.

In this case, the virtual motion vector difference calculated for MVP1 itself is (2, 0), which is always equal to the actual motion vector difference, and thus can not be smaller than the actual motion vector size. Therefore, when calculating the virtual motion vector difference for each of all candidates of the predicted motion vector, the virtual motion vector difference for MVP1 itself may not be calculated.

When it is determined that MVP1 is excluded, the candidate determining unit 920 determines whether to exclude MVP2 from all candidates of the predicted motion vector. When MVP2 is added to the actual motion vector difference, the virtual motion vector based on MVP2 is (2, 0). Thus, the virtual motion vector difference for MVP1 is (2,0) and the virtual motion vector difference for MVP3 is (-1,0). Since the magnitude of the virtual motion vector difference for MVP3 is smaller than the magnitude of the actual motion vector difference, MVP2, like MVP1, can be removed from all candidates of the predicted motion vector. When determining whether to exclude MVP2, a comparison between the virtual motion vector difference and the actual motion vector difference for MVP1 is optional. It is determined that MVP1 is not a predicted motion vector of the current block. Therefore, it is possible to compare a virtual motion vector difference for each of the remaining candidates except MVP1 with an actual motion vector difference.

Further, the candidate determining unit 920 determines whether the MVP3 is also excluded. A virtual motion vector based on MVP3 is equal to the actual motion vector and even if subtracting another predicted motion vector candidate (i.e., MVP1 or MVP2) from the actual motion vector, a virtual motion vector difference smaller than the magnitude of the actual motion vector difference MVP3 is not excluded from all candidates of the predicted motion vector. According to another embodiment of the present invention, it is apparent that MVP3 determined to be used for predicting the motion vector of the current block is not excluded from all candidates of the predicted motion vector. It is possible to skip the excluded motion vector candidates determined to be used for predicting the motion vector.

That is, the candidate determining unit 920 determines whether or not to exclude the second predicted motion vector, which is one of all candidates of the predicted motion vector, and generates a virtual motion vector by adding the second predicted motion vector and the actual motion vector difference And calculates a difference vector between a virtual motion vector and another predicted motion vector for each of all the candidates to generate a plurality of virtual motion vector differences. If there is at least one virtual motion vector difference that is smaller than the actual motion vector difference among the plurality of virtual motion vector differences, it is apparent that the second predicted motion vector is not the predicted motion vector of the current block. Exclude candidates.

In addition, the candidate determining unit 920 can reduce the total number of candidates of the predicted motion vector, i.e., the number of elements of the C set, by repeating the determination on the excluded or not for each of all the candidates of the predicted motion vector. It is determined whether or not to exclude in order according to the sorting order of the candidates of the total set of predicted motion vectors in the C set. For example, when median (mv_a ', mv_b', mv_c '), mv_a', mv_b ', mv_c', mv_temporal} is determined, When the judgment is finished, it is judged whether or not mv_a 'is excluded. Then, it is determined whether or not mv_b 'is excluded. Repeat the mv_temporal exceptions according to the sort order of the C set.

In the case of repeated judgment, the comparison between the virtual motion vector difference and the actual motion vector difference can be omitted for the candidates excluded in the previous judgment process.

In addition, the C set can be rearranged according to a predetermined criterion as described later with reference to FIGS. 13A to 13D. If the C sets are rearranged, the C sets are repeatedly determined according to the order of rearrangement.

Comparing the magnitudes of the virtual motion vector difference and the actual motion vector difference described above with reference to FIG. 12 can be applied not only to the one-dimensional motion vector but also to the two-dimensional motion vector. In other words, the size of the virtual motion vector difference defined by the x-coordinate and the y-coordinate is compared with the size of the actual motion vector difference to selectively exclude a candidate of a predetermined predicted motion vector from all candidates.

However, the size, which is a criterion for comparison between the hypothetical motion vector difference and the actual motion vector difference, is illustrative, and various criteria can be used for comparing the hypothetical motion vector difference and the actual motion vector difference. Assuming that an evaluation function for generating a value for a virtual motion vector difference and an actual motion vector difference based on a predetermined criterion is 'A', a virtual motion vector difference and an actual motion vector Differences can be compared.

The candidate determining unit 920 determines whether or not to exclude 'mvx', which is one of all candidates of the predicted motion vector, from all candidate candidates of the predicted motion vector, 'mvy' satisfying the expression (1) . In Equation (1), 'MVD' denotes an actual motion vector difference. mvx + MVD-mvy ', which is a virtual motion vector difference between' mvx + MVD ', which is a virtual motion vector based on' mvx ', and' mvy ', which is another predicted motion vector candidate, (Mvx + MVD-mvy), which is a value obtained by evaluating the motion vector A using a predetermined evaluation function 'A', and compares the generated values with 'A (MVD)', which is a value for the actual motion vector difference. Mvy 'candidates other than' mvx 'among all candidates are repeatedly substituted into' mvy 'to determine whether there exists at least one of all candidates' mvy' satisfying the expression (1).

As described above, a virtual motion vector difference and an actual motion vector difference estimated by 'A' can be defined as an x coordinate and a y coordinate. In this case, the evaluation function can be defined as a sum of a value obtained by evaluating the x-coordinate and a value obtained by evaluating the y-coordinate as shown in the following equation (2).

When a virtual motion vector difference or an actual motion vector difference is defined as an x coordinate 'p' and a y coordinate 'q', each coordinate value is substituted by a predetermined function 'f' 'A' can be defined.

According to an embodiment of the present invention, the evaluation function 'A' in Equations (1) and (2) can be an evaluation function for estimating a result of entropy-coding a difference between a result of entropy encoding a virtual motion vector difference and an actual motion vector . The candidate determining unit 920 estimates the result of entropy-encoding the difference between the virtual motion vector and the actual motion vector based on the evaluation function 'A', and reduces the number of candidates of the predicted motion vector based on the estimation result . Will be described in detail with reference to Equation (3).

The function 'f' for estimating the entropy encoding result with respect to the x coordinate value or the y coordinate value can be defined as Equation (3). When an x coordinate value or a y coordinate value 'val' is input to a function 'f' for predicting a result of variable length coding (for example, Universal Variable Length Coding) 'Length' is calculated accordingly.

Equation (3) may be expressed as follows.

The x-coordinate value or the y-coordinate value may be an x-coordinate value or a y-coordinate value of a virtual motion vector difference or an actual motion vector difference.

According to Equation (3), if 'val' is a negative number or '0', 'val' is converted to a positive number and then shifted to the left by one bit to multiply the coordinate value by '2', and '1' '. If 'val' is a positive number, 'val' is shifted to the left by 1 bit, and the coordinate value is multiplied by '2' and stored as 'Temp'. Then repeat the 'while' loop until 'Temp' is '1' to calculate 'Length'.

For example, if the virtual motion vector difference or the actual motion vector difference is (2,0), A (2,0) = f (2) + f (0).

f (2) is calculated as follows. Since '2' of f (2) is a positive number, it shifts to the left by one bit and sets 'Temp' to '4'. In the first while loop, 'Temp' is set to '4', '4' is not '1', 'Temp' is set to '2' by shifting '4' to the right and multiplying by ' do. The initial value of 'Length' is set to '1', so 'Length' is '3' in the first while loop.

In the second while loop, 'Temp' is set to '2', '2' is not '1', and 'Temp' is set to '1' by shifting '2' to the right and multiplying by ' do. Currently 'Length' is '3', so 'Length' in the second while loop is '5'. The third while loop is not performed because 'Temp' is '1', and f (2) is '5'.

f (0) is calculated as follows. Since the input coordinate value of f (0) is '0', '0' is shifted left by one bit, '1' is added, and 'Temp' is set to '1'. Thus, a while loop is not performed. According to the initial value of 'Length', f (0) becomes '1'.

The predetermined evaluation function 'f' described in relation to Equation (3) is a function for estimating the result of entropy encoding using variable length coding. Accordingly, the candidate determining unit 920 estimates the result of variable length length coding of the virtual motion vector differences using the evaluation function 'A' to determine whether to exclude 'mvx' from all candidates of the predicted motion vector. As a result of estimation, if there is at least one hypothetical motion vector difference estimated to be encoded with a shorter length than the actual motion vector difference, 'mvx' is excluded from all candidates of the predicted motion vector.

However, it is easily understood by those skilled in the art that the entropy encoding result can be estimated by a method other than the variable length encoding result. For example, an entropy encoding result of an entropy encoding result and an actual motion vector difference of a virtual motion vector difference may be estimated and compared using another evaluation function 'h', where 'h' is a context adaptive binary arithmetic And may be a function for estimating the result of context adaptive binary arithmetic coding.

According to still another embodiment of the present invention, the result of evaluating the index information to increase the accuracy of the evaluation result based on the predetermined evaluation function can also be estimated together. The index information is information for specifying a predetermined predicted motion vector candidate among all candidates of the predicted motion vector. Will be described in detail with reference to Equation (4).

The candidate determining unit 920 determines whether or not to exclude 'mvx', which is one of all candidates of the predicted motion vector, from all candidate candidates of the predicted motion vector, 'mvy' satisfying the expression (4) . In Equation (4), 'MVD' denotes an actual motion vector difference, and mvxIdx and mvyIdx denote index information for specifying 'mvx' and 'mvy' in all candidates of the predicted motion vector, respectively. mvx + MVD-mvy ', which is a virtual motion vector difference between' mvx + MVD ', which is a virtual motion vector based on' mvx ', and' mvy ', which is another predicted motion vector candidate based on' mvx ' 'And' mvy 'in all candidates using a predetermined evaluation function' A ', and index information for specifying an actual motion vector difference and' mvx 'in all candidates is determined by a predetermined evaluation function' A '. As a result of the evaluation, it is determined whether 'mvy' satisfying the expression (4) is present in at least one of all candidates.

As described above, the virtual motion vector difference and the actual motion vector difference estimated by 'A' can be defined as an x coordinate and a y coordinate, and can be defined as shown in Equation (5).

(Mvx + MVD-mvy, mvyIdx) in Equation (5) is calculated based on the motion vector of the virtual motion vector A (mvx + MVD-mvy) The information for specifying 'mvy' in all candidates of difference and prediction motion vectors is also evaluated. The evaluation function 'A' may be a function for evaluating a result of entropy encoding as described above. In this case, the function 'f' is an x coordinate value of a virtual motion vector difference, is a function for estimating the entropy encoding result based on the y coordinate value, and the function 'g' is a function for estimating the entropy encoding result of 'mvxIdx'. (mvx + MVD-mvy, mvxIdx) can be calculated as shown in Equation (5) when the x coordinate value of 'mvx + MVD-mvy' is 'p1' and the y coordinate value is 'q1'.

A (MVD) on the right side of Equation (2) also evaluates only the actual motion vector difference, but A (MVD, mvxIdx) in Equation 5 indicates information for specifying 'mvx' in all candidates of the actual motion vector difference and the predicted motion vector Are evaluated together. The function 'f' is a function for estimating the entropy encoding result based on the x coordinate value or the y coordinate value of the actual motion vector difference as described above in relation to Equation (2), and the function g is a function for estimating the entropy of mvxIdx And may be a function for estimating the encoding result. When the x coordinate value of 'MVD' is 'p2' and the y coordinate value is 'q2', A (MVD, mvxIdx) can be calculated as shown in Equation (5).

Equation 4 and Equation 5 may be used to determine whether or not to exclude whether or not to exclude the object according to Equation (2). In other words, it is first determined whether 'mvx' is excluded from the candidates of the total predicted motion vector based on Equation (2), and it can be determined whether or not to exclude again according to Equations (4) and (5). For example, if A (mvx + MVD-mvy) and A (MVD) are equal to or greater than A (mvx + MVD-mvy) (MVD) ', then according to equation (2),' mvx 'is not excluded from all candidates of the predicted motion vector. However, even if 'A (mvx + MVD-mvy)' and 'A (MVD)' are the same, 'mvx' can be excluded from all candidates of the predicted motion vector based on the determination result according to Equations have.

When the candidate determining unit 920 determines whether or not to exclude a predicted motion vector candidate based on Equations 1 to 5, it has been described above that it is repeated for all candidates of the predicted motion vector according to the sorting order of the C set. According to another embodiment of the present invention, the candidate determining unit 920 may rearrange the C sets according to a predetermined criterion, and repeat the determination of whether to exclude the C sets according to the rearrangement order. Will be described in detail with reference to Figs. 13A to 13D.

FIGS. 13A to 13D show positions of current blocks included in a coding unit of a predetermined size according to an embodiment of the present invention.

When the candidates of the entire predictive motion vector are equal to C = {median (mv_a ', mv_b', mv_c '), mv_a', mv_b ', mv_c', mv_temporal}, binary numbers are assigned to each of the C set of predicted motion vector candidates The predictive motion vector used to predict the motion vector of the current block among the candidates of the predictive motion vector can be specified.

In this case, a binary number is assigned according to the sort order of the candidates of the predicted motion vector included in the C set, and the binary number may be a variable length code based on the Huffman code. Therefore, a smaller number of bits can be allocated to the predicted motion vector candidate located before the C-th sort order. For example, in the C set, a '0' bit is allocated to 'median (mv_a', mv_b ', mv_c'), a '00' bit is assigned to mv_a and a '01' bit is allocated to mv_b . Accordingly, the candidate determining unit 920 determines the candidates of the predicted motion vector in a predetermined order so that the predicted motion vector candidates likely to be used in predicting the motion vector of the current block among the candidates of the predicted motion vector are located in front of the C set. .

A prediction motion vector highly likely to be used for predicting the motion vector of the current block can be determined according to the position of the current block in the coding unit. 13A, if the current block is positioned at the lower end of the coding unit, the motion vector of the current block is likely to be the same as or similar to the motion vector of the block adjacent to the left of the coding unit or the block adjacent to the lower left. Therefore, it is necessary to change the sorting order so that the candidate of the predicted motion vector corresponding to the motion vector of the block adjacent to the left side or the motion vector of the block adjacent to the lower left side is located in front of the C set. Since the mv_b 'candidates of the predicted motion vector of the set C are candidates of the predicted motion vector corresponding to the motion vector of the block adjacent to the left, the C set is a sequence of mv_b' and median (mv_a ', mv_b', mv_c ' Can be rearranged as C = {mv_b ', mv_a', median (mv_a ', mv_b', mv_c '), mv_c', mv_temporal}.

Likewise, if the current block is located on the left side of the coding unit as shown in FIG. 13B, the motion vector of the block adjacent to the left of the coding unit and the predicted motion vector candidate corresponding to the motion vector of the block adjacent to the upper block are predicted There is a high possibility of being used. Since the mv_b 'of the candidates of the predicted motion vector of the C set is the predicted motion vector candidate corresponding to the motion vector of the block adjacent to the left, the C set includes the order of mv_b' and median (mv_a ', mv_b', mv_c ') Alternatively, they may be rearranged as C = {mv_b ', mv_a', median (mv_a ', mv_b', mv_c '), mv_c', mv_temporal}.

13C, when the current block is located above the coding unit, the motion vector of the block adjacent to the left of the coding unit and the predicted motion vector candidate corresponding to the motion vector of the block adjacent to the upper block are used as the predicted motion vector of the current block There is a high possibility. Since the mv_a 'of the candidates of the predicted motion vector of the C set is the predicted motion vector candidate corresponding to the motion vector of the block adjacent to the upper side, the C set is a sequence of mv_a' and median (mv_a ', mv_b', mv_c ') Alternatively, they can be rearranged as C = {mv_a ', median (mv_a', mv_b ', mv_c'), mv_b ', mv_c', mv_temporal}.

If the current block is located on the right side of the coding unit as shown in FIG. 13D, there is a high possibility that the predicted motion vector candidate corresponding to the motion vector of the block adjacent to the upper right of the coding unit is used to predict the motion vector of the current block. Since the mv_c 'among the candidates of the predicted motion vector of the C set is the predicted motion vector candidate corresponding to the motion vector of the block adjacent to the upper right, the C set is a sequence of mv_c' and median (mv_a ', mv_b', mv_c ' Can be rearranged as C = {mv_c ', mv_a', mv_b ', median (mv_a', mv_b ', mv_c'), mv_temporal}.

The position of the current block in the encoding unit as a reference for rearranging the candidates of the predicted motion vector is exemplary. In other words, various criteria can be used as a reference to rearrange the candidates of the predicted motion vector. Various criteria that allow the prediction motion vector candidates likely to be similar to the motion vector of the current block to be aligned to the beginning of the C set can be used as a reference for rearranging the candidates of the prediction motion vector. It is possible to determine a predicted motion vector candidate that is likely to be similar to a motion vector of the current block based on predetermined information related to other blocks encoded before the current block, and rearrange the C sets based on the determination.

In addition, a motion vector candidate that is likely to be similar to the motion vector of the current block is determined based on other information encoded or decoded on the current block before coding the motion vector of the current block, and based on the determination, Can be rearranged.

Also, the reordering can be performed except for the redundant predicted motion vector candidates at the reordering of the C set. If there are overlapping predicted motion vector candidates in all candidates of the predicted motion vector, it is possible to exclude the predicted motion vector candidates and determine whether to exclude each predicted motion vector candidate according to Equations 1 to 5 above.

Referring again to FIG. 9, the motion vector coding unit 930 encodes information on a motion vector and information on a predicted motion vector. The information about the motion vector is a difference vector between the actual motion vector of the current block and the actual predicted motion vector, and the information about the predicted motion vector is the motion vector of the current block in the candidates excluding at least one predicted motion vector from all candidates of the predicted motion vector. And information for specifying a predicted motion vector used in the motion vector prediction of the current frame. In other words, information for specifying a predictive motion vector of the current block in the candidates of the predictive motion vector not excluded in the candidate decision unit 920 is encoded as information on the predictive motion vector.

The motion vector difference is received from the motion vector estimation unit 910 and is encoded according to a predetermined entropy coding method, and candidates of the predicted motion vector determined by selectively excluding at least one predicted motion vector candidate in the candidate determination unit 920 Information for specifying a predicted motion vector candidate used for predicting the motion vector of the current block determined by the motion vector estimation unit 910. [

When the candidate determining unit 920 determines candidates of the predicted motion vector by excluding candidates of at least one predicted motion vector from all candidates of the predicted motion vector according to Equations 1 to 5 described above, And information for specifying a predicted motion vector candidate used for predicting the motion vector of the current block is encoded. The motion vector coding unit 930 may index each of the candidates of the predicted motion vector not excluded in the candidate decision unit 920 and entropy-encode the index information as information on the predicted motion vector. Indexing means assigning a predetermined binary number to each of the candidates of the predicted motion vector and information on the predicted motion vector means information used to predict a motion vector of the current block among the candidates of the predicted motion vector. As a result of the candidate decision unit 920 selectively excluding at least one predicted motion vector candidate, there is no need to separately encode the information about the predicted motion vector in the motion vector coding unit 930 . The predicted motion vector candidate to be used for predicting the motion vector of the current block is implicitly determined.

13A to 13D, the candidate determining unit 920 rearranges the candidates of the entire predicted motion vector according to a predetermined criterion, and predicts at least one predicted motion vector And indexes each candidate of the generated predicted motion vector and entropy-encodes the index information.

As a result of rearrangement of the candidate decision unit 920, since the binary number of the smallest number of bits is allocated to the prediction motion vector candidates likely to be used in predicting the motion vector of the current block, information on the predicted motion vector is encoded with a higher compression rate can do.

FIG. 14 illustrates an apparatus for decoding a motion vector according to an embodiment of the present invention.

An apparatus for decoding a motion vector included in the image decoding apparatus 200 of FIG. 2 or the image encoding unit 500 of FIG. 5 is shown in detail in FIG. Referring to FIG. 14, a motion vector decoding apparatus 1400 according to an embodiment of the present invention includes a motion vector decoding unit 1410, a candidate determination unit 1420, and a motion vector restoration unit 1430.

The motion vector decoding apparatus 1400 of FIG. 14 shows an apparatus for decoding a motion vector of a current block when a motion vector of the current block is coded according to the explicit mode among the explicit mode and the implicit mode described above.

The motion vector decoding unit 1410 receives the bitstream of the motion vector of the current block and decodes the received bitstream. And decodes information on a motion vector included in the bitstream. And decodes the actual motion vector difference of the current block. The actual motion vector difference can be decoded according to a predetermined entropy decoding method. The actual motion vector difference is a difference vector between the motion vector of the current block and the predicted motion vector candidate used to predict the motion vector of the current block.

According to the method of coding a motion vector of the present invention, candidates of a predicted motion vector are determined by excluding at least one predicted motion vector candidate among all candidates of the predicted motion vector according to the above-described Equations 1 to 5. The candidates of the predicted motion vector are not fixed, but can be continuously changed as the decoding proceeds on a block-by-block basis. Therefore, even if the information on the candidate of the predicted motion vector is the same, if the predicted motion vector candidates are not determined, the predicted motion vector candidate used for predicting the motion vector of the current block can not be accurately reconstructed.

Therefore, the candidate determining unit 1420 determines the candidates of the predicted motion vector before determining the predicted motion vector candidate. And selects candidates of the predicted motion vector by selectively excluding at least one predicted motion vector candidate among all candidates of the motion vector according to Equations (1) to (5). Based on a predetermined evaluation function, excludes a predicted motion vector candidate that is not used to predict a motion vector of the current block among all candidates determined based on motion vectors of blocks included in a previously decoded area adjacent to the current block do.

A virtual motion vector is generated based on information on a motion vector predicted by a motion vector decoding unit and a motion vector predicted by a predetermined motion vector candidate among all candidates of the predicted motion vector, A difference of a virtual motion vector, which is a difference, is calculated for each of all candidates. The calculated virtual motion vector differences are compared with the information on the motion vector decoded by the motion vector decoding unit 1410, that is, the difference between the actual motion vectors, thereby excluding candidates of the predicted motion vectors. The entropy encoding result of the virtual motion vector differences may be compared with the entropy encoding result of the actual motion vector difference to determine whether to exclude the candidate of the predetermined predicted motion vector. In order to increase the estimation accuracy of the entropy encoding result, the result of entropy encoding the index information can also be estimated and used to determine whether or not to exclude the entropy encoding result. The method of excluding candidates of the predicted motion vector has been described above with reference to Equations (1) to (5).

According to another embodiment of the present invention, the candidate determination unit 1420 rearranges the entire candidates of the predicted motion vector according to a predetermined criterion, and performs the exclusion determination according to Equations 1 to 5 on all the reordered candidates It is possible to selectively exclude at least one predictive motion vector candidate repeatedly. It is possible to repeat the exclusion determination according to Equations 1 to 5 except for the redundant predicted motion vector candidates in all the reordered candidates.

If the candidate decision unit 1420 has left a plurality of predicted motion vector candidates among all the candidates of the predicted motion vector as a result of excluding at least one predicted motion vector candidate among all candidates of the predicted motion vector, the motion vector decoding unit 1410 decides And decodes the information on the motion vector. And decodes information on the predicted motion vector according to a predetermined entropy decoding method. The information on the predictive motion vector is information for specifying a predictive motion vector candidate used to predict a motion vector of a current block among the candidates of the predictive motion vector from which at least one predictive motion vector candidate is excluded. Information for specifying a predicted motion vector candidate used for predicting the motion vector of the current block among the candidates of the predicted motion vector not excluded in the candidate decision unit 1420 is decoded.

If the candidate decision unit 1420 has left only one predicted motion vector candidate as a result of excluding at least one predicted motion vector candidate among all candidates of the predicted motion vector, the remaining predicted motion vector candidate predicts the motion vector of the current block The motion vector decoding unit 1410 does not need to separately decode the information on the candidate of the predicted motion vector.

The motion vector restoring unit 1430 restores the motion vector of the current block based on the information on the motion vector decoded by the motion vector decoding unit 1410. The motion vector decoding unit 1410 adds the motion vector difference decoded by the motion vector decoding unit 1410 and the predicted motion vector candidate used to predict the motion vector of the current block to recover the motion vector of the current block. A prediction motion vector candidate to be used for predicting the motion vector of the current block among the candidates of the prediction motion vector determined in the candidate determination unit 1420 is determined and the determined predicted motion vector candidate is added to the actual motion vector difference. In the case where candidates of a plurality of predictive motion vectors other than one are left as the result of the exclusion in the candidate decision unit 1420, the predicted motion vector candidate used for predicting the motion vector of the current block is decoded in the motion vector decoding unit 1410 May be determined based on information on the predicted motion vector.

Since the candidates of the predicted motion vector are determined by the candidate determining unit 1420, even if the information on the decoded predicted motion vector is the same, the predicted motion vector candidate used for predicting the motion vector of the current block is May be a motion vector.

15 is a flowchart illustrating a method of coding a motion vector according to an embodiment of the present invention.

Referring to FIG. 15, in step 1510, the motion vector encoding apparatus estimates a motion vector of a current block and determines a predicted motion vector candidate used for predicting a motion vector of the current block, from all candidates of the predicted motion vector. A motion vector which is the same or similar to the current block is searched in a plurality of reference pictures and a motion vector which is a relative position difference between the current block and the reference block is estimated according to the search result.

Then, the motion vector of the current block is predicted based on the motion vectors of the blocks included in the previously encoded area adjacent to the current block. In other words, the motion vectors of the blocks included in the previously coded area adjacent to the current block are set as all candidates of the predicted motion vector, and a predicted motion vector which is the closest to the motion vector of the estimated current block among all candidates of the predicted motion vector The candidate is decided. A difference vector between the motion vector of the current block and the determined predicted motion vector candidate, that is, an actual motion vector difference, is generated.

In step 1520, the image encoding apparatus selectively excludes at least one predicted motion vector candidate from all candidates of the predicted motion vector. Excludes predicted motion vector candidates that are unlikely to be used to predict the motion vector of the current block among all candidates of the predicted motion vector.

The image encoding apparatus generates a virtual motion vector using a predetermined prediction motion vector candidate among all candidates of the predicted motion vector and an actual motion vector difference generated in step 1510. The generated virtual motion vector and another predicted motion vector candidate To generate a virtual motion vector difference. A virtual motion vector difference is generated for each of all the candidates, and a predetermined predicted motion vector candidate is selectively excluded by comparing the generated virtual motion difference and the actual motion vector difference.

The process of generating and optionally excluding the virtual motion vector of step 1520 may be repeated for all of the candidates so that at least one predicted motion vector candidate may be excluded from all candidates. The motion vector difference may be calculated for each of the remaining predicted motion vector candidates obtained by subtracting the excluded prediction motion vector candidates from each other and the calculated virtual motion vector difference may be compared with the actual motion vector difference. have.

The virtual motion vector difference and the actual motion vector difference can be evaluated and compared based on a predetermined evaluation function, and the predetermined evaluation function can be a function for predicting the entropy encoding result. It is possible to compare the result obtained by entropy encoding a virtual motion vector difference and the function of estimating a result of entropy encoding an actual motion vector difference. In addition, the entropy-encoded index information may be estimated together with the index information to improve the accuracy of the evaluation. The method of excluding at least one predicted motion vector candidate in all candidates of the predicted motion vector has been described above with reference to Equations 1 to 5. [

In addition, the motion vector coding apparatus may rearrange the entire candidates of the motion vector according to a predetermined criterion, as described above with reference to FIGS. 13A to 13D, and selectively exclude at least one predictive motion vector candidate from all the reordered candidates have. It is possible to repeat the exclusion determination according to Equations 1 to 5 except for the redundant predicted motion vector candidates in all the reordered candidates.

In step 1530, the motion vector coding apparatus encodes information on a motion vector and information on a predicted motion vector. Information for specifying an actual motion vector difference and a predicted motion vector candidate used to predict a motion vector of the current block is encoded. The information on the predicted motion vector may be information for specifying a predicted motion vector candidate used in predicting the motion vector of the current block in the candidates of the predicted motion vector not excluded through steps 1520 and 1530. [

If only one predicted motion vector candidate is left as a result of excluding at least one predicted motion vector candidate in all candidates of the predicted motion vector, information on the predicted motion vector may not be encoded.

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

Referring to FIG. 16, in step 1610, the motion vector decoding apparatus decodes information on the motion vector of the current block from the received bitstream. The information on the motion vector may be the difference between the actual motion vector of the current block and the predicted motion vector of the current block.

In step 1620, the motion vector decoding apparatus generates information on the motion vector decoded in step 1610 and a virtual motion vector based on the predicted motion vector candidate of all the candidates of the predicted motion vector.

When a virtual motion vector is generated, the motion vector decoding apparatus excludes at least one predicted motion vector candidate from all candidates of the predicted motion vector. All candidates of the predicted motion vector are determined based on the motion vectors of the blocks of the previously decoded area adjacent to the current block. The motion vector decoding apparatus may selectively exclude at least one predicted motion vector candidate from all candidates of the predicted motion vector. The virtual motion vector difference based on the predetermined evaluation function and the actual motion vector difference decoded in step 1610 are evaluated to selectively exclude a predetermined predicted motion vector candidate. The method of excluding the predicted motion vector candidate from all candidates is the same as step 1530, and has been described above with reference to equations (1) to (5).

The process of generating the virtual motion vector of step 1620 and the process of selectively excluding the virtual motion vector may be repeatedly performed for all of the candidates so that at least one predicted motion vector candidate may be excluded from all candidates.

In addition, the motion vector coding apparatus may rearrange the entire candidates of the motion vector according to a predetermined criterion, as described above with reference to FIGS. 13A to 13D, and selectively exclude at least one predictive motion vector candidate from all the reordered candidates have. It is possible to repeat the exclusion determination according to Equations 1 to 5 except for the redundant predicted motion vector candidates in all the reordered candidates.

As a result of the exclusion, if the candidates of the plurality of predicted motion vectors remain, information on the predicted motion vector is decoded. If only one predicted motion vector candidate is left, information on the predicted motion vector is not decoded.

In step 1630, the motion vector decoding apparatus determines a predicted motion vector candidate used for predicting the motion vector of the current block among the candidates of the predicted motion vector not excluded in step 1620.

A prediction motion vector candidate used for predicting the motion vector of the current block among the candidates of the predicted motion vector may be determined based on information on the predicted motion vector of the current block. If only one predicted motion vector candidate is left as a result of step 1620, the remaining predicted motion vector candidate is determined as a predicted motion vector candidate used to predict a motion vector of the current block.

When the predicted motion vector candidate is determined, the motion vector of the current block is restored by adding the determined predicted motion vector candidate and the difference of the actual motion vector decoded in step 1610. [

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, Modification is possible. Accordingly, the spirit of the present invention should be understood only in accordance with the following claims, and all of the equivalent or equivalent variations will fall within the scope 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, the image encoding apparatus, the image decoding apparatus, the motion vector encoding apparatus, and the motion vector decoding apparatus according to exemplary embodiments of the present invention may be applied to a device as shown in FIGS. 1, 2, 4, 5, 9 and 14 A bus coupled to each of the units, and at least one processor coupled to the bus. It may also include a memory coupled to the bus for storing instructions, received messages or generated messages and coupled to the at least one processor for performing the instructions as described above.

In addition, the computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer system is stored. Examples of the recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like. The computer-readable recording medium may also be distributed over a networked computer system so that computer readable code can be stored and executed in a distributed manner.

Claims (9)

A method for decoding an image,
Decoding information on a motion vector difference of a current block and information on a prediction motion vector for specifying a prediction motion vector of the current block from a bitstream;
Constructing a predicted motion vector candidate set;
Adjusting the predicted motion vector candidate set based on the values of the predicted motion vector candidates in the predicted motion vector candidate set;
Determining a predicted motion vector of the current block based on the adjusted predicted motion vector candidate set and information on the predicted motion vector; And
And determining a motion vector of the current block based on the predicted motion vector of the current block and the motion vector difference of the current block,
The adjusted predicted motion vector candidate set includes first-order predicted motion vector candidates based on motion vectors of blocks adjacent to the current block, second-order predicted motion vector candidates based on motion vectors of blocks on the reference picture at the same position as the current block, , ≪ / RTI >
The adjacent blocks include a first block on the upper right side of the current block, a second block on the left side of the first block, a third block on the upper left side of the current block, and a fourth block on the lower left side of the current block The decoding method
The method according to claim 1,
Wherein the step of adjusting the predicted motion vector candidate set comprises: if there are first-order predicted motion vector candidates having overlapping values in the predicted motion vector candidate set, selecting one of the first- And removing from the predicted motion vector candidate set.
The method according to claim 1,
Wherein the adjusted predicted motion vector candidate set includes a total of two predicted motion vector candidates,
Wherein the information on the predictive motion vector represents one of the two predictive motion vector candidates through a 1-bit binary value.
A decoding unit decoding information on a motion vector difference of a current block and information on a prediction motion vector for specifying a prediction motion vector of the current block from a bitstream;
A candidate decision unit for constructing a predicted motion vector candidate set and adjusting the predicted motion vector candidate set based on the value of the predicted motion vector candidates in the predicted motion vector candidate set; And
Determining a predicted motion vector of the current block based on the adjusted predicted motion vector candidate set and information about the predicted motion vector, and calculating information on the predicted motion vector of the current block and the motion vector difference of the current block, And a motion vector restoring unit for determining a motion vector of the current block,
The adjusted predicted motion vector candidate set includes first-order predicted motion vector candidates based on motion vectors of blocks adjacent to the current block, second-order predicted motion vector candidates based on motion vectors of blocks on the reference picture at the same position as the current block, , ≪ / RTI >
The adjacent blocks include a first block on the upper right side of the current block, a second block on the left side of the first block, a third block on the upper left side of the current block, and a fourth block on the lower left side of the current block And decrypting the encrypted data.
5. The method of claim 4,
In the case where the candidate motion vector candidates having overlapping values are present in the motion vector candidate set of the predictive motion vector, the candidate decision unit determines the candidate motion vectors of the primary predicted motion vector candidates From the predicted motion vector candidate set. ≪ Desc / Clms Page number 19 >
5. The method of claim 4,
Wherein the candidate determining unit adjusts the predicted motion vector candidate set so as to include a total of two predicted motion vector candidates,
Wherein the candidate determining unit determines one of the two predicted motion vector candidates through information on the predictive motion vector having a 1-bit binary value. A method of encoding an image,
Performing inter-prediction on a current block to determine a motion vector of the current block;
Constructing a predicted motion vector candidate set;
Adjusting the predicted motion vector candidate set based on the values of the predicted motion vector candidates in the predicted motion vector candidate set;
A predictive motion vector of the current block among the predicted motion vector candidates included in the adjusted predictive motion vector candidate set is determined based on the motion vector of the current block to determine a predictive motion vector for specifying the predictive motion vector of the current block Encoding the information about the data;
Encoding information on a difference between a motion vector of the current block and a motion vector of the current block indicating a difference between the motion vector of the current block and the predicted motion vector of the current block; And
Generating a bitstream including information on a motion vector difference of the current block and information on a predictive motion vector for specifying a predictive motion vector of the current block,
The adjusted predicted motion vector candidate set includes first-order predicted motion vector candidates based on motion vectors of blocks adjacent to the current block, second-order predicted motion vector candidates based on motion vectors of blocks on the reference picture at the same position as the current block, , ≪ / RTI >
The adjacent blocks include a first block on the upper right side of the current block, a second block on the left side of the first block, a third block on the upper left side of the current block, and a fourth block on the lower left side of the current block Encoding method < RTI ID = 0.0 >
An apparatus for encoding an image, the apparatus comprising:
A motion vector estimator for performing motion vector prediction on the current block to determine a motion vector of the current block;
A candidate decision unit for constructing a predicted motion vector candidate set and adjusting the predicted motion vector candidate set based on the value of the predicted motion vector candidates in the predicted motion vector candidate set; And
A predictive motion vector of the current block among the predicted motion vector candidates included in the adjusted predictive motion vector candidate set is determined based on the motion vector of the current block to determine a predictive motion vector for specifying the predictive motion vector of the current block And encodes information on the difference between the motion vector of the current block and the motion vector of the current block indicating the difference between the motion vector of the current block and the predicted motion vector of the current block, And a motion vector encoding unit for generating a bitstream including information on a predictive motion vector for specifying a predictive motion vector of the current block,
The adjusted predicted motion vector candidate set includes first-order predicted motion vector candidates based on motion vectors of blocks adjacent to the current block, second-order predicted motion vector candidates based on motion vectors of blocks on the reference picture at the same position as the current block, , ≪ / RTI >
The adjacent blocks include a first block on the upper right side of the current block, a second block on the left side of the first block, a third block on the upper left side of the current block, and a fourth block on the lower left side of the current block .
Information on a predicted motion vector for specifying a predicted motion vector of a current block; And
And information on a difference between a motion vector of a current block and a motion vector of a current block indicating a difference between the predicted motion vectors of the current block,
The motion vector of the current block is determined by performing inter-prediction on the current block,
A predicted motion vector candidate set is constructed, the predicted motion vector candidate set is adjusted based on the value of the predicted motion vector candidates in the predicted motion vector candidate set,
Wherein the information on the predictive motion vector is information determined by determining one of predictive motion vector candidates included in the adjusted predictive motion vector candidate set based on the motion vector of the current block,
The adjusted predicted motion vector candidate set includes first-order predicted motion vector candidates based on motion vectors of blocks adjacent to the current block, second-order predicted motion vector candidates based on motion vectors of blocks on the reference picture at the same position as the current block, , ≪ / RTI >
The adjacent blocks include a first block on the upper right side of the current block, a second block on the left side of the first block, a third block on the upper left side of the current block, and a fourth block on the lower left side of the current block A computer-readable recording medium storing a bit stream.

Images