KR101601813B1 - A Video Decoding Method and Apparatus Using Inter Prediction - Google Patents

Abstract

The present invention relates to a method of decoding an image using inter-prediction, comprising: reconstructing a first differential motion vector and a second differential motion vector of a current block located in a current picture by decoding encoded data; Deriving a first predicted motion vector and a second predicted motion vector of the current block from one or more neighboring blocks of the current block; A second motion vector prediction unit for generating a first motion vector of the current block by adding the first difference motion vector and the first predicted motion vector to each other and adding the second difference motion vector and the second predicted motion vector, Generating a motion vector; Generating a prediction block of the current block using a first motion vector and a second motion vector of the current block; Decoding the residual signal included in the encoded data and selectively performing an inverse quantization process and an inverse transform process on the decoded residual signal to recover a residual block; And adding each pixel of the prediction block and a corresponding pixel of the residual block.

Description

[0001] The present invention relates to a video decoding method and apparatus using inter prediction,

The present invention relates to a video decoding method and apparatus using inter prediction.

[0002] Recently, multimedia technology has been rapidly developed, and accordingly, demand for high-quality multimedia data including audio, image, and moving picture is increasing. As part of this trend, an international standard for high-efficiency image compression has been established to meet the demand to transmit, store, and retrieve multimedia data in a restricted network environment. In particular, the H.264 / AVC MPEG-4 Part.10 standard, established by ISO / IEC JTC1 / SC29 MPEG group and ITU-T VCEG group as an international standard for video compression, Block prediction, motion estimation and compensation for block size, intra prediction, and the like.

Predictive coding is widely used for compressing various data as an effective way to reduce correlation between data. In particular, since the list0 motion vector and the list1 motion vector, which are motion vectors (MVs) of the current block for the two reference pictures in the B picture, are closely correlated with the motion vectors of the neighboring blocks, A prediction value (PMV) for a motion vector of a block is calculated and a list value of a list0 motion vector and a list1 motion vector of the current block are not encoded but a difference value ( DMV (Differential Motion Vector, hereinafter referred to as 'differential motion vector'), the amount of bits to be encoded can be significantly reduced and the coding efficiency can be increased accordingly.

Generally, in motion vector coding using such a predictive motion vector, the coding efficiency can be increased as the predicted motion vector is more accurate for effective compression. Therefore, a finite number of predicted motion vectors, which are not only motion vectors of spatially adjacent blocks but also motion vectors of temporally, spatially or temporally adjacent blocks, or other motion vectors calculated by combining them, are generated, The coding efficiency can be further increased by selecting a most suitable motion vector for the prediction coding from the motion vectors.

In this case, in order to correctly recover the original motion vector from the information on the encoded motion vector, it is necessary to know which prediction motion vector is used among a finite number of predicted motion vectors. The most simple motion vector prediction coding method for this purpose is to encode information about which prediction motion vector is used for motion vector prediction coding. The current H.264 / AVC standard uses the list0 and list1 motions of the neighboring blocks (left, top, right top) to reduce the amount of bits that occur to encode additional information to indicate which prediction mode The median of each of the horizontal component and the vertical component of the vectors is used as a predictive motion vector for predictive coding of the motion vector. In this method, a predetermined predictive value calculation method known together in the image encoding apparatus and the image decoding apparatus such as the intermediate value calculation is determined, and a predictive motion vector is calculated using a predetermined predictive value calculation method in the image encoding apparatus or the image decoding apparatus , And eliminates the need to encode information about which prediction motion vector is used together. An existing method of predefining and using a predetermined prediction value calculation method has an advantage that it is possible to maintain an improved coding efficiency without transmitting additional information as to which motion vector is used as a predicted motion vector. There is a problem that the motion vector is not an optimal predicted motion vector that always generates the minimum amount of bits required for coding the differential vector.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and it is an object of the present invention to improve the compression efficiency by reducing the amount of bits required for coding motion vectors and residual signals in image coding and decoding using inter prediction.

According to an aspect of the present invention, there is provided a method of decoding an image using inter prediction, the method comprising: decoding a coded data to restore a first differential motion vector and a second differential motion vector of a current block located in a current picture ; Deriving a first predicted motion vector and a second predicted motion vector of the current block from one or more neighboring blocks of the current block; A second motion vector prediction unit for generating a first motion vector of the current block by adding the first difference motion vector and the first predicted motion vector to each other and adding the second difference motion vector and the second predicted motion vector, Generating a motion vector; Generating a prediction block of the current block using a first motion vector and a second motion vector of the current block; Decoding the residual signal included in the encoded data and selectively performing an inverse quantization process and an inverse transform process on the decoded residual signal to recover a residual block; And adding each pixel of the prediction block and a corresponding pixel of the residual block.

The step of reconstructing the residual block comprises: identifying one or more conditions for inverse quantization and inverse transform; And skipping the inverse quantization and inverse transform of the decoded residual signal based on the identified condition, dequantizing the decoded residual signal and skipping the inverse transform of the dequantized residual signal, And performing the inverse quantization and inverse transform on the residual residual signal, thereby reconstructing the residual block.

The step of deriving the first and second predicted motion vectors may include deriving a first candidate motion vector set and a second candidate motion vector set using the motion vectors of the at least one neighboring block. Selecting a first candidate motion vector in the first set of candidate motion vectors and a second candidate motion vector in a second set of backward motion vectors; And setting the first candidate motion vector to the first predictive motion vector and the second candidate motion vector to the second predictive motion vector.

Wherein the one or more neighboring blocks include one or more blocks spatially adjacent to the current block in the current picture and one or more blocks located in a reference picture decoded prior to the current picture, And may be determined based on the position of the current block in the current picture.

The step of deriving the first set of candidate motion vectors may include the step of including only one of the two candidate motion vectors derived from the motion vectors of the at least one neighboring block in the first candidate motion vector set Wherein the step of deriving the second set of candidate motion vectors comprises the step of including only one of the two candidate motion vectors derived from the motion vector of the at least one neighboring block in the second candidate motion vector set . ≪ / RTI >

The first candidate motion vector and the second candidate motion vector may be selected based on information included in the encoded data.

In addition, one or more conditions for the dequantization and inverse transformation may be identified based on information included in the encoded data.

According to another aspect of the present invention, there is provided an apparatus and method for decoding an image using inter prediction, the apparatus comprising: an information extracting unit that decodes encoded data to reconstruct a first difference motion vector and a second difference motion vector of a current block located in a current picture, part; A first predictive motion vector and a second predictive motion vector of the current block are derived from one or more neighboring blocks of the current block, and the first differential motion vector and the first predictive motion vector are added, Generating a second motion vector of the current block by adding the second difference motion vector and the second predicted motion vector to generate a motion vector, and generating a second motion vector of the current block using the first motion vector and the second motion vector of the current block, A prediction unit for generating a prediction block of a current block; A decoding unit decoding a residual signal included in the encoded data and selectively performing an inverse quantization process and an inverse transform process on the decoded residual signal to recover a residual block; And an adder for adding each pixel of the prediction block and a corresponding pixel of the residual block.

The decoding unit identifies one or more conditions for inverse quantization and inverse transform, and skips inverse quantization and inverse transform of the decoded residual signal based on the identified condition, dequantizes the decoded residual signal A process of skipping the inverse transform of the inverse quantized residual signal, and a process of performing both the inverse quantization and inverse transform of the decoded residual signal, thereby restoring the residual block.

As described above, according to the present invention, the prediction motion vector of the current motion vector for a plurality of reference pictures can be more accurately predicted, and the amount of bits required to encode the motion vector can be reduced to improve the compression efficiency.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram illustrating a block for coding a current motion vector of a list0 according to an embodiment of the present invention;
FIG. 2 is a block diagram illustrating a block for encoding a current motion vector list1 according to an embodiment of the present invention. FIG.
3 is a block diagram schematically illustrating a configuration of a motion vector coding apparatus according to an embodiment of the present invention.
4 is a flowchart for explaining a motion vector coding method according to an embodiment of the present invention.
FIG. 5 is a block diagram schematically showing a configuration of a motion vector decoding apparatus according to an embodiment of the present invention;
FIG. 6 is a flowchart for explaining a motion vector decoding method according to an embodiment of the present invention;
7 is a block diagram schematically illustrating a configuration of an image encoding apparatus according to an embodiment of the present invention.
8 is a flowchart illustrating a method of encoding an image according to an embodiment of the present invention.
9 is a block diagram schematically illustrating a configuration of an image decoding apparatus according to an embodiment of the present invention.
10 is a flowchart illustrating an image decoding method according to an embodiment of the present invention.

Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. It should be noted that, in adding reference numerals to the constituent elements of the drawings, the same constituent elements are denoted by the same reference numerals whenever possible, even if they are shown in different drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

In describing the components of the present invention, terms such as first, second, A, B, (a), and (b) may be used. These terms are intended to distinguish the constituent elements from other constituent elements, and the terms do not limit the nature, order or order of the constituent elements. When a component is described as being "connected", "coupled", or "connected" to another component, the component may be directly connected or connected to the other component, Quot; may be "connected," "coupled," or "connected. &Quot;

1 is a block diagram illustrating a block for encoding a current motion vector of list0 according to an embodiment of the present invention.

Referring to FIG. 1, all the blocks shown in FIG. 1 are blocks that reference a reference picture whose reference picture identifier is list0. Block D represents a current block having a current motion vector (MV), which is a current motion vector to be coded. Block A, block B, and block C represent neighboring blocks for block D.

1, MV ^A0 , MV ^B0 , and MV ^C0 And MV ^D0, respectively, block A, block B, block C, block D is a motion vector (hereinafter referred to "list0 motion vector" hereinafter) that refer to list0 reference picture having, each of the horizontal component (MV ^A0 _x, MV ^B0 _x , MV ^C0 _x And MV ^D0 _x ) and the vertical components (MV ^A0 _y , MV ^B0 _y , MV ^C0 _y And MV ^D0 _y ). Hereinafter, the list0 motion vector MV ^D0 of the block D, which is the current block, is referred to as 'list0 current motion vector'. list0 the current motion vector MV ^D0 is (2,0), and each motion vector MV list0 ^{A0, B0} and MV MV ^C0 is the neighboring blocks (2,0), (2,1) and (2,2) .

The above-described present block (block D) of list0 list0 predicted motion vector for a current motion vector: calculated as the (PMV ^D0 list0 Predicted Motion Vector), and expression (1) and, list0 predicted motion vector (PMV ^D0) is the horizontal component (PMV ^D0 _x ) and a vertical component (PMV ^D0 _y ).

Referring to Equation (1), the prediction list0 motion vector PMV ^D0 for the current motion vector list0 can be calculated using a specific function F The list0 motion vectors MV ^A0 , MV ^B0 and MV ^{C0 of the} block A, the block B, and the block C can be used.

FIG. 2 is a block diagram illustrating a block for coding a current motion vector list1 according to an embodiment of the present invention. Referring to FIG.

Referring to FIG. 2, all the blocks shown in FIG. 2 are blocks that reference a reference picture whose reference picture identifier is list1. Block D represents a current block having a current motion vector to be encoded, and is assumed to be the same block as block D shown in FIG. Block A, block B, and block C represent neighboring blocks for block D, and are assumed to be the same blocks as block A, block B, and block C shown in FIG.

Referring to FIG. 2, MV ^A1 , MV ^B1 , MV ^C1 And MV ^D1 are list1 motion vectors (MVs) of the block A, the block B, the block C, and the block D, and each of them is a horizontal component (MV ^A1 _x , MV ^B1 _x , MV ^C1 _x And MV ^D1 _x ) and the vertical components (MV ^A1 _y , MV ^B1 _y , MV ^C1 _y And MV ^D1 _y ). Here, the list1 motion vector MV ^D1 of the current block, block D, is referred to as 'list1 current motion vector'. As shown in Figure 2, the MV list1 ^D1 current motion vector is (0,2), and each of list1 motion vector MV ^{A1, B1} and MV MV ^C1 in the neighboring blocks (0,2), (1,1) And (2, 0). The predicted motion vector for the current motion vector of list1 of the current block (block D) is referred to as 'list1 predicted motion vector PMV ^D1 '. The list1 predicted motion vector can be calculated as shown in Equation (2), and is defined as having a horizontal component (PMV ^D1 _x ) and a vertical component (PMV ^D1 _y ).

Referring to Equation (2), the list1 predicted motion vector for the current motion vector list1 is obtained by substituting the motion vectors MV ^A1 , MV ^B1 and MV ^C1 of the neighboring blocks (blocks A, B, and ^C ) ) Is substituted for the variable.

In the H.264 / AVC standard, a list0 predictive motion vector for the current motion vector list0 and a list1 predictive motion vector for the current motion vector list1 are computed using a function calculating the median value as a specific function F . That is, the list0 predicted motion vector PMV ^D0 for the current motion vector of list0 is the median of the motion vectors MV ^A0 , MV ^B0, and MV ^C0 of the neighboring blocks (block A, block B, and block C) And the list1 predicted motion vector PMV ^D1 for the current motion vector of list1 is a median for the motion vectors MV ^A1 , MV ^B1 and MV ^C1 of the neighboring blocks A, B, . Represents such a calculation method by the following equation, list0 current motion vector (MV ^D0) list0 predicted motion vector (PMV ^D0) for the list1 predicted motion for a can be represented as shown in Equation 3, list1 current motion vector (MV ^D1) The vector PMV ^D1 can be expressed by Equation (4).

Using equation 1 (or equation 3) list0 current motion vector (MV ^D0) of list0 predicted motion vector (PMV ^D0) to obtain, by using the equation 2 (or equation 4) list1 current motion vector (MV ^D1 ) of the list 1 predicted motion vector (PMV ^D1 ). Can be compressed using the equation 5 list0 current motion vector (MV ^D0) thereto for list0 predicted motion vector (PMV ^D0) the soothing motion vector list0 difference vector (DMV ^D0 from: list0 Differential Motion Vector, 'list0 motion vector residual Quot; signal "). The list 1 predicted motion vector PMV ^D1 is subtracted from the list 1 current motion vector MV ^D1 to be compressed using Equation 6, and the list 1 differential vector DMV ^D1 : list 1 differential motion vector, can be obtained a signal also called '), list0 difference vector (DMV ^D0) and list1 difference vector (DMV ^D1) is transmitted is coded by a predetermined method, such as predefined, each entropy encoded.

As shown in Figs. 1 and 2, list0 the conventional motion vector coding method for calculating a current motion vector of the case, list0 predicted motion vector (PMV ^D0) through the median value is (2,0) of (MV ^D0) Using Equation (3), the list0 predicted motion vector PMV ^D0 becomes (2,1). Using Equation (4) according to the conventional motion vector coding method for calculating the list 1 predicted motion vector (PMV ^D1 ) through the intermediate value when the value of the current motion vector (MV ^D1 ) of the list 1 is (0, 2) the list 1 predicted motion vector PMV ^D1 becomes (1,1).

As described above, the conventional motion vector coding method using the intermediate values as the list0 predictive motion vector and the list1 predictive motion vector is a method in which the motion vector coding device and the motion vector decoding device promise to calculate the predictive motion vector using the intermediate value in advance , It is not necessary to encode and transmit 'additional information' as to whether a list0 motion vector and a list1 motion vector are used as list0 current motion vector and list0 predictive motion vector of list1 current motion vector and prediction list1 motion vector, respectively, in the motion vector coding apparatus , The coding efficiency, that is, the compression efficiency can be improved.

However, the list0 predicted motion calculated using the median vector (PMV ^D0) is, and may be different actual list0 current motion vector (MV ^D0), calculated using the median list1 predicted motion vector (PMV ^D1 ) May differ from the actual list1 current motion vector (MV ^D1 ). Of (2, 1) FIG. 1 and the list0 predicted motion vector (PMV ^D0) calculated using the intermediate value by way of example in Fig. 2 is not different from the list0 (2,0) the current motion vector (MV ^D0) . By using the equation (5) Obtaining a list0 difference vector (DMV ^D0), list0 difference vector (DMV ^D0) to be coded is (0, -1) is, the predicted motion list1 calculated using a median vector (PMV ^D1) (1, 1) is different from (0,2) which is the current motion vector MV ^D1 of the list 1, and the list 1 differential vector DMV ^D1 is calculated using the equation (6) The vector (DMV ^D1 ) becomes (-1, 1).

However, if the block A of the motion vector MV list0 ^A0 is (2, 0) the predictive motion vector list0 the (2,0) and the difference (PMV ^D0), the actual list0 current motion vector (MV ^D0) when used as a does not occur, by using the equation (5) in the obtained list0 difference vector (DMV ^D0) is a is a (0,0) list0 difference vector (DMV ^D0) to be encoded. If, using the motion vector MV list1 ^A1 is (0,2) of the block A in list1 predicted motion vector (PMV ^D1), the difference does not occur and the (0,2) real list1 current motion vector (MV ^D1) rather, using the equation (6) in the list1 obtained difference vector (DMV ^D1) is a is a (0,0) list1 difference vector (DMV ^D1) to be encoded.

In other words, calculated using the median list0 predicted motion vector (PMV ^D0) is (2, 1) will be more, list0 list0 the motion vector MV of the ^A0 (2,0) of the block A predicted motion vector using the ( PMV ^D0 ), the list 0 differential vector DMV ^D0 becomes (0, 0), and the amount of bits used to encode the same can be reduced. Further, the calculation by using the intermediate value list1 predicted motion vector (PMV ^D1) of (1, 1) will than, list1 list1 the motion vector MV of ^A1 (0,2) of the block A predicted motion vector using the ( PMV ^D1 , the list 1 differential vector DMV ^D1 becomes (0, 0), and the amount of bits used to encode it can be further reduced.

However, in the conventional motion vector coding method of using a median list0 current motion vector (MV ^D0) of list0 predicted motion vector (PMV ^D0) always so use the median value to compute, the MV ^A0 list0 motion vector of the block A Is used as the list0 predicted motion vector (PMV ^D0 ). And list1 it should always intermediate value to calculate the current motion vector list1 predicted motion vector (PMV ^D1) of (MV ^D1), using the list1 motion vector MV ^A1 of block A in list1 predicted motion vector (PMV ^D1) It is impossible.

Even if the list0 motion vector MV ^A0 of the block A is used as the list0 predicted motion vector PMV ^D0 , MV ^A0 and MV ^B0 And MV ^C0 It is necessary to transmit 'additional information' as to which list0 motion vector is used as the list0 predicted motion vector PMV ^D0 . Even if the list ¹ motion vector MV ^A1 of the block A is used as the list ¹ predicted motion vector PMV ^D1 , MV ^A1 and MV ^B1 And MV ^C1 And 'additional information' indicating which list 1 motion vector is used as the list 1 predicted motion vector PMV ^D1 . Therefore, there is another problem that it is impossible to guarantee the improvement of the compression efficiency by encoding additional information.

Accordingly, the motion vector coding method for a plurality of reference pictures according to an embodiment of the present invention can more accurately select a predictive motion vector for a plurality of reference pictures, thereby obtaining a motion vector using a more accurately predicted motion vector So that it can be encoded. In addition, the motion vector coding method for a plurality of reference pictures according to an embodiment of the present invention may improve the coding efficiency by selecting more accurate prediction motion vectors, and may further include additional information to inform a predicted motion vector for a plurality of selected reference pictures Can be reduced.

Hereinafter, a description will be given of an embodiment of the present invention in which blocks (block A, block B, block C, and block D) shown in FIG. 1 and motion vectors MV ^A0 , MV ^B0 , and MV ^C0 And, blocks shown in FIG. 2 and MV ^D0 (block A, block B, block C, and block D), and thereto (hereinafter referred to as "list1 motion vector" hereinafter) for vector movement of the list1 reference pictures of MV ^A1, MV ^B1 , MV ^C1 And MV ^D1 .

1 and 2, the list ⁰ motion vectors MV ^A0 , MV ^B0 , and MV ^C0 And MV ^D0 ) and list1 motion vectors (MV ^A1 , MV ^B1 , MV ^C1 And MV ^{D1 are} shown as a two dimensional vector having a vertical component and a horizontal component. However, this is for convenience of description, and the motion vector used in the present invention is not necessarily limited to this, can do. In FIGS. 1 and 2, neighboring blocks of the current block (block D) are represented by three blocks A, B, and C in accordance with spatial adjacency. However, The peripheral blocks used are not necessarily limited to this, and may be one or more peripheral blocks temporally or spatially surrounding.

1 and 2, a motion vector for a reference picture is referred to as a list0 motion vector MV ^A0 , MV ^B0 , MV ^C0 And MV ^D0 ) and list1 motion vectors (MV ^A1 , MV ^B1 , MV ^C1 And MV ^D1 . However, since a B picture has either list0 or list1 or both list0 and list1 as reference pictures, it is not necessary to necessarily have motion vectors for two reference pictures (list0 and list1). Furthermore, it is possible to have two or more reference pictures, in which case a plurality of reference pictures may be list0, list1, ... , list n can be represented by reference pictures and the corresponding motion vectors are denoted by list0 motion vector, list1 motion vector, ... , list n motion vectors, and so on.

The motion vector coding mode according to an embodiment of the present invention includes a predictable mode and an unpredictable mode. For example, the predictive mode may be a predictive motion vector (hereinafter referred to as an " optimal motion vector ") determined by predicting a current motion vector, which is a motion vector of a current block, ) &Quot;) can be predicted by a motion vector decoding apparatus or an image decoding apparatus. The unpredictable mode refers to a mode for identifying that an optimal motion vector can not be predicted in a motion vector decoding apparatus or an image decoding apparatus.

In the present invention, the optimal motion vector means only a predictive motion vector determined by predicting the current motion vector according to a preset reference or method in the motion vector coding apparatus, and thus the predicted motion vector thus determined is always the optimal predicted value Does not mean that it has. Also, the default motion vector may be a motion vector coding apparatus, a motion vector decoding apparatus, a motion vector decoding apparatus, or a motion vector decoding apparatus, ) Calculation method, and the like), respectively.

3 is a block diagram schematically illustrating a configuration of a motion vector coding apparatus according to an embodiment of the present invention.

3, a motion vector coding apparatus 300 according to an embodiment of the present invention includes an optimal motion vector determination unit 310, a motion vector coding mode determination unit 320, a first motion vector coding unit 330 And a second motion vector coding unit 340.

The optimal motion vector determination unit 310 determines an optimal motion vector for a plurality of reference pictures by predicting the current motion vector of the current block with respect to a plurality of reference pictures. That is, the optimal motion vector determination unit 310 selects a set of candidate motion vectors that can be used as list0 and list1 optimal motion vectors for list0 and list1 current motion vectors of the current block, and selects one candidate motion vector from the selected list0 and list1 candidate motion vectors. The candidate motion vectors are determined as a set of optimal motion vectors for list0 and list1. Here, the list0 and list1 candidate motion vector sets may include one or more candidate motion vectors.

Here, the optimal motion vector determination unit 310 selects a candidate motion vector set for the list0 optimal motion vector set and a candidate motion vector set for the list1 optimal motion vector, or sets a candidate motion vector set for the list0 and list1 optimal motion vectors, respectively You can choose to share. When selecting a candidate motion vector set for the list0 optimum motion vector and a candidate motion vector set for the list1 optimal motion vector, the optimal motion vector determination unit 310 selects candidate motion vectors Selects a candidate motion vector set for the list 0 best motion vector among the vectors and selects a candidate motion vector set for the list 1 best motion vector among the candidate motion vectors that refer to the list 1 reference picture when selecting the list 1 best motion vector.

When the candidate motion vector set for the list0 and list1 optimal motion vectors is shared and selected, the optimal motion vector determination unit 310 determines candidate optimal motion vectors for the list0 optimal motion vector (i.e., Candidate motion vectors referring to the list 1 reference pictures among the candidate motion vectors for the list 1 optimal motion vector are scaled appropriately considering the temporal distance and selected as the candidate motion vectors for the list 0 optimal motion vector, When selecting a motion vector, candidate motion vectors that refer to the list0 reference picture among the candidate motion vectors to be shared are appropriately scaled considering the temporal distance and selected as a candidate motion vector set for the list1 optimal motion vector.

In addition, the optimal motion vector determination unit 310 may search for one or more neighboring blocks of the current block, and may calculate and collect the list0 and list1 motion vectors of the searched one or more neighboring blocks to select as a candidate motion vector set. 1 and 2, the set of candidate motion vectors includes a motion vector of a block A, a block B, and a block C, which are neighboring blocks on the left, top, MV ^A , MV ^B , MV ^C }.

However, the optimal motion vector determination unit 310 may select more various motion vectors as a candidate motion vector set according to an implementation method or a need. For example, a motion vector of a block located at the same position as the current block in a previous reference picture on the time axis or a motion vector of a block located at the upper left of the spatial axis may be selected as a candidate motion vector set. Another motion vector selected using motion vectors (e.g., an average or median value of one or more motion vectors, etc.) may also be included. The candidate motion vector set may be defined in various ways on the assumption that the motion vector coding apparatus and the motion vector decoding apparatus are defined in advance. If some or all of the candidate motion vectors in the candidate motion vector set have the same value Only candidate motion vectors having different values can be selected.

As described above, the optimal motion vector determiner 310 selects one candidate motion vector from a set of candidate motion vectors selected by various methods, and determines the candidate motion vectors as list0 and list1 optimal motion vectors. The optimal motion vector determination unit 310 calculates a selection function value for each of the candidate motion vectors using a predetermined selection function in the motion vector coding apparatus and the motion vector decoding apparatus, The motion vectors are determined as list0 and list1 optimal motion vectors.

As an example, the above-mentioned selection function value may include a bit amount required to encode list0 and list1 differential vectors which are the differences between list0 and list1 current motion vectors for one or more candidate motion vectors in the selected candidate motion vector set, A list0 difference between list0 and list1 current motion vectors for each of the candidate motion vectors, and a bit quantity required to encode the motion vector coding mode and the size of the list1 differential vector. If the bit values of the list0 and list1 difference vectors are used as the selection function values, the optimal motion vector determination unit 310 calculates the bit amounts required to encode the list0 and list1 differential vectors for the selected one or more candidate motion vectors And a candidate motion vector generating a minimum amount of bits of the calculated bit amount can be selected as list0 and list1 optimal motion vectors.

As another example, when a motion vector of one or more selected candidate motion vectors is selected, the optimal motion vector determination unit 310 determines a rate of considering a bit amount required for encoding and a quality of a reconstructed image generated at this time - The optimal list0 and list1 motion vectors may be determined using Rate-Distortion Optimization. In this case, the above-mentioned selection function value may be a Rate-Distortion Cost.

As another example, the optimal motion vector determiner 310 may determine a Lagrangian Cost function defined by Equations (7), (8), and (9) as a list0 and a list1 optimal motion vector . In this case, the above-mentioned selection function value can be a cost of raglane.

J ⁰ denotes a Lagrangian cost, J ⁰ denotes a Lagrangian cost for a reference picture list ⁰ , J ¹ denotes a Lagrangian cost for a reference picture list ¹ , D ⁰ denotes original picture data, and list ⁰ reference picture Represents an error between reconstructed image data, D ¹ represents an error between original image data and reconstructed image data using list ¹ reference pictures, and λ represents a Lagrangian multiplier. R _H represents the amount of bits required to code the motion vector coding mode, R ⁰ _M represents the amount of bits required to code the list0 differential vector of the current motion vector, R ¹ _M are list1 the differential vector of the current motion vector Indicates the amount of bits required for encoding.

w is a weight value, and may have a value of 1 or 2, depending on the implementation. In the embodiment, when the motion vector coding modes for the list0 and list1 reference pictures are respectively determined, the motion vector coding mode becomes two, so the value of w can be two. In another embodiment, when the motion vector coding modes for list0 and list1 are determined to be one, the motion vector coding mode becomes one, so that the value of w may be one.

J, J ⁰ , J ¹ , D ⁰ , D ¹ , R _H , R ⁰ _M , and R ¹ _M in Equations (7) to (9) are all expressed by n representing the number of the current picture in which the current block is located, Since it is defined according to k indicating the number of blocks, the determination of the optimal motion vector using the Lagrangian cost can be selectively applied on a picture or block basis. In the case where D, which is an error between the original image data and the reconstructed image data, does not change in the process of determining the optimal motion vector or for convenience of calculation, Equation 7 and Equation 8 for calculating the Lagrangian cost J It is also possible to simplify the equation by removing D ⁰ , D ¹ and λ.

In the process of calculating the Lagrangian cost, R _H in Equation (9) is the amount of bits required for coding the motion vector coding mode, and R ⁰ _M and R ¹ _M in Equations (7) and (8) The amount of bits required for coding the differential vector, and the calculation method thereof depends on the motion vector coding mode. That is, when the motion vector coding mode is in the unpredictable mode, R ⁰ _M and R ¹ _M are set to list0 and list1 predictive motion vectors (hereinafter referred to as 'list0' and 'list0') generated by a predetermined method such as predetermined or median And the list1 default motion vector) and the difference between the list0 and list1 current motion vectors (list0 and list1 differential vectors). When the motion vector coding mode is the predictive mode, R ⁰ _M and R ¹ _M are the bit amounts required to encode the list0 and list1 differential vectors, which are differences between the determined list0 and list1 optimal motion vectors, and the current motion vectors of list0 and list1, .

The optimal motion vector determiner 310 may determine the list0 and list1 optimal motion vectors using the Lagrangian Cost function described above through Equations 7 through 9 as a selection function, And a more generalized selection function such as Equation (11) may be used to determine optimal list0 and list1 motion vectors. However, it was represented by assuming a motion vector MV ^D0 in equation (10) and Equation (11) shows a block at list0 1 also the current motion vector of a current block to code D, also the list1 current motion vector of the current block to be coded a motion vector MV D ^D1 of one block shown in Figure 2 were expressed on the assumption.

In Equation 10, PMV ⁰ _enc is determined list0 best represents the motion vector, PMVC ⁰ is list0 current motion vector MV ^D0 list0 best motion vector candidate motion a set of candidate motion vectors a set of vectors that can be determined by the (CS: Candidate Set (I. E., A motion vector). h () is a selection function for determining the list 0 best motion vector for the current motion vector MV ^D0 .

In Equation (11), PMV ¹ _enc is a list ¹ best motion vector, PMVC ¹ is a list ¹ candidate motion vector set (CS 1) which is a set of candidate motion vectors that can be determined as a list ¹ best motion vector for list ¹ current motion vector MV ^D1 (I. E., A candidate motion vector).

In Equations (10) and (11), h () is a selection function for selecting the list 1 best motion vector for the current motion vector MV ^D1 . For example, the selection function h () can be obtained by coding the bit amount and the motion vector coding mode used for coding by differentiating the list0 and list1 from the current motion vector, The sum of the amounts of bits required to perform the above operation can be used. In order to simplify the calculation, it is also possible to use the sizes of the list0 and list1 difference vectors, which are differences between the current motion vectors list0 and list1 and the optimal motion vectors list0 and list1, instead of the actual bit amounts. More generally, the selection function h () may be variously defined on the assumption that it is preset in the motion vector coding apparatus and the motion vector decoding apparatus. Given this selection function h (), one candidate motion vector (PMVC ⁰ ) that optimizes the selection function h () from the candidate motion vector set (CS) containing the candidate motion vector as a candidate for the best motion vector is called the list0 optimum It can be determined by a motion vector (PMV ⁰ _enc ).

The motion vector coding mode determination unit 320 determines a motion vector coding mode according to whether or not an optimal motion vector for a plurality of reference pictures can be predicted by the motion vector decoding apparatus. That is, the motion vector coding mode determination unit 320 determines a motion vector coding mode as a predictive mode or a prediction mode according to whether a motion vector decoding apparatus can predict the list0 and list1 optimal motion vectors determined by the optimal motion vector determination unit 310. [ And controls the first motion vector coding unit 330 or the second motion vector coding unit 340 to encode the current motion vectors of list0 and list1 according to the determined motion vector coding mode.

Here, the motion vector coding mode determination unit 320 calculates a determination function value for each of the candidate motion vectors using a predetermined determination function between the motion vector coding apparatus and the motion vector decoding apparatus, The candidate motion vectors selected from among the one or more candidate motion vectors are determined as list0 and list1 estimated optimal motion vectors for the current motion vectors list0 and list1, respectively, and the determined list0 and list1 estimated optimal motion vectors, it is possible to determine whether the list0 and list1 optimal motion vectors can be predicted by the motion vector decoding apparatus by comparing list1 best motion vectors.

For example, a list0 and list1 differential vectors calculated using list0 and list1 optimal motion vectors (PMV ⁰ _enc and PMV ¹ _enc ), a finite number of candidate motion vectors that can be candidates of list0 and list1 optimal motion vectors, And the residual signal which is a difference between the predicted pixel value of the prediction block generated by motion compensation of the pixel value of the current block and the predicted pixel value of the current block is used as the optimal motion vectors of list0 and list1, It is possible to judge whether or not it can be predicted by the vector decoding apparatus.

To this end, the motion vector coding mode determining unit 320 determines a motion vector coding mode of the motion vector coding unit 300 based on the list 0 differential vector DMV ^D0 (= MV ^D0 -PMV ⁰ _enc ) for the list0 current motion vector MV ^D0 to be calculated and transmitted by the motion vector coding apparatus 300, PMV ⁰ _dec, which is a list ⁰ estimated optimal motion vector, is determined using information of a neighboring block that has already been coded, decoded and reconstructed, a reference picture to be used for motion compensation, and a decision function such as Equation (12) the list1 differential vector DMV to ^{D1) D1 (= MV D1 -PMV} 1 enc), list1 using a criterion function, such as a reference picture, and the equation (13) used for the already encoded and decoded information of the reconstructed neighboring blocks, the motion compensation The estimated optimal motion vector PMV ¹ _dec can be determined.

In Equation (12), the motion vector coding mode determination unit 320 of the motion vector coding apparatus 300 uses the information of the list0 differential vector and information of the neighboring blocks that have been already coded and decoded and restored , And determines whether the list0 optimal motion vector (PMV ⁰ _enc ) can be predicted in the motion vector decoding apparatus or the video decoding apparatus. The determination function g () may also be used in determining a list0 estimated optimal motion vector in the motion vector decoding apparatus.

In Equation 13, the motion vector coding mode determination unit 320 of the motion vector coding apparatus 300 uses the information of the list 1 differential vector and neighboring blocks that have been already coded and decoded and restored , And determines whether the list ¹ best motion vector (PMV ¹ _enc ) can be predicted in the motion vector decoding apparatus. Further, the determination function g () is also used in predicting the list 1 estimated optimal motion vector in a motion vector decoding apparatus or an image decoding apparatus to be described later. This determination function g () may be defined in various ways on the premise that it is previously set in the motion vector coding apparatus 300 and the motion vector decoding apparatus.

According to Equation (12), the motion vector coding mode determining unit 320 previously calculates the list0 estimated optimal motion vector PMV ⁰ _dec , which is a motion vector to be estimated by the motion vector decoding apparatus, and the motion vector decoding apparatus calculates the list0 differential vector DMV ^D0 MV = ^D0 -PMV using ⁰ _enc) list0 judges whether the optimal motion vector (PMV ⁰ can be correctly predicted by the _enc) to restore the image data correctly. That is, the motion vector coding mode determination unit 320 performs a process of determining a list0 estimated optimal motion vector to be performed by the video decoding apparatus, and uses the resultant list0 estimated optimal motion vector to encode the motion vector.

According to Equation (13), the motion vector coding mode determination unit 320 previously calculates the list ¹ estimated optimal motion vector PMV ¹ _dec , which is a motion vector to be estimated by the motion vector decoding apparatus, and the motion vector decoding apparatus calculates the list ¹ differential vector DMV ^D1 = MV ^D1 -PMV ¹ _enc ) is used to correctly predict the list ¹ best motion vector PMV ¹ _enc to determine whether the image data can be correctly restored. That is, the motion vector coding mode determination unit 320 performs a process of finding the list 1 estimated optimal motion vector to be performed by the video decoding apparatus in advance, and uses the resultant list 1 estimated optimal motion vector to encode the motion vector.

For example, the motion vector coding mode decision unit 320 is calculated using Equation 12 list0 estimated optimal motion vector (PMV ⁰ _dec) and list0 optimal motion vector (PMV ⁰ determined by the optimal motion vector determiner 310 _the motion vector decoding apparatus can correctly restore the current motion vector MV ^D0 by adding the list0 estimated optimal motion vector PMV ⁰ _dec directly estimated to the list0 differential vector DMV ^D0 provided from the motion vector coding apparatus So that the correctly reconstructed image data can be obtained.

Also, the motion vector coding mode determination unit 320 may determine that the list ¹ estimated optimal motion vector PMV ¹ _dec calculated using Equation (13) is the list ¹ best motion vector PMV ¹ _enc , The motion vector decoding apparatus can correctly restore the current motion vector MV ^D1 by adding the list1 estimated optimal motion vector PMV ¹ _dec directly estimated to the list1 differential vector DMV ^D1 provided from the motion vector coding apparatus Accordingly, it is possible to obtain correctly reconstructed image data.

Accordingly, the motion vector coding mode determination unit 320 determines the motion vectors of the list ⁰ and list ¹ predicted motion vectors PMV ⁰ _enc and PMV ¹ _enc determined by the optimal motion vector determination unit 310, list0, which is estimated by the motion vector decoding apparatus, If the list1 estimated optimal motion vectors (PMV ⁰ _dec and PMV ¹ _dec ) are the same, it is determined that the list0 and list1 optimal motion vectors (PMV ⁰ _enc and PMV ¹ _enc ) can be predicted by the motion vector decoding apparatus, Can not be predicted.

In addition, the motion vector coding mode determination unit 320 may determine whether the list0 and list1 optimal motion vectors (PMV ⁰ _enc and PMV ¹ _enc ) determined by the optimal motion vector determination unit 310 and the motion vectors estimating list0 and list1 estimated optimal motion vector (PMV ⁰ _dec and PMV ¹ _dec) even if more than the predetermined vector threshold value, list0 in the motion vector decoding apparatus and a list1 optimal motion vector (PMV ⁰ _enc and PMV ¹ _enc) the difference between the It can be judged that it can not be predicted. Here, the vector boundary value refers to a value that can be set freely in a unit of magnitude of a motion vector or through empirical expression.

As another example, in a case where the compression rate of the image is high, the change of the pixel value of the image is not large, or the change of the motion vector of the image is not large, the motion vector coding mode determination unit 320 determines that the list ⁰ best motion vector PMV ⁰ _enc ) And list0 estimated optimal motion vector (PMV ⁰ _dec ) are different from each other by the list ⁰ estimated optimal motion vector PMV ⁰ _dec , ^D0 = DMV ^D0 + PMV ⁰ _dec) the motion compensated image data and list0 optimal motion vector (PMV ⁰ _enc) list0 a current motion vector restored by using the (ie, MV ^D0 using = DMV ^D0 + PMV ⁰ _enc ), It is determined that the motion vector decoding apparatus can predict the list0 optimal motion vector (PMV ⁰ _enc ) using the list ⁰ estimated optimal motion vector PMV ⁰ _dec , and the other It can be judged that it can not be predicted.

Also, even if the list ¹ optimal motion vector PMV ¹ _enc and the list ¹ estimated optimal motion vector PMV ¹ _dec are not the same, the motion vector coding mode determination unit 320 uses the list ¹ estimated optimal motion vector PMV ¹ _dec The reconstructed list1 current motion vector (i.e. MV ' ^D1 = DMV ^D1 + PMV _dec ¹⁾ and motion-compensated image data and list1 optimal motion vector (PMV ¹ _enc) list1 a current motion vector restored by using the (ie, MV ^D1 using = DMV ^D1 + PMV ¹ _enc (For example, when the sum of absolute differences (SAD) between the two reconstructed image data is '0'), the motion vector decoding apparatus calculates the list 1 estimated optimal motion vector PMV ¹ _dec ) can be used to determine that the list ¹ optimal motion vector PMV ¹ _enc can be predicted, and otherwise, it can be determined that the prediction can not be performed.

In some cases, the motion vector coding mode determination unit 320, in order to increase the compression ratio further, list0 estimated optimal motion vector (PMV ⁰ _dec) list0 a current motion vector (i.e., MV ^D0 restored using = DMV ^D0 + PMV ⁰ _dec) the motion compensated image data and list0 optimal motion vector (PMV ⁰ _enc) list0 a current motion vector restored by the (i. E., Using the MV ^{D0 D0} = DMV + PMV ⁰ _enc (For example, when the sum of absolute differences (SAD) between two reconstructed image data is equal to or less than a predetermined boundary value), motion vector decoding apparatus that can be determined list0 estimated optimal motion vector (PMV ⁰ _dec) using a list0 optimum motion vector determined that predict the (PMV ⁰ _enc), and can not predict when the other. Here, the data boundary value refers to a value that can be freely set by a calculation formula or empirically as a unit capable of expressing the size of data such as a bit amount of data.

Also, even if the list ¹ optimal motion vector PMV ¹ _enc and the list ¹ estimated optimal motion vector PMV ¹ _dec are not the same, the motion vector coding mode determination unit 320 uses the list ¹ estimated optimal motion vector PMV ¹ _dec The reconstructed list1 current motion vector (i.e. MV ' ^D1 = DMV ^D1 + PMV _dec ¹⁾ and motion-compensated image data and list1 optimal motion vector (PMV ¹ _enc) list1 a current motion vector restored by using the (ie, MV ^D1 using = DMV ^D1 + PMV ¹ _enc ), The motion vector decoding apparatus calculates the list ¹ optimal motion vector PMV ¹ _enc using the list ¹ estimated optimal motion vector PMV ¹ _dec , It can be judged that it can be predicted, and in the other cases, it can be judged that it can not be predicted.

various types of functions may be applied based on the assumption that the decision function used for calculating the list 0 and list 1 estimated optimal motion vectors is previously set in the motion vector coding apparatus 300 and the motion vector decoding apparatus. However, since the predetermined decision function can be used in both list0 and list1, the reference pictures are not separately described.

As a determination function g () in Equations (12) and (13), a function using Template Matching (TM) or a function using boundary matching (BM) can be used.

For example, when a function using Template Matching (TM) is used as a determination function, a Template Matching Set (TMS) May be defined as a set of indices indicating the relative positions of the selected pixels, for example, the positions of one or more surrounding pixels adjacent to the left, top left, and top of a designated block. If necessary, the TMS can be selected by another method. However, if the number of pixels indicated by the TMS is usually large, more accurate matching is possible, but the calculation amount is increased.

The template matching method selects all candidate motion vector sets (CS) that can be selected as the estimated optimal motion vector, and then selects a block (hereinafter referred to as a reference block) designated by each candidate motion vector in the selected candidate motion vector set. And the difference between the pixels indicated by the TMS for the current block and the pixels indicated by the TMS for the current block are calculated using Equation 14 (Equations expressing one embodiment of Equations 12 and 13) And the vector having the smallest matching error among the calculated matching errors is determined as the estimated optimal motion vector (PMV _dec ).

In Equation ^{14, f (PMVC 1 + DMV} D1, i) (PMVC 1 + DMV D1) of the denotes a pixel location indicated by the reference block surrounding of the index i in the reference points, the index i (Included in TMS) picture, f (PMVC ¹ + DMV ^D1 , i) represents the pixel value at this pixel position. C (i) represents the pixel value around the current block indicated by the index i.

The decision function g PMVC ¹ DMV ^D1 is added to the difference vector DMV ^D1 provided by the motion vector coding apparatus 300 to the motion vector decoding apparatus 300 as a candidate motion vector PMVC ¹ ), Adds the pixel value of the reference block indicated by the calculated motion vector (PMVC ¹ + DMV ^D1 ) and the motion vector of the reference block indicated by the motion vector (PMVC ¹ + DMV ^D1 ) And a residual signal of the residual block obtained by subtracting the residual signal from the residual block. In order to calculate this, a Sum of Squared Error is used in Equation (14), but various functions such as sum of absolute difference (SAD) may be used depending on the application. The estimated optimal motion vector PMV ¹ _dec may be the candidate motion vector PMVC ¹ having the minimum value of the determination function g (PMVC ¹ | DMV ^D1 ).

That is, the motion vector coding mode determining unit 320 determines the motion vector coding mode of the current block, based on the pixels indicated by the TMS and the pixels indicated by the TMS with respect to the reference block indicated by each of the one or more candidate motion vectors included in the selected candidate motion vector set The pixel value difference may be calculated and a matching error for each of the candidate motion vectors may be calculated as a judgment function value based on the calculated pixel value difference.

As another example, when a function using boundary pixel matching is used as a determination function, a BMS (Bounding Matching Index Set) (hereinafter, referred to as BMS) But it may be defined as the position of all or some of the pixels located at the block boundary in the current block depending on the application.

In the boundary pixel matching method, after selecting all the candidate motion vector sets CS that can be selected as the predicted motion vectors, in order to determine which candidate motion vector PMVC among the candidate motion vector sets CS is the most optimal, The candidate motion vector PMVC that minimizes the error of the boundary pixel matching is determined as the estimated optimal motion vector PMV _dec by performing boundary pixel matching among the candidate motion vector sets CS. For this, the motion vector coding mode determination unit 320 not only calculates the matching error of each candidate motion vector by the sum of squares of differences as shown in Equation (15), but also calculates the sum of the differences of the absolute values SAD). ≪ / RTI >

In Equation (15), C (i) represents a candidate motion vector PMVC ^{1 that} is an element of the candidate motion vector set CS and a difference vector DMV ^{D !} Provided from the motion vector coding apparatus 300 to the motion vector decoding apparatus ^. ), Adds the pixel value of the reference block indicated by the calculated motion vector (PMVC ¹ + DMV ^D1 ) and the motion vector of the reference block indicated by the motion vector (PMVC ¹ + DMV ^D1 ) Is added to the residual signal of the residual block and the pixel value indicated by the index i in the BMS of the reconstructed pixels of the reconstructed current block.

Also, f (i) represents the pixel value of the pixel immediately adjacent to the pixel indicated by the index i of the BMS among the boundary pixels in the neighboring block adjacent to the current block. A matching error of the boundary pixel matching is calculated for each candidate motion vector PMVC ^{1 in} the candidate motion vector set using Equation 15 and a candidate motion vector for generating the minimum matching error among the calculated matching errors is estimated Is determined by the motion vector (PMV ¹ _dec ). That is, the estimated optimal motion vector refers to a predicted motion vector to be estimated by the motion vector decoding apparatus.

That is, the motion vector coding mode determination unit 320 determines a motion vector coding mode by using a difference vector between at least one candidate motion vector included in the selected candidate motion vector set, a pixel value of a reference block indicated by the corresponding candidate motion vector, A residual value of a residual block generated by compensating for a motion using a restored motion vector is added to a pixel value indicated by an index in the BMS of the reconstructed pixels of the restored current block, A matching error for each of the one or more candidate motion vectors may be calculated as a determination function value based on the difference between the pixel value of the pixel indicated by the index of the BMS and the pixel value of the adjacent pixel.

In addition, the motion vector coding mode determination unit 320 can determine the motion vector coding mode according to whether the optimal motion vector for a plurality of reference pictures can be predicted in the motion vector decoding apparatus. In this case, if the motion vector coding mode determination unit 320 can predict all the optimal motion vectors for a plurality of reference pictures in the motion vector decoding apparatus, the predictive mode can be determined as a motion vector coding mode. If it is not possible to predict at least one of the best motion vectors for a plurality of reference pictures, the predictive mode can be determined as a motion vector coding mode.

In addition, the motion vector coding mode determination unit 320 determines the motion vector coding mode independently for each of the plurality of optimal motion vectors according to whether the motion vector decoding apparatus can predict each of the optimal motion vectors for the plurality of reference pictures . At this time, the motion vector coding mode determination unit 320 can determine the predictive mode as the motion vector coding mode for the best motion vector predicted by the motion vector decoding apparatus among the optimal motion vectors for the plurality of reference pictures, The optimal motion vector for the reference picture that can not be predicted by the motion vector decoding apparatus can be determined as the motion vector coding mode.

When the motion vector coding mode is the predictive mode, the first motion vector coding unit 330 predicts an optimal motion vector for a plurality of reference pictures determined by the best motion vector determination unit 310 as a prediction motion for a plurality of reference pictures And motion information on a plurality of reference pictures is encoded using a predicted motion vector for a plurality of reference pictures and a current motion vector for a plurality of reference pictures. That is, when the motion vector coding mode is the predictable mode, the first motion vector coding unit 330 determines list0 and list1 optimal motion vectors as list0 and list1 predicted motion vectors for list0 and list1 current motion vectors, list1 generates list0 and list1 motion information using the current motion vector and the list0 and list1 optimal motion vectors, encodes the list0 and list1 motion information, and encodes the motion vector coding mode (i.e., the predictable mode).

Here, the first motion vector encoding unit 330 calculates list0 and list1 differential vectors, which are differences between the current motion vectors list0 and list1 and the optimal motion vectors list0 and list1, and codes the calculated list0 and list1 differential vectors, list1 < / RTI > and the current motion vector of list1 using the optimal motion vector. That is, list0 and list1 differential vectors can be generated and encoded as list0 and list1 motion information. The first motion vector encoding unit 330 not only encodes the list0 and list1 differential vectors as the list0 and list1 motion information but also generates the current motion vectors list0 and list1 as list0 and list1 motion information, The list0 and list1 motion information, i.e., list0 and list1 current motion vectors are encoded differently (for example, by using different variable length coding tables), depending on the characteristics (e.g., direction, size, etc.) And so on.

When the motion vector coding mode is in the unpredictable mode, the second motion vector coding unit 340 determines a default motion vector for a predetermined plurality of reference pictures as a predictive motion vector for a plurality of reference pictures, And motion information on a plurality of reference pictures is generated and coded using the motion vectors for the reference pictures and the current motion vectors for the plurality of reference pictures. That is, when the motion vector coding mode is in an unpredictable mode, the second motion vector coding unit 340 determines list0 and list1 default motion vectors as list0 and list1 predicted motion vectors for list0 and list1 current motion vectors, respectively, list0 and list1 motion information of list0 and list1 are generated and encoded using the current motion vector and list0 and list1 default motion vectors and the motion vector coding mode (i.e., the unpredictable mode) is encoded.

Here, the second motion vector coding unit 340 calculates list0 and list1 differential vectors, which are differences between list0 and list1 current motion vectors and list0 and list1 default motion vectors, as list0 and list1 motion information, respectively, and outputs the calculated list0 and list1 differential vectors The list0 and list1 default motion vectors can be used to encode the current motion vectors of list0 and list1. The second motion vector coding unit 340 not only encodes the list0 and list1 difference vectors but also the list0 and list1 differential motion vectors according to the characteristics (e.g., direction, size, etc.) (For example, coding is performed using different variable length coding tables), and the like.

In FIG. 3, the first motion vector coding unit 330 and the second motion vector coding unit 340 are independently implemented. However, the motion vector coding unit 330 and the second motion vector coding unit 340 may be implemented as a single motion vector coding unit. In this case, the combined motion vector coding unit may encode a current motion vector for a plurality of reference pictures using an optimal motion vector for a plurality of reference pictures or a default motion vector for a plurality of reference pictures according to a motion vector coding mode do.

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

The motion vector coding apparatus 300 determines an optimal motion vector for a plurality of reference pictures to encode the current motion vector for a plurality of reference pictures (S410). That is, the motion vector coding apparatus 300 determines list0 and list1 optimal motion vectors to encode the current motion vectors of list0 and list1.

The motion vector coding apparatus 300 determines an optimal motion vector for a plurality of reference pictures. The motion vector decoding apparatus 300 determines whether an optimal motion vector for a plurality of reference pictures can be predicted (S420) When it is determined that the optimal motion vector for a plurality of reference pictures can be predicted, list0 and list1 optimal motion vectors are determined as list0 and list1 predicted motion vectors (S430), and the predictable mode is determined as a motion vector encoding mode S440). When the motion vector decoding apparatus determines that the optimal motion vector for a plurality of reference pictures can not be predicted, the default list0 and list1 default motion vectors are determined as list0 and list1 predicted motion vectors (S432) As a motion vector coding mode (S442). That is, the motion vector coding apparatus 300 determines whether the motion vector decoding apparatus can predict the list0 and list1 optimal motion vectors. If it is determined that the list0 and list1 optimal motion vectors can be predicted, the motion vector coding apparatus 300 determines the predictable mode as a motion vector coding mode, The optimal motion vectors are determined as list0 and list1 predictive motion vectors. If it is determined that the optimal motion vectors are not predictable, the non-predictable mode is determined as a motion vector coding mode and the list0 and list1 default motion vectors are determined as list0 and list1 predicted motion vectors.

The motion vector coding apparatus 300 determines a predicted motion vector for a plurality of reference pictures. Using the predicted motion vector for the plurality of reference pictures and the current motion vector for the plurality of reference pictures determined in step S430 or step S432, Motion information on the reference picture is generated and encoded (S450). That is, the motion vector coding apparatus 300 generates and codes list0 and list1 motion information using list0 and list1 predictive motion vectors and list0 and list1 current motion vectors.

The motion vector coding apparatus 300 encodes the motion vector coding mode determined in step S440 or step S442 (S460). Motion vector coded data including motion information on a plurality of coded reference pictures and an encoded motion vector coding mode is generated and output (S470).

In step S410, the motion vector coding apparatus 300 selects a candidate motion vector set, selects one candidate motion vector from the set of candidate motion vectors, and determines the candidate motion vectors as list0 and list1 optimal motion vectors. In step S420, A method for determining whether the motion vector coding apparatus 300 can predict list0 and list1 optimal motion vectors in an image decoding apparatus and a method for determining whether or not the motion vector coding apparatus 300 determines whether the list0 and list1 current motion vectors, The method of generating and encoding the list0 and list1 motion information using the list1 predictive motion vector has already been described with reference to FIG. 3, and a detailed description thereof will be omitted.

As described above, the encoded motion vector encoded data can be decoded by a motion vector decoding apparatus to be described later.

FIG. 5 is a block diagram schematically illustrating a configuration of a motion vector decoding apparatus according to an embodiment of the present invention. Referring to FIG.

The video decoding apparatus 500 according to an embodiment of the present invention includes a decoding unit 510, a first predictive motion vector determination unit 520, a second predictive motion vector determination unit 530, and a motion vector restoration unit 540, As shown in FIG.

The decoding unit 510 reconstructs a coded motion vector coding mode included in the input motion vector coding data and motion information on a plurality of reference pictures. That is, the decoding unit 510 extracts the coded motion vector coding mode and the encoded list0 and list1 motion information from the motion vector coded data, and decodes them to restore the motion vector coding mode and the list0 and list1 motion information.

The decoding unit 510 analyzes the reconstructed motion vector coding mode, and when the motion vector coding mode is the predictable mode, the decoding unit 510 activates the first predictive motion vector determiner 520 or the first predictive motion vector determiner 520 520 to determine list0 and list1 optimal motion vectors and determine determined list0 and list1 optimal motion vectors as list0 and list1 predicted motion vectors. The decoding unit 510 analyzes the reconstructed motion vector coding mode and activates the second predictive motion vector determiner 530 or the second predictive motion vector determiner 530 when the motion vector coding mode is the predictive- To determine the list0 and list1 optimal motion vectors and determine the determined list0 and list1 optimal motion vectors as list0 and list1 predicted motion vectors.

The decoding unit 510 activates or controls the first predictive motion vector determination unit 520 and the second predictive motion vector determination unit 530 in the case where the motion vector encoding mode is one, There can be more than one mode each. That is, as described above with reference to FIG. 3, there may be a motion vector coding mode for each of a plurality of reference pictures, and there may be one motion vector coding mode for a plurality of reference pictures.

For example, if the motion vector coding mode is encoded for the list0 and list1 reference pictures in the motion vector coding apparatus 300, the motion vector coding mode will have a motion vector coding mode for list0 and list1, and the decoding unit 510 May analyze each of the motion vector coding modes for list0 and list1 and may activate or control the first predictive motion vector determination unit 520 and the second predictive motion vector determination unit 520 depending on the respective modes. That is, if the motion vector coding modes for list0 and list1 are both the predictive mode or the non-predictable mode, the first predictive motion vector determination unit 520 determines list0 and list1 estimated optimal motion vectors, Or the second predicted motion vector determining unit 530 may determine the list0 and list1 default motion vectors as list0 and list1 predicted motion vectors and determine motion of list0 in the motion vector coding mode for list0 and list1 If the vector coding mode is the predictive mode and the motion vector coding mode for list 1 is the unpredictable mode, the first predictive motion vector determination unit 520 may determine the list0 estimated optimal motion vector to be determined as the list0 predictive motion vector. And the second predicted motion vector determining unit 530 sets the predicted motion vector The vector Im can be determined as a predictive motion vector list1.

When the reconstructed motion vector coding mode is a predictive mode, the first predictive motion vector determination unit 520 determines an estimated optimal motion vector for a plurality of reference pictures and outputs an estimated optimal motion vector for a plurality of reference pictures, Are determined as prediction motion vectors for the reference pictures. The second predictive motion vector determination unit 530 determines a default motion vector for a predetermined plurality of reference pictures as a predictive motion vector for a plurality of reference pictures, when the reconstructed motion vector coding mode is an unpredictable mode.

That is, when the motion vector coding mode is the predictive mode, the first predictive motion vector determiner 520 determines list0 and list1 estimated optimal motion vectors, and determines list0 and list1 estimated optimal motion vectors as list0 and list1 predictive motion vectors . When the restored motion vector coding mode is in an unpredictable mode, the second predictive motion vector determiner 530 determines the list0 and list1 default motion vectors, which are generated according to a predetermined or preset reference, as list0 and list1 predictive motion vectors . Here, the method for determining the estimated optimal motion vectors list0 and list1 by the first predictive motion vector determination unit 520 will be described with reference to FIG. 3 by the motion vector encoding apparatus 300 or the optimal motion vector determination unit 310, Function is used to determine the estimated optimal motion vectors list0 and list1, the detailed description thereof will be omitted.

In FIG. 5, the first predictive motion vector determination unit 520 and the second predictive motion vector determination unit 530 are illustrated as being independently configured. However, the predictive motion vector determination unit 520 may be configured as a predictive motion vector determination unit It is possible. In this case, the predictive motion vector determination unit, which is composed of one unit, determines a best motion vector or a motion vector coding unit for a plurality of reference pictures determined by predicting a current motion vector of a current block in accordance with a motion vector coding mode, It is possible to determine a default motion vector for a plurality of predetermined reference pictures as a predictive motion vector for a plurality of reference pictures.

The motion vector restoring unit 540 restores the list0 and list1 predicted motion vectors determined by at least one of the first predicted motion vector determiner 520 and the second predicted motion vector determiner 530 and the list0 And list1 motion information to restore the current motion vectors of list0 and list1 and output them. Here, when the list0 and list1 motion information are list0 and list1 differential vectors, list0 and list1 predictive motion vectors may be added to list0 and list1 differential vectors to restore the current motion vectors of list0 and list1, but the present invention is not limited thereto, It is possible to restore the current motion vectors of list0 and list1 in the reverse of the manner in which list0 and list1 motion information are generated in the device 300. In this case, the motion vector coding apparatus 300, the first coding unit 330, And the motion vector decoding apparatus 500 and the motion vector restoring unit 540 may be set to the same.

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

In operation S610, the motion vector decoding apparatus 500 decodes the input motion vector coding data to restore the motion vector coding mode and list0 and list1 motion information, and determines whether the motion vector coding mode is the predictable mode in operation S620. , List0 and list1 estimated optimal motion vectors are determined and determined as list0 and list1 predicted motion vectors (S630). If the motion vector coding mode is in the unpredictable mode, the list0 and list1 default motion vectors are determined The motion vectors are determined as list0 and list1 predicted motion vectors (S632).

The motion vector decoding apparatus 500 restores the current motion vectors of list0 and list1 using the motion vector coding mode restored in step S610 and the list0 and list1 predictive motion vectors determined in step S630 or step S632 in step S640.

Here, the motion vector decoding apparatus 500 analyzes the motion vector coding mode to determine list0 and list1 estimated optimal motion vectors, and determines list0 and list1 as predicted motion vectors, or list0 and list1 default motion vectors as list0 and list1 predicted motion vectors And a method of restoring the current motion vectors of list0 and list1 using the motion vector coding mode and the list0 and list1 predictive motion vectors are described above with reference to FIG. 5, and thus detailed description thereof will be omitted.

The motion vector encoding / decoding apparatus according to an embodiment of the present invention can be applied to an image encoding apparatus for encoding an image and an image decoding apparatus for decoding an image.

FIG. 7 is a block diagram of a video encoding apparatus according to an embodiment of the present invention. Referring to FIG.

The image encoding apparatus 700 according to an embodiment of the present invention includes a block mode determination unit 710, a prediction unit 720, a subtraction unit 730, a first encoding unit 740, a second encoding unit 750, An encoded data generating unit 760, a decoding unit 770, an adding unit 780, and a reference picture storage unit 790. The image encoding apparatus 700 may be a personal computer (PC), a notebook computer, a personal digital assistant (PDA), a portable multimedia player (PMP), a PlayStation Portable ), A mobile communication terminal, and the like, and may be a communication device such as a communication modem for performing communication with various devices or wired / wireless communication networks, a memory for storing various programs for encoding an image and data, a program And a microprocessor for calculating and controlling the microprocessor.

The block mode determination unit 710 applies a predetermined optimum criterion (e.g., a rate-distortion optimization criterion) to the block modes that can be selected for the current block to be coded in the image, (E.g., a block mode with a minimum rate-distortion cost). If the block mode is preset in the image encoding apparatus 700, the block mode determination unit 710 is not necessarily included in the image encoding apparatus 700 and may be selectively omitted.

The prediction unit 720 predicts the current block to generate and output a prediction block. That is, the predicting unit 720 predicts a pixel value of each pixel of a current block to be encoded in an image and outputs a predicted block having a predicted pixel value of each predicted pixel . The prediction unit 720 may be configured to include a motion vector coding unit 722 and a motion compensation unit 724 as shown in the case of performing inter prediction, The optimal motion vector or motion vector decoding apparatus and a default motion vector for a predetermined plurality of reference pictures are used according to whether or not an optimal motion vector for a plurality of reference pictures determined by predicting a current motion vector can be predicted by a motion vector decoding apparatus Generates motion vector coding data by generating and encoding motion information on a plurality of reference pictures, and generates a prediction block of a current block using a current motion vector for a plurality of reference pictures.

The motion vector coding unit 722 may be implemented by the motion vector coding apparatus 300 according to an embodiment of the present invention described above with reference to FIG. 3, and the detailed description thereof will be omitted since it has been described with reference to FIG. However, the motion vector coding unit 722 can determine a current motion vector for a plurality of reference pictures, which are motion vectors of a current block for a plurality of reference pictures. When determining a current motion vector for a plurality of reference pictures, - distortion optimization, and so on.

The motion compensation unit 724 stores a current motion vector for a plurality of reference pictures output from the motion vector coding unit 722 in a reference picture indicated by index information about a reference picture output from the motion vector coding unit 722, And generates and outputs a prediction block of the current block.

The subtraction unit 730 subtracts the prediction block from the current block to generate a residual block. That is, the subtractor 730 calculates the difference between the pixel value of each pixel of the current block to be encoded and the predicted pixel value of each pixel of the predicted block predicted by the predictor 720, and outputs a residual signal ) ≪ / RTI >

The first encoder 740 transforms and quantizes the residual block and outputs the quantized residual block. That is, the first encoder 740 converts the residual signal of the residual block into a frequency domain, transforms each pixel value of the residual block into a frequency coefficient, and quantizes the residual block having the frequency coefficient. Here, the first encoder 740 may include various transformation techniques for transforming image signals on the spatial axis into a frequency axis such as a Hadamard Transform and a DCT based Transform (DCT based Transform) The residual signal transformed into the frequency domain is the frequency coefficient. The first encoding unit 740 encodes the transformed residual block into a dead zone uniform boundary quantization (DZUTQ), a quantization weighted matrix, Method, or the like.

In the above description, the first encoder 740 transforms and quantizes the residual block. However, the residual signal of the residual block may be transformed to generate the residual block having the frequency coefficient, and the quantization process may not be performed. Not only the quantization process can be performed without converting the residual signal of the block into the frequency coefficient, and even the conversion and the quantization process can not be performed. In the case where the transformation and quantization are not performed, the first encoder 740 in the image encoding apparatus 740 according to the embodiment of the present invention may be omitted.

The second encoder 750 encodes the residual block output from the first encoder 740 to generate and output residual block encoded data. That is, the second encoder 750 scans the quantization frequency coefficient, the frequency coefficient, or the residual signal of the residual block according to various scan methods such as zigzag scan to generate a quantization frequency coefficient sequence, a frequency coefficient sequence or a signal sequence, Entropy coding) technique. The functions of the first encoder 740 and the second encoder 750 may be integrated into one encoder.

The encoded data generation unit 760 generates encoded data including the residual block encoded data output from the encoding unit 750 and the motion vector encoded data output from the motion vector encoding unit 722 and outputs the encoded data. The encoded data generation unit 760 may further include information on the block mode output from the block mode determination unit 710 or the preset current block in addition to the encoded data. The coded data generator 760 may be implemented as a multiplexer (MUX).

The decoding unit 770 performs inverse quantization and inverse transform of the residual block quantized by the first encoding unit 740. That is, the decoding unit 770 dequantizes the quantized frequency coefficients of the coarse residual block to generate a residual block having a frequency coefficient, and inversely transforming the dequantized residual block to generate a residual block having a pixel value, And generates a residual block. Here, the decoding unit 770 can perform inverse transform and inverse quantization using the transform method and the quantization method used in the first encoding unit 740 inversely. If the first encoder 740 does not perform the quantization but only the inverse transform, the decoder 770 does not perform the inverse quantization, and only the quantization is performed in the first encoder 740 If the transform is not performed, only the inverse quantization is performed and the inverse transform is not performed. If the first encoding unit 740 does not perform both the conversion and the quantization or the first encoding unit 740 is omitted in the image encoding apparatus 700 and is omitted, the decoding unit 770 also performs inverse transform and inverse transform The quantization may not be performed at all or may be omitted in the image encoding apparatus 700.

The adding unit 780 adds the prediction block predicted by the predicting unit 720 and the residual block restored by the decoding unit 770 to restore the current block. The reference picture storage unit 790 stores the reconstructed current block output from the addition unit 780 as a reference picture on a picture-by-picture basis, and when the prediction unit 720 encodes the next block of the current block or another block in the future, .

Although not shown in FIG. 7, the image encoding apparatus 700 according to an embodiment of the present invention described above includes an intra prediction unit for intra prediction, an intra prediction unit for intra prediction based on the H.264 / AVC standard, And a deblocking filter unit for deblocking filtering the received signal. The first encoding unit 740 and the decoding unit 770 convert and quantize (or inverse transform and inverse-quantize) a specific picture (e.g., intra picture) based on the H.264 / AVC standard May be performed. Here, deblocking filtering refers to a task of reducing block distortion caused by coding an image block by block. The deblocking filter may be applied to a block boundary or a macro block boundary, a deblocking filter may be applied to a macro block boundary, Can be selectively used.

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

The image coding apparatus 700 determines a motion vector coding mode according to whether or not an optimal motion vector for a plurality of reference pictures determined by predicting a current motion vector of a current block for a plurality of reference pictures can be predicted by a motion vector decoding apparatus (S810). That is, as described above with reference to FIG. 3, the image encoding apparatus 700 determines list0 and list1 optimal predicted motion vectors, determines whether the list0 and list1 optimal predicted motion vectors can be determined in the image decoding apparatus, Mode. At this time, the motion vector coding mode can be determined to be a predictive mode and an unpredictable mode, as described above with reference to FIG. 3, and may be determined for each of a plurality of reference pictures or may be determined as one.

The image coding apparatus 700 generates motion information on a plurality of reference pictures using an optimal motion vector for a plurality of reference pictures or a default motion vector for a plurality of reference pictures according to a motion vector coding mode in operation S820. That is, as described above with reference to FIG. 3, the image coding apparatus 700 generates list0 and list1 motion information using list0 and list1 optimal motion vectors or list0 and list1 default motion vectors according to a motion vector coding mode.

The image coding apparatus 700 generates motion vector coding data by coding motion information and a motion vector coding mode for a plurality of reference pictures (S830), and predicts the current block using the current motion vector for a plurality of reference pictures (S840). The residual block generated by subtracting the current block from the prediction block is coded to generate residual block coded data (S850), and coded data including the motion coded data and the residual block coded data is generated Output.

As described above, the image encoded with the encoded data by the image encoding apparatus 700 can be transmitted through the wired / wireless communication network such as the Internet, the local area wireless communication network, the wireless LAN network, the WiBro network, the mobile communication network, , A universal serial bus (USB), or the like, to be transmitted to a video decoding apparatus to be described later, decoded by the video decoding apparatus, and restored and reproduced as an image.

FIG. 9 is a block diagram of a video decoding apparatus according to an embodiment of the present invention. Referring to FIG.

An image decoding apparatus 900 according to an exemplary embodiment of the present invention includes an information extracting unit 910, a first decoding unit 920, a second decoding unit 930, a predicting unit 940, an adding unit 950, And a reference picture storage unit 960. The video decoding apparatus 900 may be a personal computer (PC), a notebook computer, a personal digital assistant (PDA), a portable multimedia player (PMP), a PlayStation Portable ), A mobile communication terminal, and the like, and may be a communication device such as a communication modem for performing communication with various devices or wired / wireless communication networks, a memory for storing various programs and data for decrypting video, a program And a microprocessor for calculating and controlling the microprocessor.

The information extraction unit 910 receives the encoded data, extracts information on the block mode (for example, an identifier), and outputs information on the extracted block mode. If the block mode is the motion vector skip mode (for example, the block mode is the intra 16x16 mode or the intra 4x4 mode), the information extracting unit 910 extracts the motion vector encoded data from the encoded data, Only the encoded data can be extracted and output. On the other hand, when the block mode is not the motion vector skip mode (for example, the block mode is the inter 16x16 mode, the inter 4x4 mode, the P8x8 mode, etc.), the information extracting unit 910 extracts, from the encoded data, And extracts and outputs block coded data. At this time, the information extracting unit 910 can further extract and output index information about the reference picture from the encoded data.

The first decoding unit 920 decodes the residual block coded data output from the information extracting unit 910. That is, the first decoding unit 920 decodes the binary data of the residual block coded data using an entropy coding technique or the like to generate a quantized frequency coefficient sequence, performs inverse scanning by various scanning methods such as inverse zigzag scanning, Lt; / RTI > If the binary data of the residual block coded data is binary data in which the frequency coefficients are coded, the residual block decoded by the first decoding unit 920 will be a residual block having a frequency coefficient, and the binary data of the residual block coded data The residual block decoded by the first decoding unit 920 will be a residual block having the residual signal if the non-quantized residual signal is the encoded binary data.

The second decoding unit 930 dequantizes and inversely transforms the residual block decoded by the first decoding unit 920 and restores the residual block. That is, the second decoding unit 930 dequantizes the quantized frequency coefficients of the decoded residual block output from the first decoding unit 920, and inversely transforms the dequantized frequency coefficients to recover the residual block having the residual signal . If the residual block decoded by the first decoding unit 920 has a quantization frequency coefficient, the second decoding unit 930 performs both of the inverse quantization and the inverse transform. However, if the first decoding unit 920 If the decoded residual block has a frequency coefficient, only the inverse transform can be performed without performing inverse quantization. If the residual block decoded by the first decoding unit 920 has only a residual signal, inverse quantization and inverse transform Or the second decoding unit 930 may not be configured in the image decoding apparatus 900 and may be omitted. 9, the first decoding unit 920 and the second decoding unit 930 are illustrated as being independently configured, but they may be configured as one decoding unit (not shown) that integrates the functions of the first decoding unit 920 and the second decoding unit 930 .

The prediction unit 940 predicts the current block to generate a prediction block. The prediction unit 940 may include a motion vector decoding unit 942 and a motion compensation unit 944. The prediction unit 940 may decode the motion vector coding data output from the information extraction unit 910, And reconstructs motion information for a plurality of reference pictures and predicts a current motion vector for a plurality of reference pictures according to the reconstructed motion information and the motion vector coding mode, And restores the current motion vector for a plurality of reference pictures using the apparatus and a preset default motion vector and generates a prediction block of the current block using the current motion vector for the plurality of reference pictures reconstructed.

The motion vector decoding unit 942 may be implemented as a motion vector decoding apparatus 500 according to an embodiment of the present invention described above with reference to FIG. That is, the motion vector decoding unit 942 decodes the motion vector coding data to restore the motion vector coding mode and the list0 and list1 motion information, and determines whether the motion vector coding mode is predictive mode or predictive mode, And list1 estimated optimal motion vectors to determine list0 and list1 predicted motion vectors or determined list0 and list1 default motion vectors as list0 and list1 predicted motion vectors, and list0 and list1 predicted motion vectors and list0 and list1 motion information, respectively. To restore the current motion vectors of list0 and list1.

The motion compensation unit 944 predicts the reference blocks indicated by the current motion vectors of the list0 and list1 restored by the motion vector decoding unit 942 in the reference pictures stored in the reference picture storage unit 960 as prediction blocks of the current block Thereby generating a prediction block. Here, when using the reference picture, the motion vector decoding unit 942 outputs the index information of the reference picture from the information extracting unit 910 to the reference picture among the many reference pictures stored in the reference picture storage unit 960 It is possible to use the reference picture identified by the index information.

The adder 950 restores the current block by adding the restored residual block output from the second decoding unit 930 to the predicted block output from the predicting unit 940. The reconstructed current block may be accumulated as a reconstructed picture on a picture-by-picture basis or may be stored in a reference picture storage unit 960 as a reference picture and used to predict a next block.

Although not shown in FIG. 9, the video decoding apparatus 900 according to an embodiment of the present invention described above includes an intra prediction unit for intraprediction based on the H.264 / AVC standard, a deblocking filtering A deblocking filter unit for performing deblocking filtering, and the like. In addition, the second decoding unit 930 may further perform inverse transform and inverse quantization on a specific picture (e.g., intra picture) based on the H.264 / AVC standard.

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

The video decoding apparatus 900 receives encoded video data through a wired / wireless communication network or a cable and stores the encoded data. The video decoding apparatus 900 decodes and restores the video in order to reproduce the video according to a user's selection or an algorithm of another program being executed.

For this, the video decoding apparatus 900 decodes input encoded data to recover residual blocks, a motion vector coding mode, and motion information on a plurality of reference pictures (S1010) The current motion vector for a plurality of reference pictures is reconstructed using the estimated optimal motion vector for the reference picture or the default motion vector for the plurality of reference pictures (S1020).

That is, the image decoding apparatus 900 decodes the input encoded data to recover the residual block, the motion vector coding mode, and the list0 and list1 motion information, and outputs list0 and list1 estimated optimal motion vectors or list0 And list1 default motion vectors to restore the current motion vectors of list0 and list1.

The video decoding apparatus 900 generates a prediction block by predicting the current block using the current motion vectors of the restored reference pictures, that is, list0 and list1 current motion vectors (S1030), and outputs the reconstructed residual block and the prediction block And restores the current block (S1040).

4, 6, 8, and 10 and the description thereof is only an order according to one embodiment of the present invention, and the order of each step within the scope of the essential characteristics of the present invention May be modified and implemented.

In the above description, the motion vector coding mode is classified into the predictive mode and the unpredictable mode. However, the present invention is not limited to this, and a default motion vector for a plurality of reference pictures may be divided into a plurality of reference pictures And a mode for using a best motion vector for a plurality of reference pictures as a predicted motion vector for a plurality of reference pictures according to a mode used as a predictive motion vector and a preset reference or method.

According to an embodiment of the present invention described above, the motion vector coding apparatus 300 or the image coding apparatus 700 can select and determine a prediction mode for a plurality of reference pictures, The amount of bits required to encode the current motion vector for a plurality of reference pictures can be minimized by using a motion vector equal to or more similar to a vector as a predictive motion vector for a plurality of reference pictures, The efficiency can be improved.

Also, according to an embodiment of the present invention, the image coding apparatus 300 or the motion vector coding apparatus 300 can improve the coding efficiency by using the predicted motion vectors for the more accurate reference pictures, A function for transmitting or searching only information such as motion information and a motion vector coding mode so as to be directly obtained from the motion vector decoding apparatus 500 or the video decoding apparatus 700 without notifying the predicted motion vector for the picture directly to the decoding apparatus, It is possible to reduce an increase in the amount of additional bits generated for notifying the predicted motion vector for a plurality of reference pictures, so that the coding efficiency and the decoding efficiency can be further improved.

Further, when an embodiment of the present invention is applied to an image processing service or the like, it is possible to encode an image with a small bit amount, thereby providing a highly satisfactory service to a user. Especially, it can be expected to be more effective in a wireless mobile environment that can have a relatively small bandwidth, large data loss and delay compared to a wired environment.

While the present invention has been described in connection with what is presently considered to be the most practical and preferred embodiment, it is to be understood that the invention is not limited to the disclosed embodiments. That is, within the scope of the present invention, all of the components may be selectively coupled to one or more of them. In addition, although all of the components may be implemented as one independent hardware, some or all of the components may be selectively combined to perform a part or all of the functions in one or a plurality of hardware. As shown in FIG. The codes and code segments constituting the computer program may be easily deduced by those skilled in the art. Such a computer program can be stored in a computer-readable storage medium, readable and executed by a computer, thereby realizing an embodiment of the present invention. As the storage medium of the computer program, a magnetic recording medium, an optical recording medium, a carrier wave medium, or the like may be included.

Furthermore, the terms "comprises", "comprising", or "having" described above mean that a component can be implanted unless otherwise specifically stated, But should be construed as including other elements. All terms, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. Commonly used terms, such as predefined terms, should be interpreted to be consistent with the contextual meanings of the related art, and are not to be construed as ideal or overly formal, unless expressly defined to the contrary.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are intended to illustrate rather than limit the scope of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

As described above, the present invention is applied to an image processing field for encoding and decoding an image, so as to more accurately predict a predicted motion vector of a current motion vector for a plurality of reference pictures, to calculate a bit amount And the compression efficiency can be improved.

Claims (16)

In an image decoding method using inter prediction,
Reconstructing a first differential motion vector and a second differential motion vector of a current block located in a current picture by decoding the encoded data;
Deriving a first predicted motion vector and a second predicted motion vector of the current block from one or more neighboring blocks of the current block;
A second motion vector prediction unit for generating a first motion vector of the current block by adding the first difference motion vector and the first predicted motion vector to each other and adding the second difference motion vector and the second predicted motion vector, Generating a motion vector;
Generating a prediction block of the current block using a first motion vector and a second motion vector of the current block;
Reconstructing a residual block by decoding the residual signal included in the encoded data; And
And adding each pixel of the prediction block and a corresponding pixel of the residual block,
Wherein the step of restoring the residual block comprises:
Identifying one or more conditions for dequantization and inverse transformation; And
Quantizing the decoded residual signal, skipping the inverse quantization and inverse transform of the decoded residual signal, dequantizing the decoded residual signal and skipping the inverse transform of the dequantized residual signal, Reconstructing the residual block by selecting one of the processes of performing both inverse quantization and inverse transform on the residual signal
And decoding the decoded image. delete The method according to claim 1,
Wherein deriving the first and second predicted motion vectors comprises:
Deriving a first candidate motion vector set and a second candidate motion vector set using the motion vectors of the at least one neighboring block;
Selecting a first candidate motion vector in the first set of candidate motion vectors and a second candidate motion vector in a second set of backward motion vectors; And
Setting the first candidate motion vector to the first predicted motion vector and setting the second candidate motion vector to the second predicted motion vector
And decoding the decoded image. The method of claim 3,
Wherein the one or more neighboring blocks include one or more blocks spatially adjacent to the current block in the current picture and one or more blocks located in a reference picture decoded prior to the current picture,
Wherein a block located within a coded reference picture is determined based on a position of the current block in the current picture. 5. The method of claim 4,
Wherein at least one block spatially adjacent to the current block includes blocks located at left, upper, upper right, and upper left of the current block. The method of claim 3,
Wherein the step of deriving the first set of candidate motion vectors comprises the step of including only one of the two candidate motion vectors derived from the motion vectors of the at least one neighboring block in the first candidate motion vector set ,
Wherein the step of deriving the second set of candidate motion vectors comprises the step of including only one of the two candidate motion vectors derived from the motion vectors of the at least one neighboring block in the second candidate motion vector set And decodes the decoded image. The method of claim 3,
Wherein the first candidate motion vector and the second candidate motion vector are selected based on information included in the encoded data. The method according to claim 1,
Wherein one or more conditions for the dequantization and inverse transformation are identified based on information included in the encoded data. In an image decoding apparatus using inter prediction,
An information extraction unit for decoding the encoded data to reconstruct a first difference motion vector and a second difference motion vector of a current block located in a current picture;
A first predictive motion vector and a second predictive motion vector of the current block are derived from one or more neighboring blocks of the current block, and the first differential motion vector and the first predictive motion vector are added, Generating a second motion vector of the current block by adding the second difference motion vector and the second predicted motion vector to generate a motion vector, and generating a second motion vector of the current block using the first motion vector and the second motion vector of the current block, A prediction unit for generating a prediction block of a current block;
A decoding unit for decoding a residual signal included in the encoded data to reconstruct a residual block; And
And an adder for adding each pixel of the prediction block and a corresponding pixel of the residual block,
Wherein the decoding unit comprises:
Identifying one or more conditions for dequantization and inverse transformation,
Quantizing the decoded residual signal, skipping the inverse quantization and inverse transform of the decoded residual signal, dequantizing the decoded residual signal and skipping the inverse transform on the dequantized residual signal, And performing the inverse quantization and inverse transform on the residual signal, thereby reconstructing the residual block. delete 10. The method of claim 9,
The predicting unit,
Deriving a first candidate motion vector set and a second candidate motion vector set using the motion vectors of the at least one neighboring block,
Selecting a first candidate motion vector in the first set of candidate motion vectors and a second candidate motion vector in a second set of backward motion vectors,
Sets the first candidate motion vector to the first predictive motion vector and sets the second candidate motion vector to the second predictive motion vector. 12. The method of claim 11,
Wherein the one or more neighboring blocks include one or more blocks spatially adjacent to the current block in the current picture and one or more blocks located in a reference picture decoded prior to the current picture,
Wherein a block located within a coded reference picture is determined based on a position of the current block in the current picture. 13. The method of claim 12,
Wherein at least one block spatially adjacent to the current block includes blocks located at the left, upper, upper right, and upper left of the current block. 12. The method of claim 11,
Wherein the step of deriving the first candidate motion vector set includes the step of including only one of the two candidate motion vectors derived from the motion vectors of the at least one neighboring block in the first candidate motion vector set, Including,
Wherein the step of deriving the second set of candidate motion vectors includes the step of including only one of the two candidate motion vectors derived from the motion vectors of the at least one neighboring block in the second candidate motion vector set Wherein the video decoding apparatus comprises: 12. The method of claim 11,
Wherein the first candidate motion vector and the second candidate motion vector are selected based on information included in the encoded data. 10. The method of claim 9,
Wherein the decoding unit identifies one or more conditions for the inverse quantization and inverse transform based on information included in the encoded data.

Images