KR101567964B1 - Method and Apparatus for Encoding and Decoding Motion Vector - Google Patents

Abstract

The present invention relates to a motion vector coding / decoding method and apparatus, and more particularly, to a motion vector coding / decoding method and apparatus for selecting a candidate motion vector set for a current motion vector; Selecting a predicted motion vector from a set of candidate motion vectors; Determining whether a predicted motion vector is predictable in a decoding apparatus; Encoding the current motion vector using a motion vector prediction mode indicating a predicted motion vector and a predictable motion vector; And a current motion vector is encoded using a motion vector prediction mode that indicates a default predicted motion vector and a non-predictable motion vector, when it is determined that the motion vector can not be predicted.
According to the present invention, by sharing a predetermined function for giving or searching for information so as to find a predicted motion vector without notifying the predicted motion vector for the current motion vector directly to the decoding apparatus, a predicted motion vector is determined according to the motion vector prediction mode And the motion vector prediction mode is transmitted to the decoding apparatus by selecting and encoding the current motion vector, thereby improving the coding bit amount reduction, the motion prediction performance, and the motion compensation performance.

Description

[0001] The present invention relates to a motion vector coding / decoding method,

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving picture encoding / decoding method, and more particularly, to a motion vector coding / decoding method and apparatus in a moving picture coding / decoding field.

[0002] Recently, multimedia technology has been rapidly developed, and accordingly, demand for high-quality multimedia data including audio, image, and moving picture is increasing. Accordingly, an international standard for high-efficiency video compression has been established to meet the demand for transmitting, storing and re-reading of such multimedia data in a limited network environment. In particular, for video, 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, block prediction, intra prediction, and intra predictive coding. Predictive coding is widely used for compressing various data as an effective way to reduce the correlation between data. In particular, since the motion vector has a close correlation with the motion vector of the neighboring block, the predicted value (PMV: predictive motion vector, also referred to as "predicted motion vector") of the motion vector of the current block is calculated from the motion vector of the neighboring block Only the difference vector (DMV: "residual signal", "difference value") for the predictive value is encoded without coding the motion vector value of the current block after the current block, thereby significantly reducing the amount of bits to be encoded The coding efficiency can be increased.

Generally, in the motion vector coding using such a predictive motion vector, the coding efficiency increases as the predicted motion vector becomes more accurate for effective compression. Therefore, not only a motion vector of a spatial adjacent block but also a motion vector of a block temporally, spatially or temporally or spatially adjacent to another block or a motion vector of a finite number of other motion vectors calculated by combining them is generated, The prediction encoding efficiency can be further increased by selecting and using the most suitable one for predictive encoding. In this case, in order to correctly recover the original motion vector from the predictively encoded motion vector data, it is necessary to know which prediction motion vector is used among the finite number of predicted motion vector objects. The most simple motion vector prediction coding method for this purpose is to jointly encode information on which prediction value is used for motion vector prediction coding. Or the amount of bits required for additional data coding to indicate such a prediction motion vector selection, the current H.264 / AVC standard specifies a horizontal component of motion vectors of neighboring blocks (left, top, right top) And a median of each of the vertical components is used as a predictive motion vector (PMV) for predictive coding of the motion vector. In this method, a predetermined default method known in common for coding and decoding is called intermediate value, and information on which prediction value is used is calculated by calculating a predictive value (predictive motion vector) by using this default method in coding and decoding Thereby eliminating the need for coding. An existing method of predefining and using a predetermined default method of intermediate values is advantageous in that it can maintain an improved coding efficiency without transmitting additional information about 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 which always generates the minimum amount of bits required for coding the differential vector.

It is an object of the present invention to solve the above problems and provide a motion vector prediction method and a motion vector prediction method for selecting a predictive motion vector for a current motion vector by using a plurality of motion vector prediction modes, And to improve the performance of motion prediction and / or motion compensation.

Another object of the present invention is to provide a motion vector decoding apparatus and a motion vector decoding method in which a coding apparatus decodes information for searching for a predicted motion vector, instead of notifying a predictive motion vector used for coding directly, The present invention is to improve coding efficiency and decoding efficiency by reducing the increase in the amount of coding bits for additional information generated for informing a prediction motion vector by sharing a function for giving or searching for a device with a decoding device.

According to an aspect of the present invention, there is provided a motion vector coding method including: (a) selecting a set of candidate motion vectors for a current motion vector of a current block; (b) selecting a predicted motion vector in the selected set of candidate motion vectors; (c) determining whether the selected predictive motion vector is predictable in a decoding apparatus; (d) encoding the current motion vector using the selected predictive motion vector and a motion vector prediction mode indicating predictability, if it is determined to be predictable in step (c); And encoding the current motion vector using a predetermined default predictive motion vector and a motion vector prediction mode indicating a non-predictable mode if it is determined that the current motion vector is not predictable in step (c) And a motion vector coding method.

According to another aspect of the present invention, there is provided a motion vector decoding method comprising the steps of: (a) decoding an encoded motion vector prediction mode and a coded differential vector; (b) determining whether the decoded motion vector prediction mode indicates predictable or unpredictable; (c) if the decoded motion vector prediction mode indicates prediction, selecting a candidate motion vector set that can be selected as a predicted motion vector for restoring a current motion vector of a current block, and selecting one of the selected candidate motion vector sets Determining a candidate motion vector of the current motion vector as a predicted motion vector for the current motion vector; (d) determining the predetermined default predicted motion vector as a predicted motion vector for the current motion vector, if the decoded motion vector prediction mode indicates unexpected motion; And (e) reconstructing a current motion vector of the current block by adding the predictive motion vector determined in step (c) or step (d) and the decoded difference vector. Decoding method.

According to another aspect of the present invention, there is provided a motion vector coding apparatus for selecting a candidate motion vector set for a current motion vector of a current block, A predictive motion vector selection unit for selecting a vector; A motion vector prediction mode determining unit that determines whether the selected predictive motion vector is predictable by a decoding apparatus and determines a motion vector prediction mode indicating predictable or unpredictable according to the determination result; Determining a predictive motion vector for the current motion vector as the predictive motion vector if the determined motion vector prediction mode is predictable and determining a difference vector between the current motion vector and the selected predictive motion vector and the determined motion A first encoding unit encoding the current motion vector using a vector prediction mode; And determining a predetermined default predicted motion vector as a predicted motion vector for the current motion vector if the determined motion vector prediction mode is indicated as unpredictable and determining a difference vector between the current motion vector and the predetermined default predicted motion vector And a second encoding unit encoding the current motion vector using the determined motion vector prediction mode.

According to another aspect of the present invention, there is provided a motion vector decoding apparatus comprising: a motion vector prediction mode coded by an encoder; a motion vector prediction mode for decoding the coded difference vector; part; A motion vector prediction mode determiner for determining whether the decoded motion vector prediction mode indicates predictability or non-predictability; If the decoded motion vector prediction mode is predictable, selects a candidate motion vector set that can be selected as a predicted motion vector for recovering a current motion vector of a current block, and outputs a predicted motion vector in the selected candidate motion vector set A first predicted motion vector determining unit for selecting the selected predictive motion vector as a predictive motion vector for the current motion vector; A second predicted motion vector determiner for determining a predetermined default predicted motion vector as a predicted motion vector for the current motion vector, if the decoded motion vector prediction mode indicates unexpected motion; And a current motion vector reconstruction unit for reconstructing a current motion vector of the current block by adding the decoded difference vector to a predictive motion vector determined by the first predictive motion vector determination unit or the second predictive motion vector determination unit. The motion vector decoding apparatus comprising:

According to another aspect of the present invention, there is provided a motion vector coding method comprising: (a) selecting a predicted motion vector of a current motion vector according to a selected motion vector prediction mode among a plurality of motion vector prediction modes; ; (b) coding a differential vector between the current motion vector and the predictive motion vector; (c) encoding motion vector prediction mode information indicating the selected motion vector prediction mode; And (d) generating a bitstream including the coded difference vector and motion vector prediction mode information.

According to another aspect of the present invention, there is provided a motion vector decoding method comprising: (a) decoding a differential vector between a current motion vector and a predictive motion vector and motion vector prediction mode information of the current motion vector; ; (b) selecting a predicted motion vector of the current motion vector based on the motion vector prediction mode information; And (c) reconstructing the current motion vector using the selected predictive motion vector.

As described above, according to the present invention, a prediction motion vector for a current motion vector can be selected using a plurality of motion vector prediction modes to more accurately select a prediction motion vector, and a coding bit amount And the performance of motion prediction and / or motion compensation is improved.

In addition, according to the present invention, instead of notifying the predictive motion vector used for coding directly to the decoding apparatus while selecting a more accurate predictive motion vector to improve the coding efficiency, the present invention provides information for finding a predictive motion vector to a decoding apparatus The coding efficiency and the decoding efficiency can be further improved by reducing the increase in the amount of coding bits for additional information generated for informing a prediction motion vector by sharing a function for giving or searching.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram illustrating a block for coding a motion vector according to an exemplary embodiment of the present invention;
FIG. 2 is a block diagram of a motion vector coding apparatus according to an embodiment of the present invention. FIG.
3 is a block diagram of a motion vector decoding apparatus according to an embodiment of the present invention.
FIG. 4 is a schematic overall flowchart of a motion vector coding method according to an embodiment of the present invention;
FIG. 5 is a detailed overall flowchart of a motion vector coding method according to an embodiment of the present invention;
6 is a flowchart of a candidate motion vector set selection step in a motion vector coding method according to an embodiment of the present invention;
FIG. 7 is a flowchart illustrating a predictive motion vector selection step in the motion vector coding method according to an embodiment of the present invention;
8 is a flowchart illustrating a step of determining whether a predicted motion vector can be predicted in the motion vector coding method according to an embodiment of the present invention.
9 is a flowchart of a first encoding step in a motion vector encoding method according to an embodiment of the present invention,
FIG. 10 is a flowchart of a second encoding step in a motion vector encoding method according to an embodiment of the present invention;
11 is a schematic overall flowchart of a motion vector decoding method according to an embodiment of the present invention,
12 is a detailed overall flowchart of a motion vector decoding method according to an embodiment of the present invention,
13 is a flowchart illustrating a first predictive motion vector selection step in a motion vector decoding method according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the drawings, the same reference numerals are used to designate the same or similar components throughout the 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 unclear.

FIG. 1 is a block diagram illustrating a block for coding a motion vector according to an embodiment of the present invention. Referring to FIG.

In Fig. 1, block D is a "current block " corresponding to a motion vector to be coded, and block A, block B and block C are assumed to be" neighboring blocks "

Referring to FIG. 1, MV ^A , MV ^B , and MV ^C And MV ^D are the motion vectors (MVs) of the block A, the block B, the block C, and the block D, and each of the motion vectors MV ^A _x , MV ^B _x , MV ^C _x And MV ^D _x ) and the vertical components (MV ^A _y , MV ^B _y , and MV ^C _y And MV ^D _y ). Here, the motion vector MV ^D of the current block, block D, is referred to as a 'current motion vector'.

Referring to FIG. 1, the current motion vector MV ^D is (2,0), MV ^A , MV ^B and MV ^C , which are motion vectors of the neighboring blocks, are (2,0), (2,1) , 2).

A predictive motion vector PMV ^D for a current motion vector of the current block D is calculated as shown in Equation 1 and a predicted motion vector PMV ^D is calculated for a horizontal component PMV ^D _x , And a vertical component (PMV ^D _y ).

Referring to Equation (1), a predicted motion vector (PMV ^D ) for a current motion vector is calculated by dividing the motion vectors MV ^A , MV ^B, and MV ^C of neighboring blocks A, B, Is calculated as a variable of a specific function (F ()).

In the H.264 / AVC standard, a predictive motion vector (PMV ^D ) for the current motion vector is calculated by using a function for calculating a median (Median) as a specific function F (). That is, the predicted motion vector PMV ^D for the current motion vector is obtained as an intermediate value (Median) for the motion vectors MV ^A , MV ^B, and MV ^C of the neighboring blocks (blocks A, B, and ^C ) will be. The predicted motion vector PMV ^D for the current motion vector MV ^D calculated in this manner is expressed by the following equation (2).

To it in the equation (1) (or Equation (2)) to the current motion vector (MV ^D) a predictive motion vector the current motion vector (MV ^D) when determined that (PMV ^D), to be compressed by using the equation (3) using a predictive motion vector (PMV ^D) the soothing motion vector differential vector: can be obtained a (DMV ^D also known as differential motion Vector, "motion vector residual signal"), a difference vector (DMV ^D) is pre-defined, such as entropy encoding And is transmitted by a predetermined method.

As illustrated in FIG. 1, when the value of the current motion vector MV ^D is (2, 0), Equation 2 according to the conventional method for calculating the predicted motion vector PMV ^D through the median , The predicted motion vector PMV ^D is (2, 1).

In this manner, the intermediate value is used as the predictive motion vector because the encoder and the decoder know each other, so that there is no need to encode 'additional information' as to which motion vector is used as a predictive motion vector of the current motion vector It is not necessary to transmit the data, and the coding efficiency can be improved.

However, as described above, the predicted motion vector PMV ^D calculated using the intermediate value may be different from the actual current motion vector MV ^D. In the example of FIG. 1, it can be seen that the predicted motion vector PMV ^D (2,1) calculated using the intermediate value is different from (2,0), which is the current motion vector MV ^D , When using the equation (3) obtained the difference vector (DMV ^D) is a is (0, -1), the difference vector (DMV ^D) to be encoded.

However, if (2, 0) MV ^A , which is the motion vector of block A, is used as the predicted motion vector PMV ^D , there is no difference from (2, 0) which is the actual motion vector MV ^D , When using the equation (3) obtained the difference vector (DMV ^D) it is a is a (0, 0), the differential signal ^(D DMV) to be encoded.

That is, the calculated predicted using the median motion vector (PMV ^D) of (2, 1) will look predicted motion vector (PMV ^D) of the (2,0) block motion vector MV ^A of A using a The difference vector DMV ^D becomes (0, 0), and the amount of bits used to encode the differential vector DMV can be reduced.

However, in the case of using the intermediate value, since the median must always be used to calculate the predicted motion vector PMV ^D of the current motion vector MV ^D , the motion vector MV ^A of the block A is used as the predicted motion vector PMV ^D ).

Even if a motion vector (MV ^A) of the A block to a prediction motion vector ^{(PMV D), MV A,} MV B And MV ^C Since additional information is transmitted together with 'additional information' as to which motion vector is used as the predicted motion vector PMV ^D , there is another problem that it is impossible to guarantee the improvement of the compression efficiency.

Therefore, in the motion vector coding method according to the embodiment of the present invention, it is possible to select a predicted motion vector by using a plurality of motion vector prediction modes, thereby making it possible to select a predicted motion vector more accurately, As a predictive motion vector, to solve the problem in the conventional method.

In addition, the motion vector coding method according to an embodiment of the present invention proposes an efficient method of notifying a decoding apparatus of a selected predictive motion vector while improving the coding efficiency by selecting a more accurate predictive motion vector, And also to solve the problem of the increase of the encoding bit amount with respect to the additional information generated by the notification.

(Block A, block B, block C, and block D) shown in FIG. 1 and motion vectors MV ^A , MV ^B , and MV ^C And MV ^D are generally used in the following description.

1, the motion vectors MV ^A , MV ^B , and MV ^C And MV ^{D are} shown as a two-dimensional vector having a vertical component and a horizontal component, but this is for convenience of explanation and not limited thereto. That is, the motion vector coding and decoding of the present invention can be applied to an n-dimensional motion vector. In FIG. 1, the neighboring blocks of the current block (block D) are represented by three blocks A, B, and C according to the spatial adjacency. However, the present invention is not limited thereto. ≪ / RTI >

The method for achieving the above object according to the motion vector coding method according to an embodiment of the present invention will be described in more detail below.

The motion vector prediction mode according to an embodiment of the present invention includes a default prediction mode and a non-default prediction mode. Hereinafter, the default prediction mode will be described as an example of a prediction mode indicating that prediction of a predicted motion vector (optimal predicted motion vector) selected according to a predetermined criterion or method is not possible. On the contrary, the non-default prediction mode describes, as an example, a prediction mode indicating that prediction of a predicted motion vector (optimal predicted motion vector) selected according to a predetermined criterion or method is possible. Here, the term "optimal predicted motion vector" means a predicted motion vector of a current motion vector obtained according to a predetermined reference or method, but does not mean that the predicted motion vector thus obtained is always an optimal predicted value.

2 is a block diagram of a motion vector coding apparatus according to an embodiment of the present invention.

2, a motion vector coding apparatus 200 according to an embodiment of the present invention includes a predictive motion vector selection unit 210, a motion vector prediction mode determination unit 220, a first coding unit 230, A second encoder 240, and the like. Hereinafter, the motion vector coding apparatus 200 will be briefly described as an encoder hereinafter.

The predictive motion vector selection unit 210 selects a candidate motion vector set that is a set of candidate motion vectors that can be used as a predictive motion vector with respect to the current motion vector of the current block, Select. A method of selecting a predictive motion vector in a candidate motion vector set may, for example, select a candidate motion vector that minimizes a difference from a current motion vector in a candidate motion vector set as a predicted motion vector. The predictive motion vector selected from the candidate motion vector set by the predictive motion vector selection unit 210 is referred to as an optimal predictive motion vector for convenience of explanation. For the sake of convenience, however, the optimal predicted motion vector is defined as a specific example. In the present invention, the best predicted motion vector means a predicted motion vector of the current motion vector obtained according to a predetermined reference or method .

The motion vector prediction mode determination unit 220 determines whether the optimal predictive motion vector selected by the predictive motion vector selection unit 210 is predictable in the decoding apparatus and outputs a predicted motion vector Prediction mode "

If the predictive motion vector prediction mode is determined by the motion vector prediction mode determination unit 220, the first encoder 230 converts the "optimal predicted motion vector" selected by the predictive motion vector selection unit 210 to the current A difference vector between the current motion vector and the best predicted motion vector is calculated, and the current motion vector is encoded using the calculated difference vector and the motion vector prediction mode indicating the predictability do.

If the motion vector prediction mode determined by the motion vector prediction mode determination unit 220 is not predictable, the second encoder 240 converts a predetermined "default predicted motion vector" that is commonly known to the decoding apparatus to a current motion vector And calculates a difference vector between the current motion vector and a predetermined default predicted motion vector and encodes the current motion vector using the calculated difference vector and a motion vector prediction mode indicating an un-predictable motion vector do. Here, the "default predicted motion vector" is a predicted motion vector that is preset between the encoder and the decoder, as compared with the "optimal predicted motion vector" used in the first encoder 230. For example, 0.0 > (median) < / RTI >

The motion vector prediction mode determined by the motion vector prediction mode determination unit 220 and the differential vector coded by the first encoder 230 or the second encoder 240 are inserted into the transmission bitstream. The transmission bitstream is transmitted to a decoder through a transmission channel such as a wire, wireless, or storage medium, and then decoded by a motion vector decoding apparatus according to the present invention.

The motion vector coding method by the motion vector coding apparatus 200 according to an embodiment of the present invention described above with reference to FIG. 2 will be described in detail with reference to FIG. 4 to FIG.

3 is a block diagram of a motion vector decoding apparatus 300 according to an embodiment of the present invention.

3, a motion vector decoding apparatus 300 according to an embodiment of the present invention includes a motion vector prediction mode / differential vector decoding unit 310, a motion vector prediction mode determination unit 320, A second predictive motion vector determination unit 340, a current motion vector restoration unit 350, and the like. However, in the following, the motion vector decoding apparatus 300 is simply referred to as a decoding apparatus.

The motion vector prediction mode / differential vector decoding unit 310 decodes the motion vector prediction mode and the coded differential vector coded by the coding apparatus 200.

The motion vector prediction mode determination unit 320 determines whether the decoded motion vector prediction mode indicates predictability or impossibility of prediction.

When the decoded motion vector prediction mode is predictable, the first predictive motion vector determiner 330 selects a candidate motion vector set that can be selected as a predictive motion vector for restoring the current motion vector of the current block, Selects an " optimal predicted motion vector "from the set of candidate motion vectors, and determines the selected optimal predicted motion vector as a predicted motion vector for the current motion vector.

The second predictive motion vector determiner 340 determines a predictive motion vector for a current motion vector based on a predetermined " default predicted motion vector "that is known to the encoder 200, if the decoded motion vector prediction mode is predictable, It is determined as a vector.

The current motion vector reconstruction unit 350 reconstructs the current motion vector using the predicted motion vector (the best predicted motion vector or the default predicted motion vector) determined in the first predicted motion vector determiner 330 or the second predicted motion vector determiner 340, And adds the difference vector to restore the current motion vector of the current block.

The motion vector decoding method by the motion vector decoding apparatus 300 according to an embodiment of the present invention described above with reference to FIG. 3 will be described in more detail with reference to FIG. 11 to FIG.

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

Referring to FIG. 4, a motion vector coding method according to an exemplary embodiment of the present invention includes: selecting a predicted motion vector of a current motion vector according to a selected motion vector prediction mode among a plurality of motion vector prediction modes (S400); Encoding a difference vector between the current motion vector and the predicted motion vector (S402); Encoding motion vector prediction mode information indicating a selected motion vector prediction mode (S404); And generating a bitstream including the encoded difference vector and motion vector prediction mode information (S406).

In step S400, when the selected motion vector prediction mode is the default prediction mode, the predetermined default predicted motion vector is selected as the predicted motion vector of the current motion vector.

In the case where the selected motion vector prediction mode is the non-default prediction mode, a candidate motion vector of a plurality of candidate motion vectors is determined as a predicted motion vector of the current motion vector Vector).

4, it is to be understood that the present invention is not limited to the above-described exemplary embodiments, and that various modifications and changes may be made without departing from the scope of the present invention. You can also implement it by changing the order in. For example, although steps S402 and S404 are shown in FIG. 4, steps S404 and S402 may be performed. That is, it does not matter whether either one of them is performed first.

The motion vector coding method according to an embodiment of the present invention described above with reference to FIG. 4 will be described in detail with reference to FIG. 5 to FIG.

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

Referring to FIG. 5, a motion vector coding method according to an exemplary embodiment of the present invention includes a candidate motion vector set selection step S500, a predicted motion vector selection step S502, a predicted motion vector prediction availability determination step S504, A first encoding step S506, a second encoding step S508, and the like.

The candidate motion vector set selection step (S500) selects a set of candidate motion vectors for the current motion vector of the current block.

The predictive motion vector selection step S502 selects the best predicted motion vector from the selected candidate motion vector set.

The predictive motion vector predictability determination step S504 determines whether the selected optimal predictive motion vector is predictable in the decoding apparatus.

If it is determined in step S504 that the predicted motion vector is predictable, the first encoding step S506 encodes the current motion vector using a selected optimal predictive motion vector and a motion vector prediction mode indicating predictability.

If it is determined in step S504 that the predicted motion vector is not predictable, the second encoding step S508 codes the current motion vector using a predetermined default predictive motion vector and a motion vector prediction mode indicating a non-predictable motion vector.

The finally determined motion vector prediction mode and the differential vector coded in the first coding step S506 or the second coding step S508 are inserted into the transmission bitstream. The transport bit stream is transmitted to the decoder through a transport channel such as a wire, wireless or storage medium.

The above-described steps S500, S502, S504, S506, and S508 will be described in detail with reference to Figs. 6, 7, 8, 9, and 10, respectively.

6 is a flowchart illustrating a candidate motion vector set selection step (S500) in the motion vector coding method according to an exemplary embodiment of the present invention.

Referring to FIG. 6, a candidate motion vector set selection step (S500) in the motion vector coding method according to an exemplary embodiment of the present invention includes searching (S600) one or more neighboring blocks for a current block (S600) A step S602 of calculating a motion vector of a neighboring block, a step S604 of selecting a candidate motion vector set by combining the calculated candidate motion vectors, and the like.

1, the set of candidate motion vectors selected in the candidate motion vector set selection step (S500) described above is a block A, which is a neighboring block at the upper left, upper right, and upper right of the current block D, and B is a block ^{{MV a, MV B, MV} C} C is made up with the motion vector.

However, more generally, the candidate motion vector set can select more various motion vectors as a set of candidate motion vectors (a set of candidate motion vectors) according to the implementation method or the necessity. For example, a motion vector of a co-located block of a picture existing on the time axis or a motion vector of a block located on the upper left of the spatial axis may be used as a candidate motion vector. Further, another motion vector selected using these motion vectors (for example, an average value or an intermediate value of several motion vectors, etc.) may be included.

Accordingly, the step S600 may include: one or more first blocks located at the same position as the current block in at least one of a picture existing in a time axis and a picture existing in a later time; And one or more second blocks in a position adjacent to the current block on the spatial axis as one or more neighboring blocks.

In addition, step S602 includes: a motion vector of each of the at least one first block; Each motion vector in one or more second blocks; And an average or median value for at least one of each motion vector in the first block and each motion vector in the second block; May be calculated as a motion vector of one or more neighboring blocks searched.

The above-described candidate motion vector set can be defined in various ways on the assumption that the encoding device 200 and the decoding device 300 know the definition in advance. When a part or all of the candidate motion vectors, which are elements constituting the candidate motion vector set, have the same value, the candidate motion vector may be composed of only a set of candidate motion vectors having different values.

7 is a flowchart illustrating a predictive motion vector selection step S502 in the motion vector coding method according to an embodiment of the present invention.

Referring to FIG. 7, the step of selecting a predicted motion vector S502 in the motion vector coding method according to an embodiment of the present invention is performed by using a predetermined selection function between the coding apparatus 200 and the decoding apparatus 300 Calculating a selection function value for each of the one or more candidate motion vectors included in the selected candidate motion vector set (S700); And selecting one candidate motion vector as an optimal predicted motion vector based on the computed selection function values (S702).

As an example, the above-mentioned "selection function value" includes a bit amount required to code a differential vector between the current motion vector and each of one or more candidate motion vectors included in the selected candidate motion vector set; A size of a difference vector between the current motion vector and the current motion vector for each of the one or more candidate motion vectors included in the selected candidate motion vector set; And the amount of bits required to encode the motion vector prediction mode; Or a combination of one or more of them.

If the bit amount of the difference vector is used as the selection function value, in step S702 described above, a bit required to encode the difference vector for each of the one or more candidate motion vectors included in the selected candidate motion vector set And a candidate motion vector which is a minimum value among the calculated bit amounts can be selected as an optimum predicted motion vector.

In addition, the predictive motion vector selection step S502 in the motion vector coding method according to an embodiment of the present invention may include a method of selecting the best predicted motion vector based on the selected function value, A Rate-Distortion Optimization scheme that considers both the amount of bits required for coding occurring when one motion vector of one or more candidate motion vectors included in the motion vector set is selected and the reconstructed image quality to be generated at this time The optimal predicted motion vector can be selected.

In the above-described predictive motion vector selection step S502, the Lagrangian cost function in Equation (4) can be used as a selection function for selecting an optimal predicted motion vector.

Where J is the Lagrangian cost, D is the error between the original image and the reconstructed image, and λ is the Lagrangian multiplier. R _H is the amount of bits required to encode the motion vector prediction mode, and R _M is the amount of bits required to encode the difference vector of the current motion vector. J, D, R _H , and R _M in Equation (4) are all defined according to n representing a picture number where the current block is located and k representing the block number. Therefore, the present invention can be applied selectively in a picture or a block unit.

Also, in the case where D, which is an error between the original image and the reconstructed image, does not change in the process of selecting an optimal predicted motion vector, or for convenience of calculation, D and λ are removed from Equation 4 for calculating the Lagrangian cost J, It is possible.

In the process of estimating the raglane cost caused by the motion vector coding method, R _H in Equation (4) is a bit amount required for coding a motion vector prediction mode, and R _M is a difference vector for a current motion vector As the bit amount, the calculation method thereof depends on the motion vector prediction mode. That is, when the motion vector prediction mode indicates that the (optimal) prediction motion vector is not predictable by the decoding apparatus 300, R _M is a prediction generated by a predefined default method such as a median calculation Is the amount of bits required to code the difference vector of the current motion vector with a motion vector (referred to as a "default predicted motion vector" or "default predicted value"). In addition, when the motion vector prediction mode indicates that the decoding apparatus 300 can predict the best predicted motion vector, R _M is a vector that is used to encode a difference vector (differential value) between the selected optimal predicted motion vector and the current motion vector Bit amount.

In addition, in the above-described predictive motion vector selection step S502, in addition to selecting an optimal predicted motion vector using the Lagrangian cost function in Equation (4), a more generalized selection function expression The optimal predicted motion vector may be selected using Equation (5). Equation (5) is expressed by assuming that the current motion vector of the current block to be encoded is a motion vector MV ^D of the block D in FIG.

In Equation (5), PMV _enc is a selected optimum motion vector and PMVC is an element belonging to a candidate motion vector set (CS) which is a set of candidate motion vectors selectable as a predicted motion vector of the current motion vector MV ^D Vector). "h ()" is a selection function for selecting a prediction motion vector optimal for coding the current motion vector MV ^D by differentiating the current motion vector MV ^D from the predicted motion vector.

As an example of the selection function h () described above, a bit amount used for coding the current motion vector by using the difference is used, or a bit amount used for coding the current motion vector by difference and a bit amount Can be used. Also, in order to simplify the calculation, the size of the difference vector (residual signal), which is the difference between the current motion vector and the predicted motion vector, may be used instead of the actual amount of bits. More generally, the definition of the selection function h () can be defined and used in various ways on the premise that the encoding device 200 and the decryption device 300 know the definition in advance. Given this selection function h (), one candidate motion vector PMVC that optimizes the selection function h () from the candidate motion vector set CS containing the candidate motion vector that is a candidate of the predicted motion vector, Vector (PMV _enc ).

FIG. 8 is a flowchart illustrating a step S504 of determining whether a predicted motion vector can be predicted in the motion vector coding method according to an exemplary embodiment of the present invention.

Referring to FIG. 8, in step S504, it is determined whether a predicted motion vector can be predicted in the motion vector coding method according to an exemplary embodiment of the present invention. Calculating a decision function value for each of the one or more candidate motion vectors included in the selected candidate motion vector set (S800); Selecting one of the one or more candidate motion vectors as an estimated predicted motion vector for the current motion vector based on the calculated decision function value (S802); Comparing the selected estimated predicted motion vector and the selected optimal predicted motion vector (S804); And determining whether the optimal predictive motion vector selected according to the comparison result is predictable in the decoding apparatus 300 (S806).

The predicted motion vector predictability determination step S504 will be described below using the expression in Equation (5).

The predictive motion vector predictability determination step S504 includes a difference vector calculated using the optimal predictive motion vector PMV _enc selected in the preceding predictive motion vector selection step S502, (Step S502), using a finite number of candidate motion vectors that can be used for motion compensation, reference picture data to be used for motion compensation, information of the already reconstructed neighboring blocks, and motion compensation residual signals for pixel values It is determined whether the best predicted motion vector is predictable by the decoding apparatus 300 or the decoding method.

For this, the coding apparatus 300 firstly calculates DMV ^D (= MV ^D -PMV _enc ), which is a differential vector of the current motion vector MV ^D to be calculated and transmitted by the coding apparatus 200, The predictive motion vector PMV _dec is searched using the predetermined determination method as shown in Equation (6) using the reference picture data to be used for compensation.

In Equation 6, the determination function, g (), the coding unit 200 is calculated by transfer using the information of the difference vector and the already reconstructed neighboring blocks, the decoding device 300, a vector (PMV _enc) best prediction motion Is determined by the encoding device 200. [0050] This determination function g () is also used in predicting the predicted motion vector in the decoding apparatus 300. [

The determination function g () can be defined in various ways on the assumption that the encoding device 200 and the decryption device 300 know the definition in advance. A specific embodiment of the determination function g () will be described below.

The encoding apparatus 200 previously calculates PMV _dec which is the predicted motion vector to be found by the decoding apparatus 300 and then outputs DMV ^D (= MV ^D -PMV _enc ) and the information of the already reconstructed neighboring block to predict the best predicted motion vector PMV _enc to determine whether the reconstructed image data can be correctly recovered. That is, the coding apparatus 200 performs a process of finding an optimal predicted motion vector to be performed later by the decoding apparatus 300 in advance and uses the result at the time of coding

Some embodiments of steps S804 and S806 in step S504 will be described below.

An embodiment of a method for determining whether an optimal predicted motion vector PMV _enc is predictable by the decoding apparatus 300 is as follows.

If the estimated predicted motion vector PMV _dec calculated through Equation 6 in step S504 is equal to the optimal predicted motion vector PMV _enc selected in step S502, the decoding apparatus 300 determines whether the predicted motion vector PMV _dec Correct reconstructed image data can be obtained by restoring the accurate current motion vector MV ^D by adding the estimated motion vector PMV _dec , which is a prediction motion vector estimated by itself, to the difference vector DMV ^D. Therefore, when the optimal predicted motion vector PMV _enc directly selected by the decoding apparatus 300 is the same as the estimated predicted motion vector PMV _dec estimated to be found by the decoding apparatus 300, It is determined that prediction with respect to the optimum predicted motion vector PMV _enc is possible. Otherwise, it is determined that prediction is impossible. Alternatively, according to the embodiment, the coding apparatus 200 may determine that the difference between the directly selected optimal predicted motion vector PMV _enc and the estimated predicted motion vector PMV _dec that the decoding apparatus 300 estimates to find is smaller than a predetermined boundary It is determined that prediction of the best predicted motion vector PMV _enc in the decoding apparatus 300 is possible, and in the other cases, it may be determined that prediction can not be performed.

Another embodiment of a method for determining whether the optimum predicted motion vector PMV _enc is predictable by the decoding apparatus 300 is as follows.

The optimal predicted motion vector PMV _enc and the estimated predicted motion vector PMV _dec may be used in a case where the compression rate of the image is high or 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 restored using the estimated predicted motion vector PMV _dec (i.e., MV ' ^D = DMV ^D + PMV _dec ), and a motion vector reconstructed using the optimal predicted motion vector PMV _enc (i.e. MV ^D = DMV ^D + PMV _enc In this case, even if the optimal predicted motion vector PMV _enc and the estimated predicted motion vector PMV _dec are not equal to each other, It is determined that the optimum predicted motion vector PMV _enc is predictable using the vector PMV _dec , and otherwise, it is determined that prediction can not be performed.

Another embodiment of a method for determining whether the optimum predicted motion vector PMV _enc is predictable by the decoding apparatus 300 is as follows.

The reconstructed image data obtained by using the estimated predicted motion vector PMV _dec and the reconstructed image data obtained by using the optimal predicted motion vector PMV _enc are different from each other by a predetermined threshold value or less (For example, when the sum of absolute differences (SAD) between two reconstructed image data is less than a predetermined threshold value), the decoding apparatus 300 uses the estimated predicted motion vector PMV _dec the best prediction motion vector (PMV _enc) to and determined to be predictable, the control of the decoding device 300 the estimated predicted motion vector (PMV _dec) using the best predicted motion vector (PMV _enc) to be not predictable .

In the above-described steps S804 and S806, if the selected estimated predicted motion vector and the selected optimal predicted motion vector are equal to or less than a predetermined threshold value, It is possible to determine that the selected optimal predicted motion vector is predictable in the decoding apparatus 300. [

If the reconstructed image data using the selected estimated predictive motion vector is equal to or less than a predetermined threshold value in the reconstructed image data using the selected optimal predictive motion vector in steps S804 and S806, It may be determined that the selected optimal predicted motion vector is predictable by the decoding apparatus 300. [

8, the encoding apparatus 200 calculates a determination function value using a determination function (g () in Equation 6) defined beforehand between the encoding apparatus 200 and the decoding apparatus 300 (S800 And selects a candidate motion vector having a minimum decision function value as an estimated motion vector PMV _{dec in} step S802 based on the calculated determination function value. Thereafter, the estimated predicted motion vector PMV _dec and the optimal predicted motion vector PMV _enc are compared (step S804), and finally predicted (step S806).

The determination function defined beforehand between the encoding apparatus 200 and the decoding apparatus 300 can be realized in various forms.

A function using Template Matching (TM) or a function using boundary matching (BM) can be used as the determination function g () in Equation (6).

First, a determination function using Template Matching (TM) will be described below.

A Template Matching Set (TMS) can be defined as a set of indices indicating the relative positions of selected pixels based on the position of a given designated block. For example, the left side, the upper left side and the upper side And the positions of M surrounding pixels adjacent to the pixel. It goes without saying that other methods can be used as needed. In general, if the number of pixels indicated by the TMS is large, more accurate matching is possible, but the calculation amount is increased.

The template matching method includes a step of selecting all candidate predicted motion vector sets (CS) that can be selected as the predicted motion vectors, pixels corresponding to the reference blocks designated by the respective candidate motion vectors in the selected candidate predicted motion vector set, The difference between the pixels indicated by the TMS for the current block is calculated using Equation (7) (one embodiment of Equation (6)), and the matching error according to each candidate motion vector is calculated. (PMV _dec ) of the predicted motion vector.

(PMVC + DMV) in f (PMVC + DMV, i) represents the pixel position indicated by the index i around the reference block in the reference picture indicated by the index i (included in TMS) , i) denotes a pixel value at this position. C (i) denotes the pixel value around the current block indicated by the index i.

G (PMVC | DMV), which is an embodiment of the decision function, multiplies the difference vector DMV ^D provided by the encoding apparatus 200 by a candidate motion vector PMVC, which is an element of the candidate motion vector set CS, The motion vector of the corresponding block is calculated by PMVC + DMV, and motion compensation of the corresponding block using the motion vector is performed to provide a calculated value of how correct the reconstructed block is. In order to calculate this, Sum of Squared Error is used in Equation (7). However, other methods such as SAD (sum of absolute difference) can be used depending on the application. The estimated predicted motion vector PMV _dec means a candidate motion vector PMVC that minimizes g (PMVC | DMV) which is an embodiment of the determination function.

8, in step S800, the reference pixels designated by the template matching pixel index set (TMS) and the pixels designated by the template matching pixel index set (TMS) for the reference block designated by each of the one or more candidate motion vectors included in the selected candidate motion vector set, Calculating a pixel value difference of pixels indicated by the template matching pixel index for the block and calculating a matching error for each of the one or more candidate motion vectors as the determination function value based on the calculated pixel value difference, The judgment function may be performed to calculate the judgment function value.

Next, a determination function using boundary pixel matching (BM) will be described below.

Similar to the TMS of the above-described embodiment, the boundary matching pixel index set BMS (Boundary Matching Index Set) can be defined as a set of indexes indicating the positions of the pixel values located at the leftmost and topmost positions in the current block. It is needless to say that it may be defined as the position of all or some of the pixels located at the block boundary within the current block depending on the application

After selecting all the candidate motion vector sets CS selectable as the predictive motion vectors, it is necessary to determine which candidate motion vector PMVC of the candidate motion vector set CS is most optimal, In order to estimate how smoothly a block is connected to its surrounding blocks and block boundaries, edge pixel matching is performed to determine a candidate motion vector (PMVC) that minimizes the boundary pixel matching matching error as one of the candidate motion vector sets And selects it as the estimated predicted motion vector PMV _dec . And is referred to as a predictive motion vector selected by the decoding apparatus 300.

This process is described in more detail as follows. The value of the reference block designated by the candidate motion vector calculated by PMVC + DMV using the candidate motion vector PMVC, which is an element in the candidate motion vector set CS, and the difference vector DMV determined by the encoder 200, (Residual signal) calculated by the encoding device 200 to reconstruct the pixel value of the current block, and then reconstructs the motion-compensated reconstructed pixels around the current block from the motion-compensated reconstructed current block according to each candidate motion vector A prediction motion vector for generating a reconstruction block having the highest degree of boundary matching is determined. For this, a matching error of each candidate motion vector is calculated as a sum of squares of difference values as shown in Equation (8). Of course, it is obvious that other methods such as Sum of Absolute Difference (SAD) can calculate the matching error.

In Equation (8), C (i) is a candidate motion vector calculated by PMVC + DMV using a candidate motion vector PMVC that is an element of the candidate motion vector set CS and a DMV determined by the encoder 200 And a motion compensated pixel residual signal calculated by the encoding device 200 to obtain a pixel value designated by an index i in a boundary matching pixel index set BMS (Boundary Matching Set) among restored pixels of the restored current block do. Also, f (i) denotes a pixel value immediately adjacent to a pixel designated by index i of the BMS among boundary pixels in a neighboring block adjacent to the current block. A boundary pixel matching error is calculated by Equation (8) for each candidate motion vector PMVC in the candidate motion vector set CS using Equation (8), and a candidate motion vector for generating the minimum matching error is estimated Is selected as the predicted motion vector (PMV _dec ). And is referred to as a predictive motion vector selected by the decoding apparatus 300.

8, in operation S800, a difference vector determined by the encoder for each of the one or more candidate motion vectors included in the selected candidate motion vector set, and a reference block designated by the candidate motion vector calculated using the corresponding candidate motion vector And a motion compensation residual signal calculated by the encoding device, and adds the pixel value designated by the index in the boundary matching pixel index set (BMS) among the restored pixels of the restored current block; And calculating a matching error for each of the one or more candidate motion vectors as a determination function value based on a difference between a pixel designated by the index of the boundary matching pixel index set and a pixel value adjacent to the boundary pixel in the neighboring block adjacent to the current block , And the judgment function may be calculated by performing a predetermined judgment function.

Referring to FIG. 8, in step S504, in step S800 and step S802, the encoding apparatus 200 generates an estimated predicted motion vector PMV _dec with a minimum matching error determined according to a predetermined determination method such as template matching or border pixel matching. ).

In step S804, it is determined whether or not the selected estimated predicted motion vector PMV _dec is the same as the optimal predicted motion vector PMV _enc selected in step S502.

In step S806, if the estimated predicted motion vector PMV _dec and the best predicted motion vector PMV _enc are the same, the decoding apparatus 300 predicts the best predicted motion vector of the current motion vector of the current block And if not, it is judged that it can not be predicted. Also, as described above, even if the estimated predicted motion vector PMV _dec and the best predicted motion vector PMV _enc are not the same, if the restored reconstructed images are the same or the difference is smaller than the predetermined threshold value, It may be determined that the decoding apparatus 300 can predict the best predicted motion vector of the current motion vector of the current block by considering that the two estimated predicted motion vectors PMV _dec and the optimal predicted motion vector PMV _enc are the same.

9 is a flowchart illustrating a first encoding step S506 in the motion vector encoding method according to an embodiment of the present invention.

Referring to FIG. 9, a first coding step S506 in the motion vector coding method according to an embodiment of the present invention includes: determining (S900) a motion vector prediction mode indicating predictability; Determining a selected optimal predicted motion vector as a predicted motion vector for the current motion vector (S902); Calculating a difference vector between the current motion vector and the selected optimal predicted motion vector (S904); And coding the current motion vector by coding the calculated difference vector and the motion vector prediction mode indicating the predictability (S906).

10 is a flowchart of a second coding step S508 in the motion vector coding method according to an embodiment of the present invention.

Referring to FIG. 10, a second encoding step S508 in the motion vector coding method according to an exemplary embodiment of the present invention includes: a step (S1000) of determining a motion vector prediction mode indicating a non-predictable motion vector; Determining a predetermined default predicted motion vector as a predicted motion vector for a current motion vector (S1002); Calculating a difference vector between a current motion vector and a preset default predicted motion vector (S1004); And coding the current motion vector by coding the calculated difference vector and the motion vector prediction mode indicating the impossible (S1006).

The predetermined default predicted motion vector in step S1002 is a median of a motion vector of one or more neighboring blocks adjacent to the current block and is previously set between the encoding apparatus 200 and the decoding apparatus 300. [

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

Referring to FIG. 11, a motion vector decoding method according to an embodiment of the present invention includes decoding (S1100) motion vector prediction mode information of a current vector and a difference vector between a current motion vector and a predictive motion vector; Selecting a predicted motion vector of the current motion vector based on the motion vector prediction mode information (S1102); And reconstructing a current motion vector using the selected predictive motion vector (S1104).

In the step S1102, when the motion vector prediction mode indicated by the motion vector prediction mode information is the default prediction mode, the predetermined default predicted motion vector is selected.

In the case where the motion vector prediction mode indicated by the motion vector prediction mode information is the non-default prediction mode, in step S1102, one candidate motion vector among a plurality of candidate motion vectors is determined as a current motion vector As a predicted motion vector.

The predetermined method described above may be a method based on one of template matching and boundary pixel matching.

The motion vector decoding method according to an embodiment of the present invention described above with reference to FIG. 11 will be described in more detail with reference to FIG. 12 to FIG.

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

Referring to FIG. 12, a motion vector decoding method provided by a decoding apparatus 300 according to an embodiment of the present invention includes decoding a motion vector prediction mode and a coded differential vector coded by the coding apparatus 200 Step S1200; Determining whether the decoded motion vector prediction mode indicates predictability or impossibility of prediction (S1202); If the decoded motion vector prediction mode is predictable, a set of candidate motion vectors selectable as a prediction motion vector for restoring the current motion vector of the current block is selected, and an optimal predicted motion vector is selected from the selected candidate motion vector set (S1204) determining a selected optimal predicted motion vector as a predicted motion vector for the current motion vector; Determining a predetermined default predicted motion vector as a predicted motion vector for the current motion vector if the decoded motion vector prediction mode indicates that the predicted motion vector can not be predicted (S1206); And reconstructing the current motion vector of the current block by adding the predicted motion vector and the decoded difference vector determined in step 1204 or step 1206 (step 1208).

The coded difference vector in step S1200 may be a differential vector between the current motion vector and a predetermined default predicted motion vector or a difference vector between the current motion vector and the best predicted motion vector.

13 is a flowchart illustrating a first predicted motion vector determination step (S1204) in the motion vector decoding method according to an embodiment of the present invention.

Referring to FIG. 13, a first predictive motion vector selection step (S1204) in the motion vector decoding method according to an exemplary embodiment of the present invention includes a candidate motion Selecting a vector set (S1300); (S1302) calculating a decision function value for each of one or more candidate motion vectors included in the selected candidate motion vector set using a decision function commonly defined by the encoder and the decoder; And selecting a candidate motion vector having a minimum value among the calculated decision function values as an optimal predicted motion vector and selecting the selected optimal predicted motion vector as a predicted motion vector for the current motion vector (S1304).

In step S1304, the optimal predicted motion vector determined by the decoding apparatus 300 as the predicted motion vector for the current motion vector is estimated as the estimated predicted motion vector PMV _dec estimated by the decoding apparatus 300 to be found by the encoding apparatus 200, Or a difference of a predetermined boundary value.

In operation S1302, the template matching unit sets the pixels indicated by the template matching pixel index set (TMS) to the reference block designated by each of the one or more candidate motion vectors included in the selected candidate motion vector set, Calculating a difference between the pixel values of the pixels indicated by the matching pixel index and calculating a matching error for each of the one or more candidate motion vectors based on the calculated pixel value difference as a judgment function value, .

In step S1302, a difference between a value of a reference block designated by the candidate motion vector calculated using the difference vector determined by the encoder and the candidate motion vector for each of the one or more candidate motion vectors included in the selected candidate motion vector set, A pixel value designated by an index in a boundary matching pixel index set (BMS) among restored pixels of a restored current block by adding a motion compensation residual signal calculated by an encoding device; And calculating a matching error for each of the one or more candidate motion vectors as a determination function value based on a difference between a pixel designated by the index of the boundary matching pixel index set and a pixel value adjacent to the boundary pixel in the neighboring block adjacent to the current block , And a judgment function is performed to calculate a judgment function value.

It should be understood that the order of steps in the flowcharts shown in FIGS. 4 to 12 is not intended to be in the order of an embodiment for describing the essential characteristics of the present invention, and the order of each step may be changed within the scope of the present invention It can also be implemented.

According to an embodiment of the present invention, the motion vector prediction mode includes a default prediction mode and a non-default prediction mode. As an example of the default prediction mode, a motion vector prediction mode indicating the impossibility of prediction of the best predicted motion vector has been described, and a motion vector prediction mode indicating the predictability of the best predicted motion vector has been described as an example of the non-default prediction mode.

Here, the optimal predicted motion vector means a predicted motion vector selected from a finite number of predicted motion vector candidates according to a predetermined reference or method as described in the above-described embodiment.

According to another aspect of the present invention, the non-default prediction mode can be understood as a mode for selecting a predictive motion vector other than the default predictive motion vector according to a predetermined predetermined method, and then using the predictive motion vector for encoding and / or decoding the current motion vector. In this case, a predictive motion vector may be selected from a plurality of predictive motion vector candidates as described above by a predetermined method. According to the above-described motion vector encoding and decoding method of the present invention, So that the coding efficiency can be improved by selecting the motion vector that is closer to the value of the current motion vector as the prediction motion vector to minimize the amount of coding bits of the difference vector.

Also, according to the motion vector encoding and decoding method of the present invention, it is possible to improve the coding efficiency by selecting a more accurate predictive motion vector, and at the same time, For example, by sharing a function for giving or searching for motion vector prediction mode information, the coding efficiency and the decoding efficiency are further improved by reducing the increase in the amount of coding bits for additional information generated for informing the prediction motion vector .

In addition, when the motion vector encoding and decoding technique of the present invention is applied to a video service or a multimedia streaming service, the encoding technique according to the present invention enables encoding with a small bit amount, thereby providing a high-satisfaction service to the user. Especially, in a wireless mobile environment, which can have a relatively small bandwidth, large data loss and delay compared with a wired environment, a further greater effect can be expected.

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 a motion vector coding and decoding technique to select a motion vector that minimizes a difference between a current motion vector and a current motion vector as a predicted motion vector, Therefore, it is a very useful invention that has the effect of reducing coding bit amount and improving coding efficiency and decoding efficiency.

200: motion vector coding device
210: predictive motion vector selection unit
220: a motion vector prediction mode determining unit
230: first encoding unit
240: second encoding unit
300: motion vector decoding device
310: a motion vector prediction mode / differential vector decoding unit
320: a motion vector prediction mode determination unit
330: a first predicted motion vector determination unit
340: a second predicted motion vector determining unit
350: current motion vector restoration unit

Claims (4)

An image decoding method for restoring a motion vector of a current block,
Extracting, from a bitstream, mode information used for selecting one of a plurality of modes for determining a motion vector of the current block in a different manner;
Wherein if the mode information indicates a first mode for selecting a predictive motion vector for a motion vector of the current block from among a plurality of candidate motion vectors, Selecting a plurality of candidate motion vectors for the motion vector;
Selecting one candidate motion vector to be used as the predictive motion vector among the plurality of candidate motion vectors;
Setting the selected candidate motion vector to the predicted motion vector;
Decoding the differential vector from the bitstream; And
And restoring the motion vector of the current block using the set predictive motion vector and the decoded differential vector,
Wherein the plurality of neighboring blocks include at least one first block located in a first picture including the current block or at least one second block located in a second picture different from the first picture including the current block, Or a block including the first block and the second block,
Wherein the position of the second block in the second picture is determined based on the position of the current block in the first picture. delete The method according to claim 1,
Wherein the step of selecting one candidate motion vector to be used as the predictive motion vector among the plurality of candidate motion vectors comprises:
Wherein the one candidate motion vector is selected based on information included in the bitstream. The method according to claim 1,
If the decoded mode information indicates a second mode using a predetermined motion vector to recover a motion vector of the current block, restoring the motion vector of the current block using a predetermined motion vector And decodes the decoded image.

Images