KR101543310B1 - Motion Vector Coding Method and Apparatus by Using Partitioning Space and Video Coding Method and Apparatus Using Same - Google Patents

Abstract

The present invention relates to a motion vector coding / decoding method and apparatus using spatial division, and a method and apparatus for image coding / decoding using the same.
The present invention divides candidate prediction motion vectors of a current block into a plurality of groups including at least one candidate prediction motion vector and sequentially compares candidate prediction motion vectors in each group to generate one candidate prediction motion vector as a representative prediction motion vector And selects one representative predicted motion vector among the representative predicted motion vectors for each group as a predicted motion vector. The predicted motion vector is selected as a difference between the current motion vector of the current block and the predicted motion vector And a difference vector is encoded.
According to the present invention, it is possible to suppress the amount of bits generated by coding additional information to indicate which prediction motion vector is selected while decreasing the size of difference vector to be coded, thereby improving the coding efficiency of the motion vector The compression efficiency of the image can be improved.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a motion vector coding / decoding method and apparatus using spatial division, and a video coding / decoding method and apparatus using the same.

The present invention relates to a motion vector coding / decoding method and apparatus using spatial division, and a method and apparatus for image coding / decoding using the same. More particularly, the present invention relates to a method and apparatus for efficiently compressing a motion vector used for predictive encoding of an image to improve compression efficiency of an image.

In a conventional image compression technique such as H.264 / AVC, in order to predictively encode a motion vector obtained by estimating motion based on a block, a median of motion vectors of neighboring blocks of a block to be encoded is used to calculate a predicted motion vector And variable-length-codes the difference between the motion vector to be coded and the predicted motion vector to compress the motion vector.

Recently, the ITU-T Video Coding Expert Group (VCEG) has been studying a higher performance codec based on the existing H.264 / AVC with the name KTA (Key Technical Area). Of these, the Competition-based Motion Vector Coding ) To improve the existing H.264 / AVC motion vector coding scheme.

MVComp selects a candidate prediction motion vector having a minimum difference value between a current motion vector and a motion vector obtained after prediction with a plurality of candidate prediction motion vectors and transmits information about the candidate prediction motion vector selected to the decoder , Which improves the coding compression efficiency by about 5% compared to the existing H.264 / AVC. However, this technique has a problem that the amount of indexing side information to be transmitted to the decoder increases as the number of candidate prediction motion vectors increases.

Accordingly, a technique of selecting one motion vector, which is the most similar to the current motion vector, among a plurality of candidate prediction motion vectors, and sending additional information for discriminating whether or not the selected motion vector is an optimal motion vector . However, in the proposed method, even though a bit of the current frame is received without error, a decoder crash problem in which a frame before the current frame and the next intra frame can not be restored due to an error generated in the previous frame There is a problem that there is a limit to increase the amount of computation in the decoder.

Accordingly, techniques for determining a motion vector of a current block using a motion vector around a current block to be coded have been proposed. In this technique, the indexing side information is efficiently transmitted using the motion vector of the neighboring block, and the selection of the candidate motion vector is separated from the information of the previous frame to solve the decoder conflict problem. However, since a limited number of candidate motion vectors are used There is a problem that the compression performance is limited.

In order to solve the above-described problems, the present invention provides a method and apparatus for encoding a motion vector using a predicted motion vector more similar to a motion vector to be encoded, and efficiently compressing a motion vector by reducing the amount of encoded bits, There is a main purpose in making.

According to an aspect of the present invention, there is provided an apparatus for coding a motion vector, the apparatus comprising: a candidate prediction motion vector aggregator for combining candidate prediction motion vectors of a current block into a plurality of groups; A representative predictive motion vector selector for selecting a representative predictive motion vector for each group by selecting one candidate predictive motion vector among the candidate predictive motion vectors in each group as a representative predictive motion vector; A predictive motion vector selector for selecting one representative predictive motion vector among the representative predictive motion vectors per group as a predictive motion vector; And a differential vector encoder for coding a difference vector which is a difference between a current motion vector of the current block and a predictive motion vector to be selected.

According to another aspect of the present invention, there is provided an apparatus for encoding an image, the apparatus comprising: an image encoder for determining a current motion vector, which is a motion vector of a current block, and for predicting a current block using a current motion vector; And a representative predictive motion vector for each group is selected for a plurality of groups in which the candidate predictive motion vectors of the current block are grouped based on the spatial position, and one representative predictive motion vector among the representative predictive motion vectors for each group is selected as a predictive motion vector And a motion vector encoder for generating motion vector data by coding a group index indicating a group including a difference vector which is a difference between a current motion vector to be determined and a selected predicted motion vector and a representative predictive motion vector to be selected The present invention provides a video encoding apparatus characterized by:

According to another aspect of the present invention, there is provided an apparatus for decoding a motion vector, the apparatus comprising: a candidate prediction motion vector aggregator for combining candidate prediction motion vectors of a current block into a plurality of groups; A representative predictive motion vector selector for selecting a representative predictive motion vector for each group by selecting one candidate predictive motion vector among the candidate predictive motion vectors in each group as a representative predictive motion vector; A predictive motion vector restoration unit for restoring a group index from the group index data extracted from the motion vector data and selecting a representative predictive motion vector of the group identified by the group index to be restored from among the representative predictive motion vectors for each group, group; And a current motion vector reconstructor for reconstructing the differential vector extracted from the motion vector data to reconstruct the differential vector and reconstructing the reconstructed differential vector and the selected predictive motion vector as the current motion vector of the current block. A motion vector decoding apparatus is provided.

According to another aspect of the present invention, there is provided an apparatus for decoding an image, the apparatus comprising: decoding motion vector data extracted from a bitstream to restore a difference vector and a group index; Selects a representative predictive motion vector for each group to be grouped based on the group, selects a representative predictive motion vector of the group identified by the group index to be restored from among the representative predictive motion vectors for each group as a predictive motion vector, A motion vector decoder for adding a difference vector and a selected motion vector to reconstruct a current motion vector of a current block; And an image decoder that predictively decodes the image data extracted from the bitstream using the recovered current motion vector to recover the current block.

According to another aspect of the present invention, there is provided a method of coding a motion vector, the method comprising: aggregating candidate prediction motion vectors of a current block into a plurality of groups; Selecting one of the candidate prediction motion vectors in each group as a representative prediction motion vector and selecting a representative prediction motion vector per group; Selecting one representative predictive motion vector among representative predictive motion vectors per group as a predictive motion vector; And encoding a differential vector which is a difference between a current motion vector of a current block and a predictive motion vector to be selected.

According to another aspect of the present invention, there is provided a method of coding an image, the method comprising: determining a current motion vector, which is a motion vector of a current block; Predictively encoding a current block using a determined current motion vector; Selecting a group-wise representative predicted motion vector for a plurality of groups in which candidate prediction motion vectors of the current block are grouped based on spatial positions; Selecting one representative predictive motion vector among representative predictive motion vectors per group as a predictive motion vector; And encoding a group index indicating a group including a difference vector which is a difference between a current motion vector to be determined and a selected predicted motion vector and a representative predictive motion vector to be selected. .

According to another aspect of the present invention, there is provided a method of decoding a motion vector, the method comprising: decoding differential vector data and group index data extracted from motion vector data to reconstruct a difference vector and a group index; Collecting candidate prediction motion vectors of a current block into a plurality of groups; Selecting one of the candidate prediction motion vectors in each group as a representative prediction motion vector and selecting a representative prediction motion vector per group; Selecting a representative predictive motion vector of the group identified by the restored group index among the representative predictive motion vectors of each group as a predictive motion vector; And reconstructing a current motion vector of a current block by adding a differential vector to be reconstructed and a predictive motion vector to be selected.

According to another aspect of the present invention, there is provided a method of decoding an image, comprising: decoding motion vector data extracted from a bitstream to restore a difference vector and a group index; Selecting a group-wise representative predicted motion vector for a plurality of groups in which candidate prediction motion vectors of the current block are grouped based on spatial positions; Selecting a representative predictive motion vector of the group identified by the restored group index among the representative predictive motion vectors per group as a predictive motion vector; Reconstructing a current motion vector of a current block by adding a reconstructed difference vector and a selected predictive motion vector; And reconstructing a current block by predicting and decoding the image data extracted from the bitstream using a current motion vector to be reconstructed.

As described above, according to the present invention, by selecting a predictive motion vector using a large number of candidate predictive motion vectors and reducing the size of a differential vector to be encoded, additional information for indicating which predictive motion vector has been selected is encoded The amount of bits generated can be suppressed, so that the coding efficiency of a motion vector can be improved and ultimately the compression efficiency of an image can be improved.

1 is a block diagram schematically illustrating an image encoding apparatus according to an embodiment of the present invention.
2 is a block diagram schematically showing a motion vector coding apparatus according to an embodiment of the present invention.
FIGS. 3 to 5 are views illustrating a process of grouping candidate predictive motion vectors into a plurality of groups based on spatial positions according to an embodiment of the present invention; FIG.
FIG. 6 is a diagram illustrating a process of selecting representative predictive motion vectors for each group according to an embodiment of the present invention. FIG.
FIG. 7 is a flowchart illustrating a motion vector coding method according to an embodiment of the present invention. FIG.
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 showing an image decoding apparatus according to an embodiment of the present invention.
10 is a block diagram schematically showing a motion vector decoding apparatus according to an embodiment of the present invention.
11 is a flowchart for explaining a motion vector decoding method according to an embodiment of the present invention,
12 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 symbols as 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;

A motion vector encoding apparatus, a motion vector decoding apparatus, a video encoding apparatus, and a video decoding apparatus, which will be described later, may be implemented as a personal computer (PC) Such as a portable computer, a computer, a notebook computer, a personal digital assistant (PDA), a portable multimedia player (PMP), a PlayStation Portable (PSP), a mobile communication terminal, Or a server terminal such as an application server and a service server, and may be a communication device such as a communication modem for performing communication with various devices or wired / wireless communication networks, various programs and data for coding or decoding a motion vector, A memory for storing, a program is executed And a microprocessor for controlling and controlling the microprocessor.

The motion vector or the motion vector encoded in the bit stream by the motion vector coding apparatus or the video coding apparatus can be transmitted in real time or in non-real time through a wired / wireless communication network such as the Internet, a local area wireless communication network, a wireless LAN network, a WiBro network, , A universal serial bus (USB), or the like, and is transmitted to a motion vector decoding apparatus or an image decoding apparatus, decoded by a motion vector decoding apparatus and restored as a motion vector, or decoded by a video decoding apparatus, And can be reproduced.

Usually, a moving picture is composed of a series of pictures, and each picture is divided into a predetermined area such as a block. In the case where an image area is divided into blocks, the divided blocks are classified into an intra block and an inter block according to a coding method. The intra-block refers to a block that is coded using Intra Prediction Coding (P-Coding) scheme. The intra-prediction coding is performed by using the pixels of previously decoded and decoded blocks in the current picture, A prediction block is generated by predicting a pixel of a current block which is a block to be encoded and a difference value between the pixel of the current block and the current block is encoded. Inter-block refers to a block that is coded using Inter Prediction Coding. Inter-prediction coding refers to one or more past pictures or a future picture to generate a prediction block by predicting a current block in the current picture, And the difference value is encoded. Here, a picture referred to in encoding or decoding a current picture is referred to as a reference picture.

1 is a block diagram schematically illustrating an image encoding apparatus according to an embodiment of the present invention.

The image encoding apparatus 100 according to an embodiment of the present invention is an apparatus for encoding an image and may include a motion vector encoder 110 and a video encoder 120.

The motion vector encoder 110 generates a representative predictive motion vector classified by a group for a plurality of groups in which Candidate Predicted Motion Vector of the current block is aggregated based on the spatial position. Selects a representative predictive motion vector from among representative predictive motion vectors for each group as a predictive motion vector, and selects a current motion vector and a selected predictive motion vector, A group index indicating a group including a differential motion vector that is a difference between motion vectors and one representative predictive motion vector that is selected is encoded. The difference vector is encoded and generated as difference vector data, and the group index is encoded and generated as group index data.

Accordingly, the motion vector encoder 110 generates motion vector data including differential vector data and group index data. Here, in generating the difference vector, the motion vector encoder 110 uses the current motion vector determined to predictively encode the current block in the image encoder 120. The motion vector encoder 110 will be described in detail with reference to FIG. 2 through FIG. 6 in the following description.

The image encoder 120 determines the current motion vector, which is a motion vector of the current block, and predictively encodes the current block using the current motion vector. Thus, the current block is predictively encoded and image data is generated.

For this, the image encoder 120 may include a Predictor, a subtracter, a transformer and a quantizer, and an encoder. The image encoder 120 may include an inverse transformer and an inverse quantizer, an adder, a deblocking filter, a memory, and the like. Here, the predictor estimates the motion of the current block to determine a current motion vector, generates a predicted block by compensating the motion of the current block using the current motion vector, and the subtractor subtracts the current block from the prediction block The transformer and the quantizer transform and quantize the residual block to generate a quantized transform coefficient. The encoder generates the image data by encoding the quantized transform coefficient. The dequantizer and the inverse transformer dequantize and inverse transform the quantized transform coefficients to restore the residual block, the adder restores the prediction block and the restored residual block to restore the current block, and the restored current block is deblocked Is deblocked and filtered by a filter (Deblocking Filter), accumulated in a memory on a picture-by-picture basis, and stored as a reference picture to be used for predicting a next block or a next picture.

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

The motion vector coding apparatus according to an embodiment of the present invention can be implemented in the motion vector coding unit 110 in the image coding apparatus 100 described above with reference to FIG. Hereinafter, a motion vector coding apparatus according to an embodiment of the present invention will be referred to as a motion vector coding unit 110 for convenience of explanation.

The motion vector encoder 110 includes a Candidate Predicted Motion Vector Aggregator 210, a Representative Predicted Motion Vector Selector 220, a Predicted Motion Vector Selector 230, A Differential Motion Vector Encoder 240, and a Predicted Motion Vector Encoder 250.

The candidate prediction motion vector superimposer 210 divides the candidate prediction motion vectors of the current block into a plurality of groups including at least one candidate prediction motion vector.

Here, the candidate predicted motion vectors of the current block are candidates likely to be determined as the predicted motion vector of the current block. These candidate predictive motion vectors may be, for example, motion vectors such as MV _{H .264} , MV _extspa , MV _a , MV _b , MV _c , MV _d , MV _col , MV ₀ , MV _{H .264} is a motion vector used as a predictive motion vector in the H.264 / AVC compression standard, and refers to a motion vector having a median of a motion vector of a neighboring block of the current block. MV _extspa refers to a motion vector used as a predicted motion vector according to the presence or absence of neighboring blocks of the current block. MV _a , MV _b , MV _c , and MV _d are motion vectors of neighboring blocks of the current block. MV _col is a motion vector of a block located at the same position as the current block in the reference picture. MV ₀ is a zero vector which is a motion vector of (0, 0).

For example, assuming that the left block of the current block is block a, the upper block is block b, the upper right block is block c, and the upper left block is block d, the motion vector of block a is MV _a and the motion vector of block b is a motion vector of MV _b, c block is a motion vector of MV _c, d is a block MV _{d. H .264} MV MV is _a, can be calculated as the median value of the MV _{b, c} MV, MV _extspa if possible both using the current block around the block a, the block b, c block has an MV and _H.264 If one or two of the blocks _a , _b , and _c are not available, MV _a , MV _b , and MV _{c become} available motion vectors, and block a, block b, If it is not possible, it becomes MV ₀ .

However, the candidate prediction motion vectors of the current block are only examples for convenience of explanation. The candidate prediction motion vectors of the current block used in the embodiment of the present invention are not limited to the above-described example, Only a part of the above-described example may be used as candidate predictive motion vectors, and various other candidate predictive motion vectors may be added to the motion vector decoding apparatus or the image encoding apparatus, which will be described later, And all of the above examples are not used and other candidate predictive motion vectors may be used.

The candidate predictive motion vector convolution module 210 may group the motion vector candidates into a plurality of groups based on the spatial positions of the candidate predicted motion vectors of the current block. That is, the candidate prediction motion vector superimposer 210 suitably groups the candidate prediction motion vectors into a plurality of groups in consideration of the spatial position of the candidate prediction motion vectors of the current block. At this time, the candidate predictive motion vector superimposer 210 may arbitrarily select the number of groups for grouping the candidate predictive motion vectors, and may determine the number of groups predicted by the motion vector coding apparatus, the motion vector decoding apparatus, . If the number of groups is selected in advance, the information on the number of groups does not have to be coded to indicate how many candidate prediction motion vectors are grouped. However, if the number of groups is an arbitrary number The information on the number of groups is encoded and inserted into a slice header, a picture header, a sequence header, or the like of a bitstream.

The candidate prediction motion vector superimposer 210 may group the candidate prediction motion vectors of the current block into a plurality of groups based on their spatial positions using the K-Means algorithm. That is, the candidate prediction motion vector superimposer 210 sets a plurality of candidate prediction motion vectors among the candidate prediction motion vectors of the current block as a center value of a plurality of groups, and a spatial distance from the center value The candidate prediction motion vectors that fall below the distance threshold value are collected into each group and the center value is reset so that the average of the squares of the distances between the candidate prediction motion vectors to be collected for each group is less than or equal to the average threshold, Grouping and grouping the candidate prediction motion vectors of the current block into a plurality of groups by repeating the process of re-centering the groups and resetting the center value until the average of the squares of the distances between the candidate prediction motion vectors collected for each group becomes equal to or less than the average threshold value .

FIGS. 3 to 5 are diagrams illustrating a process of grouping candidate prediction motion vectors into a plurality of groups based on spatial positions according to an embodiment of the present invention.

3 to 5, when the candidate predicted motion vectors of the current block are set to MV _{H .264} , MV _extspa , MV _a , MV _b , MV _c , MV _d , MV _col , MV ₀ , The motion vector of the current block The candidate predictive motion vectors are divided into a plurality of groups based on their spatial positions.

In FIG. 3, candidate prediction motion vectors of the current block are exemplarily shown on the X-Y plane in consideration of their spatial positions. The candidate prediction motion vector superimposer 210 arbitrarily selects a plurality of candidate prediction motion vectors among the candidate prediction motion vectors shown in FIG. 3 and sets the center values of the plurality of groups.

In FIG. 4, four candidate prediction motion vectors among the candidate prediction motion vectors shown in FIG. 3 are set as the center values of four groups. The four groups are represented as group A, group B, group C, and group D. In group A, it is assumed that MV _d is set as the initial center value. In group B, MV _{H .264} is set as the initial center value In the group C, it is assumed that MV _col is set as the initial center value. In the group D, MV _extspa is assumed as the initial center value.

When the initial center value of the four groups is set as shown in FIG. 4, the candidate prediction motion vector superimposer 210 calculates a spatial distance between the center value of each group and the candidate prediction motion vectors of the current block, A candidate predicted motion vector whose distance is less than or equal to a predetermined distance threshold value is selected and is grouped into the corresponding group. In Figure 4 is set to the center value of the group A _d and an initial MV of the current block, the candidate prediction motion vector, which are _H.264 MV, MV _extspa, MV _a, MV _{b, c} MV, MV _{d, col} MV, MV ₀ each and a distance calculation, the distance is set to a predetermined distance threshold value or less, the candidate prediction motion vector MV and _a center value is calculated candidate predicted motion vector MV _d were mobilized in group a. Likewise, the distance between the center value of each group and the candidate predicted motion vectors is calculated in the same manner for the group B, the group C, and the group D in the same manner, and the candidate predicted motion vector in which the calculated distance is less than the predetermined distance threshold value MV _a and MV _H.264 were collected in group B, MV _col was _{collected in} group C, and MV _b , MV _extspa and MV ₀ were _{collected in} group D, respectively.

In FIG. 5, as shown in FIG. 4, a center value is reset for a group that is centered around an initial center value.

When a group is formed centering on the initial center value, the candidate predictive motion vector superimposer 210 calculates the distance between the candidate predictive motion vectors in each group for each group in which the candidate motion vectors are grouped based on the initial center value The center value is reset so that the average of the squares of < EMI ID = 1.0 >

In FIG. 5, the center value is reset so that the distance between MV _a and MV _d, which are candidate predicted motion vectors in the group A, may be less than a preset average threshold value, and MV _a and MV _d A new center value has been set. Likewise, a new center value is set for group B, group C, and group D as shown in FIG.

Thereafter, the candidate motion vector superimposer 210 calculates a spatial distance between the reset center value of each group and the candidate prediction motion vectors of the current block, and outputs a candidate prediction motion vector whose calculated distance is less than a preset distance threshold value And repeats the process of resetting the center value so that the average of the squares of the distances between the candidate predicted motion vectors in each group becomes equal to or less than the preset average threshold value. In this process, the candidate motion vectors to be grouped in each group can be changed, and thus, the average of the squares of the distances between the candidate predicted motion vectors in each group is converged to be less than a predetermined average threshold value. If the average of the squares of the distances between the initial center values or the candidate predicted motion vectors in the group grouped around the reset center value is already equal to or less than the average threshold value, the candidate motion vector summarizer 210 resets the center value The set center value is determined as the final center value, and the group formed by the final center value is determined as the final group. FIG. 5 shows a case where the reset center value is the final center value.

Referring again to FIG. 2, the representative predictive motion vector selector 220 sequentially compares the candidate predictive motion vectors in each group, which have been collected by the candidate motion vector summarizer 210, And selects a group-wise representative predicted motion vector. 6, the representative predictive motion vector selector 220 selects candidate candidate motion vectors in the group among the candidate predictive motion vectors in the group for each group collected by the candidate predictive motion vector superimposer 210 It is possible to select a group-wise representative predicted motion vector by selecting a candidate predicted motion vector having the closest spatial distance from the center value of the predicted motion vector.

6 is a diagram illustrating a process of selecting representative predictive motion vectors per group according to an embodiment of the present invention.

FIG. 6 exemplarily shows a representative predictive motion vector selected for each group when candidate predictive motion vectors are grouped into four groups and a final center value is set as shown in FIG.

As shown in FIG. 6, in the group A, MV _d is selected as a representative predicted motion vector since MV _d is spatially close to the position of the final center value than MV _c . Since the group B in the MV _{H .264} is close in space at the position of the final center than _a MV is selected as a representative MV _{H. 264} predicted motion vector. In group C, MV _col is the final center value, so MV _col is selected as the representative predicted motion vector. In the group D, MV _b is selected as the representative predictive motion vector because MV _b is closest in terms of the spatial center value as compared with the other motion vectors MV _extspa and MV ₀ .

Referring again to FIG. 2, the predictive motion vector selector 230 selects one representative predictive motion vector among the representative predictive motion vectors per group as a predictive motion vector. Here, the predictive motion vector selector 230 can select one representative predictive motion vector among the representative predictive motion vectors per group using the rate-distortion cost. That is, the predictive motion vector selector 230 calculates a rate-distortion cost generated by coding the current block when using each representative predictive motion vector for each group, and calculates a predictive motion vector of the group having the smallest rate- The vector can be selected as the predictive motion vector of the current block. However, the rate-distortion cost is only an example of a criterion that can be used to select a predictive motion vector. In an embodiment of the present invention, the predictive motion vector And a prediction motion vector may be selected using various coding costs.

The difference vector encoder 240 encodes a differential vector which is a difference between the current motion vector of the current block and the selected predictive motion vector. That is, the difference vector encoder 240 subtracts the current motion vector determined by the image encoder 120 and the predicted motion vector selected by the predictive motion vector selector 230 to obtain a difference vector, . However, the difference vector encoder 240 does not separately encode the difference vector. When the predictive motion vector selector 230 obtains the difference vector to obtain the rate-distortion cost and encodes the difference vector, the difference vector encoder 240 It is possible to output vector data. Entropy coding schemes such as Fixed Length Coding, Variable Length Coding, and Arithmetic Coding can be used as a technique for coding difference vectors.

The predictive motion vector encoder 250 encodes a group index indicating a group including one representative predictive motion vector selected as a predictive motion vector. That is, the predictive motion vector encoder 250 checks whether the predictive motion vector selected by the predictive motion vector selector 230 is a group of representative predictive motion vectors, encodes the group index indicating the corresponding group, . Group Index The group index is an index capable of identifying a group such as a group A, a group B, a group C, a group D, or a first group, a second group, a third group, a fourth group, Entropy coding techniques such as fixed length coding, variable length coding, and arithmetic coding can be used.

For example, when the candidate predictive motion vectors are grouped into four groups of group A, group B, group C, and group D, if the group index is coded using fixed length coding, the group index data for group A is' 00 ', the group index data for the group B is encoded as' 01', the group index data for the group C is encoded as' 10 ', and the group index data for the group D is encoded as' 11' .

Meanwhile, as described above, the predictive motion vector selector 230 may select one representative predictive motion vector among the representative predictive motion vectors per group as a predictive motion vector. However, when the spatial distance between the plurality of groups is greater than a predetermined reference distance If the reference distance is less than or equal to the reference distance, the intermediate value of the candidate prediction motion vectors of the current block is selected as the prediction motion vector. If only the reference motion vector is larger than the reference distance, A predictive motion vector may be selected as a predictive motion vector.

If the predictive motion vector selector 230 selects the intermediate value of the candidate predictive motion vectors of the current block as a predictive motion vector because the spatial distance between the plurality of groups is less than the preset reference distance, May not encode the group index. This is because the motion vector decoding apparatus or the video decoding apparatus is also used for grouping the candidate prediction motion vectors of the current block into a plurality of groups as in the motion vector coding apparatus or the video coding apparatus and determining whether the spatial distance between the plurality of groups is equal to or less than a preset reference distance As shown in FIG.

FIG. 7 is a flowchart illustrating a motion vector coding method according to an embodiment of the present invention. Referring to FIG.

According to the motion vector coding method according to the embodiment of the present invention, the motion vector encoder 110 combines the candidate prediction motion vectors of the current block into a plurality of groups (S710) In operation S720, one representative predictive motion vector is selected as a representative predictive motion vector to select a representative predictive motion vector for each group, A differential vector which is a difference between the current motion vector of the current block and the selected predictive motion vector is encoded (740). In addition, the motion vector encoder 110 may further encode a group index indicating a group including one representative predictive motion vector selected as a predictive motion vector as well as a difference vector.

In step S710, the motion vector encoder 110 may group the motion vectors into a plurality of groups based on the spatial positions of the candidate prediction motion vectors of the current block. For example, a plurality of candidate prediction motion vectors A candidate prediction motion vector having a spatial distance from a center value to a center value is less than or equal to a distance threshold value is grouped into each group, and a candidate prediction motion The center value is reset so that the average of the squares of the distances between the vectors is equal to or less than the average threshold value, and the process of collecting into each group and resetting the center value is performed by multiplying the square of the distance between candidate predicted motion vectors Until the average of the predicted motion of the current block becomes equal to or less than the average threshold value Vector can be mobilized into the plurality of groups.

In step S720, the motion vector encoder 110 selects a candidate prediction motion vector having the closest spatial distance from the center value of the candidate prediction motion vectors in the group among the candidate prediction motion vectors in the group for each group, A predictive motion vector can be selected.

In step S730, the motion vector encoder 110 may select one representative predictive motion vector among representative predictive motion vectors per group using the rate-distortion cost.

In addition, if the spatial distance between the plurality of groups is less than or equal to a preset reference distance, the motion vector encoder 110 may select a middle value of candidate prediction motion vectors of the current block as a prediction motion vector. In this case, the motion vector encoder 110 does not encode the group index.

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

According to the image encoding method of the present invention, the image encoding apparatus 100 determines a current motion vector, which is a motion vector of a current block (S810), and predictively encodes the current block using the determined current motion vector (S820), a representative predictive motion vector for each group is selected for a group of candidate predictive motion vectors of the current block based on the spatial position (S830), and one representative predictive motion A group index indicating a group including a difference vector which is a difference between a current motion vector to be determined and a predictive motion vector to be selected and a representative predictive motion vector to be selected is encoded in operation S850. .

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

The video decoding apparatus 900 according to an exemplary embodiment of the present invention may include a motion vector decoder 910 and a video decoder 920. [

The motion vector decoder 910 decodes the motion vector data extracted from the bitstream to reconstruct a difference vector and a group index, and generates a motion vector for each of a plurality of groups of candidate prediction motion vectors of the current block, Selects a representative predictive motion vector, selects a representative predictive motion vector of the group identified by the group index to be restored from the representative predictive motion vectors of each group as a predictive motion vector, adds the restored differential vector and the selected predictive motion vector And restores the current motion vector of the current block. The motion vector decoder 910 will be described in detail with reference to FIG. 10 in the following description.

The image decoder 920 predictively decodes the image data extracted from the bitstream using the recovered current motion vector to recover the current block. For this, the image decoder 920 may include a decoder, an inverse quantizer and an inverse transformer, a predictor, an adder, a deblocking filter, a memory, and the like. Here, the decoder decompresses the image data extracted from the bitstream to restore the quantized transform coefficients. The dequantizer and the inverse transformer inverse quantize and inverse transform the quantized transform coefficients to be restored, thereby restoring the residual block. The motion vector decoder 910 generates a prediction block by compensating for the motion of the current block using the current motion vector of the current block restored by the motion vector decoder 910. The adder adds the residual block to be restored and the prediction block to restore the current block . The reconstructed current block is deblocked and filtered by the deblocking filter, accumulated in units of pictures, and output as a reconstructed image or stored in memory so that the predictor can be used to predict the next block or next picture.

FIG. 10 is a block diagram schematically illustrating a motion vector decoding apparatus according to an embodiment of the present invention. Referring to FIG.

The motion vector decoding apparatus according to an embodiment of the present invention may be implemented by a motion vector decoder 910 in the video decoding apparatus 900 described above with reference to FIG. Hereinafter, a motion vector decoding apparatus according to an embodiment of the present invention will be referred to as a motion vector decoder 910 for convenience of explanation.

The motion vector decoder 910 includes a candidate prediction motion vector combiner 1010, a representative predictive motion vector selector 1020, a predictive motion vector reconstructor 1030, and a current motion vector reconstructor 1030. [ Reconstructor, 1040).

The predictive motion vector combiner 1010 combines the candidate predictive motion vectors of the current block into a plurality of groups, and the representative predictive motion vector selector 1020 selects one of the candidate predictive motion vectors in each group And selects a representative predictive motion vector for each group as a representative predictive motion vector. The predictive motion vector combiner 1010 and the representative predictive motion vector selector 1020 are the same as or similar to the predictive motion vector superimposer 210 and the representative predictive motion vector selector 220 described above with reference to FIG. 2, The description is omitted.

The predictive motion vector reconstructor 1030 reconstructs a group index by decoding the group index data extracted from the motion vector data, and predicts a representative predictive motion vector of the group identified by the group index to be restored in the representative predictive motion vectors of each group As a motion vector. For example, when the representative predictive motion vector per group is selected by the representative predictive motion vector selector 1020 as shown in FIG. 6 and the group identified by the group index to be decoded by decoding the group index data '11' is the group D , The predictive motion vector reconstructor 1030 restores the MV _b selected as the representative predictive motion vector in the group D as a predictive motion vector.

The current motion vector reconstructor 1040 decodes difference vector data extracted from the motion vector data to restore a difference vector, and restores the reconstructed difference vector and the selected predicted motion vector as the current motion vector of the current block. For example, if the difference vector recovered by decoding the differential vector data is MV ₀ and the predicted motion vector restored by the predictive motion vector restorer 1030 is MV _b as in the above example, MV _b is the current motion vector .

Meanwhile, the predictive motion vector reconstructor 1030 may select a representative predictive motion vector of the group identified by the group index to be restored from among the representative predictive motion vectors of each group, as described above, The intermediate value of the motion vectors may be selected as the predicted motion vector. That is, the predictive motion vector reconstructor 1030 determines whether or not the group index data is extracted from the motion vector data. If the group index data is extracted, And selects the representative predictive motion vector of the group identified by the predictive motion vector as a predictive motion vector. If the group index data is not extracted, the intermediate value of the candidate predictive motion vectors of the current block can be selected as a predictive motion vector.

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

According to the motion vector decoding method of the embodiment of the present invention, the motion vector decoder 910 decodes the difference vector data and the group index data extracted from the motion vector data to recover the difference vector and the group index (S1110) , The candidate predictive motion vectors of the current block are grouped into a plurality of groups (S1120), one candidate predictive motion vector among the candidate predictive motion vectors in each group is selected as a representative predictive motion vector, (S1130), selects a representative predicted motion vector of the group identified by the group index to be restored from among the representative predicted motion vectors of each group as a predicted motion vector (S1140), adds the recovered difference vector and the selected predicted motion vector And restores the current motion vector of the current block (S1150).

In step S1120, the motion vector decoder 910 may group the motion vectors into a plurality of groups based on the spatial positions of the candidate prediction motion vectors of the current block. For example, the motion vector decoder 910 sets a plurality of candidate prediction motion vectors among the candidate prediction motion vectors of the current block as a center value of a plurality of groups, and a spatial distance from the center value The candidate predictive motion vectors that fall below the distance threshold value are grouped into each group and the center value is reset so that the average of the squares of the distances between the candidate predictive motion vectors to be collected for each group is less than or equal to the average threshold value, And the process of resetting the center value are repeated until the average of the squares of the distances between the candidate predictive motion vectors collected for each group becomes equal to or less than the average threshold value so that the candidate predictive motion vectors of the current block are divided into a plurality of Group.

In step S1130, the motion vector decoder 910 selects a candidate prediction motion vector having the closest spatial distance from the center value of the candidate prediction motion vectors in the group among the candidate prediction motion vectors in the group for each group, The representative predictive motion vector can be selected.

Also, the motion vector decoder 910 determines whether group index data is extracted from the motion vector data. If the group index data is not extracted, the motion vector decoder 910 selects the intermediate value of the candidate prediction motion vectors of the current block as a prediction motion vector .

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

According to an image decoding method according to an embodiment of the present invention, the image decoding apparatus 900 decodes motion vector data extracted from a bitstream to recover a difference vector and a group index (S1210) The representative predictive motion vectors of the group identified by the group index to be restored from among the representative predictive motion vectors of each group are predicted (step S1220) (S1230). The current motion vector of the current block is reconstructed by adding the reconstructed difference vector and the predictive motion vector to be reconstructed (S1240). The reconstructed image data from the bitstream is reconstructed using the reconstructed current motion vector And reconstructs the current block (S1250).

As described above, according to an embodiment of the present invention, candidate prediction motion vectors of a block to be coded are divided into a plurality of groups in consideration of spatial positions, and a representative predictive motion vector is selected for each group A predictive motion vector is selected using a large number of candidate predictive motion vectors by selecting a representative predictive motion vector optimal in terms of rate-distortion, as a predictive motion vector, and even if a predictive motion vector is predictively encoded using the selected predictive motion vector, It is possible to improve the coding efficiency of the motion vector and ultimately improve the compression efficiency of the image.

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 compression processing field for encoding and decoding an image, so that a prediction motion vector is selected using a large number of candidate prediction motion vectors, It is possible to suppress the amount of bits generated by encoding the additional information for indicating whether the vector is selected. Therefore, it is very useful to improve the coding efficiency of the motion vector and ultimately to improve the compression efficiency of the image Invention.

Claims (5)

An apparatus for coding a motion vector of a current block,
Selecting a first motion vector in a first motion vector among a plurality of motion vectors corresponding to a block located in the periphery of the current block and selecting a second motion vector in a second motion vector group among the plurality of motion vectors A motion vector selector;
A predicted motion vector selector for selecting one of the candidate motion vector sets including the first and second motion vectors as a predicted motion vector of the current block;
A differential vector encoder for coding a difference vector which is a difference between the motion vector of the current block and the predicted motion vector; And
A predictive motion information encoder for encoding an index identifying the one motion vector selected from the set of candidate motion vectors including the first and second motion vectors,
Lt; / RTI >
Wherein the first motion vector group includes a motion vector of a left block of the current block and the second motion vector group includes a motion vector of an upper block of the current block. The method according to claim 1,
Wherein the motion vector selector sequentially selects motion vectors included in the second motion vector group to select the second motion vector. A method of coding a motion vector of a current block,
Selecting a first motion vector in a first motion vector among a plurality of motion vectors corresponding to a block located in the periphery of the current block and selecting a second motion vector in a second motion vector group among the plurality of motion vectors ;
Determining one motion vector among a set of candidate motion vectors including the first and second motion vectors as a predicted motion vector of the current block; And
Encoding a difference vector which is a difference between the motion vector of the current block and the predicted motion vector; And
Encoding an index identifying a motion vector selected as the predicted motion vector from a set of candidate motion vectors including the first and second motion vectors;
Lt; / RTI >
Wherein the first motion vector group includes a motion vector of a left block of the current block and the second motion vector group includes a motion vector of an upper block of the current block. 4. The method of claim 3, wherein selecting the second motion vector comprises:
Wherein the second motion vector is selected by sequentially checking the motion vectors included in the second motion vector group. 4. The method of claim 3, wherein determining as the predicted motion vector comprises:
Wherein the one motion vector is determined from a set of candidate motion vectors including the first and second motion vectors using a coding cost for the current block.

Images