KR101782153B1 - Method for selecting motion vector candidate and method for encoding/decoding image using the same - Google Patents

Abstract

A motion vector candidate selection method for selecting a prediction candidate from reference blocks of a reference picture including a current picture to derive motion information for a current block to be encoded or decoded in an image, and an image encoding / decoding method using the same. A motion vector candidate selection method includes: constructing a spatial motion vector candidate; determining whether a reference picture of a current block exists in a current picture; and if the determination result of the determining step is YES, And adding a spatial motion vector candidate in another block of the picture.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a motion vector candidate selection method and a video encoding / decoding method using the motion vector candidate selection method.

The present invention relates to an image encoding / decoding technique, and more particularly, to a motion vector decoding method and a motion vector decoding method for decoding a motion vector for selecting a prediction candidate from reference blocks of a reference picture including a current picture, Candidate selection method and a video encoding / decoding method using the same.

With the spread of the Internet and mobile terminals and the development of information and communication technology, the use of multimedia data is increasing rapidly. Therefore, in order to perform various services or tasks through image prediction in various systems, there is a great need for improving the performance and efficiency of the image processing system.

Meanwhile, in the conventional image encoding / decoding technique, motion information on neighboring blocks of the current block is predicted in at least one reference picture before or after the current picture according to the inter picture prediction method, The motion information for the current block is estimated by acquiring the motion information from the reference block.

However, existing inter-picture prediction has a drawback in that the calculation complexity is high because a prediction block is generated using a temporal prediction mode between pictures, and in-picture prediction has a disadvantage that it has a large coding complexity.

As described above, in the image encoding / decoding method of the related art, performance improvement for image encoding or image decoding is still required.

It is an object of the present invention to provide a motion vector candidate selection method capable of effectively selecting a motion vector candidate.

Another object of the present invention is to provide an image encoding / decoding method using a motion vector candidate selection method.

According to an aspect of the present invention, there is provided a method of generating spatial motion vector candidates, comprising the steps of: constructing a spatial motion vector candidate (first candidate); determining whether a reference picture of a current block exists in a current picture; Is added, adding a spatial motion vector candidate (second candidate) in another block of the current picture encoded before the current block, is provided.

Here, the motion vector candidate selection method may further include adding a temporal motion vector candidate (third candidate) if the determination result of the determining step is NO.

Here, the motion vector candidate selection method may further comprise, after the adding step, constructing a mixed list candidate including the first candidate, the second code and the third candidate.

Here, the motion vector candidate selection method may include: determining whether the current picture is a reference picture after constructing the mixed list candidate; And adding the fixed candidate of fixed coordinates if the current picture is a reference picture and the number of motion vector candidates in the mixed list candidate is smaller than a preset number.

Here, if the current picture is not a reference picture and the number of motion vector candidates in the mixed list candidates is smaller than a preset number, a step of adding (0, 0) as a fixed candidate is performed .

Here, another block of the current picture may include a block that is coded before the current block in the current picture, as a block that faces the current block with a neighboring block of the current block interposed therebetween. The other block of the current picture may be a block coded by inter-picture prediction prior to the current block.

According to another aspect of the present invention, there is provided a motion estimation method comprising: constructing a spatial motion vector candidate (first candidate); Determining whether a reference picture of the current block exists in the current picture; Adding a spatial motion vector candidate (second candidate) in another block of the current picture encoded before the current block if the determination result of the determining step is YES; Adding a temporal motion vector candidate (third candidate) if the determination result of the determining step is NO; And performing reference pixel filtering based on the motion vector candidate including the first candidate and the second candidate and the third candidate.

Here, the image coding method may further comprise: before the performing step, determining whether the current picture is a reference picture; If the current picture is a reference picture and the number of motion vector candidates in the mixed list candidate is smaller than a preset number, adding a fixed candidate of fixed coordinates previously set; And adding (0, 0) as a fixed candidate if the current picture is not a reference picture and the number of motion vector candidates in the mixed list candidate is smaller than a preset number.

Here, the image encoding method may further include: generating a prediction block through intra-picture prediction after the performing; And performing prediction mode encoding on the generated prediction block.

According to another aspect of the present invention, there is provided a method for generating a spatial motion vector candidate, the method comprising: constructing a spatial motion vector candidate (first candidate); Determining whether a reference picture of the current block exists in the current picture; Adding a spatial motion vector candidate (second candidate) in another block of the current picture encoded before the current block if the determination result of the determining step is YES; Adding a temporal motion vector candidate (third candidate) if the determination result of the determining step is NO; And performing motion prediction based on a motion vector candidate including at least one of the second candidate and the third candidate and the first candidate.

Here, the image coding method may further comprise: before the performing step, determining whether the current picture is a reference picture; If the current picture is a reference picture and the number of motion vector candidates in the mixed list candidate is smaller than a preset number, adding a fixed candidate of fixed coordinates previously set; And adding (0, 0) as a fixed candidate if the current picture is not a reference picture and the number of motion vector candidates in the mixed list candidate is smaller than a preset number.

Here, the image encoding method may further include performing interpolation after the performing step.

According to still another aspect of the present invention, there is provided a method of encoding a coded picture, the method comprising: entropy decoding a coded picture; Dequantizing the decoded picture; Transforming the inversely quantized picture; Selecting a motion information prediction candidate for an inversely transformed image based on the header information of the decoded picture; And decoding the inversely transformed image based on the addition / subtraction image information acquired based on the motion information prediction candidate.

Here, the selecting may include: a spatial motion vector candidate (first candidate) from a neighboring block of the current picture in the inversely transformed image; and a spatial motion vector candidate in another block of the current picture encoded prior to the current block A second candidate) for the current block based on the candidate group.

Here, the selecting may predict a motion of the current block based on a candidate group to which a temporal motion vector candidate (third candidate) is further added.

Here, the selecting step may include: forming a mixed list candidate group including the first candidate, the second code, and the third candidate, determining whether the current picture of the current block is a reference picture, If the number of motion vector candidates in the mixed list candidates is smaller than a preset number, fixed candidates of fixed coordinates may be added.

Here, if the current picture is not a reference picture and the number of motion vector candidates in the mixed list candidates is smaller than a preset number, the image decoding method may add (0, 0) as a fixed candidate have.

Here, another block of the current picture may include a block that faces the current block with a neighboring block of the current block interposed therebetween, and a block that is coded by inter-picture prediction before the current block in the current picture.

According to another aspect of the present invention for achieving the above object, there is provided a video encoding method for a reference pixel configuration in intra-frame prediction, comprising: obtaining a reference pixel of a current block from intra- Generating a prediction block of a current block by using a reference pixel to which adaptive filtering is applied as an input value according to a prediction mode of the current block, generating a post-processing filter adaptive to the prediction block, The method comprising the steps of:

Here, the acquiring step may acquire a reference pixel of the current block from the neighboring block.

Here, the acquiring step may be determined according to availability of neighboring blocks.

Here, the availability of the neighboring block may be determined by the position of the neighboring block and / or a specific flag (constrained_intra_pred_flag). In one example, a particular flag may have a value of 1 when neighboring blocks are available. That is, when the prediction mode of the neighboring block is the inter-view mode, it can mean that the reference pixel of the block can not be used for prediction of the current block.

Here, the specific flag (constrained_intra_pred_flag) may be determined according to the prediction mode of the neighboring block, and the prediction mode may be one of intra prediction or inter prediction.

Here, when the specific flag (constrained_intra_pred_flag) is 0, the availability of the neighboring block becomes 'true' irrespective of the prediction mode of the neighboring block. If the prediction mode of the neighboring block is intra- , And if the prediction is inter-picture prediction, the availability of the neighboring block may be 'false'.

Here, the inter picture prediction can generate a prediction block with reference to one or more reference pictures.

Here, the reference picture is managed through the reference picture list 0 (List 0) and the reference picture list 1 (List 1), and one or more past picture, future picture, and current picture can be included in the List 0 and List 1.

Here, List 0 and List 1 can be adaptively determined whether to insert the current picture into the reference picture list.

Here, the information for determining whether to insert the current picture into the reference picture list may be included in the sequence, the reference picture parameter set, and the like.

According to still another aspect of the present invention, there is provided a method of decoding an image performed in a computing device, the method comprising: obtaining a flag for a reference pixel availability of a neighboring block in a sequence or picture unit from an input bitstream; Determining a reference pixel availability of a neighboring block when intra prediction is performed according to a flag; When the flag is 0, the reference pixel of the neighboring block is used for prediction of the current block regardless of the prediction mode of the neighboring block. If the prediction mode of the neighboring block is intra prediction when the flag is 1, And the reference pixel of the neighboring block is not used for prediction of the current block when the prediction mode of the neighboring block is the inter-view prediction.

Here, the inter picture prediction can generate a prediction block based on block matching in the reference picture.

Here, the reference picture can be managed by List 0 in the P picture, and by List 0 and List 1 in the B picture.

Here, in the inter picture prediction, the current picture can be included in List 0.

Here, the current picture can be included in List 1 in inter-picture prediction.

Here, the inclusion of the current picture in the List 0 and List 1 can be determined based on the flag transmitted in the sequence parameter.

Here, the inclusion of the current picture in the List 0 and List 1 can be determined based on the flag transmitted from the picture parameter.

In the case of using the motion vector candidate selection method and the image coding / decoding method using the motion vector candidate method according to the embodiment of the present invention as described above, the motion vector candidates for the image coding in the image processing system and various systems including the image processing system Can be effectively selected to improve the performance and efficiency of the device or system.

Further, there is an advantage that the performance and efficiency of the image encoding apparatus, the image decoding apparatus, or the image processing system can be improved through effective selection of the motion vector candidate or the motion information prediction candidate.

1 is a view for explaining a system using an image encoding apparatus and / or an image decoding apparatus according to the present invention.
2 is a block diagram of an image encoding apparatus according to an embodiment of the present invention.
3 is a block diagram of an image decoding apparatus according to an embodiment of the present invention.
FIG. 4 is a diagram illustrating an inter-picture prediction of a P slice in an image encoding and decoding method according to an embodiment of the present invention.
FIG. 5 is a diagram illustrating an inter-picture prediction of a B slice in an image encoding and decoding method according to an embodiment of the present invention.
FIG. 6 is a diagram illustrating an example of generating a prediction block in a unidirectional manner in the image encoding and decoding method according to an embodiment of the present invention. Referring to FIG.
FIG. 7 is a diagram illustrating a reference picture list in the image encoding and decoding method according to an embodiment of the present invention. Referring to FIG.
FIG. 8 is a diagram illustrating an example of performing inter-picture prediction from a reference picture list in an image encoding and decoding method according to an embodiment of the present invention.
9 is an exemplary diagram for explaining intra prediction in the image encoding method according to an embodiment of the present invention.
10 is an exemplary diagram for explaining a prediction principle in a P slice or a B slice in the image encoding method according to an embodiment of the present invention.
11 is an exemplary diagram for explaining a case where interpolation is performed in the image encoding method of FIG.
FIG. 12 is a diagram for explaining a main procedure of an image coding method according to an embodiment of the present invention with a syntax in a coding unit.
13 is a diagram illustrating an example of supporting a symmetric type partition or an asymmetric type partition as in inter picture prediction when a prediction block is generated through block matching in the current picture used in FIG. Fig.
FIG. 14 is an exemplary diagram for explaining that 2Nx2N and NxN can be supported in inter-picture prediction (Inter) as in intra-picture prediction (Intra) in FIG.
FIG. 15 is a diagram for explaining a process of performing a horizontal 1D filter on pixels located at positions a, b, and c of an image (assuming x) in the image encoding method according to an embodiment of the present invention.
16 is an exemplary view of a current block and a neighboring block according to a comparative example.
17 is an exemplary view of a current block and neighboring blocks according to another comparative example.
18 is an exemplary view of a current block and a neighboring block according to yet another comparative example.
FIG. 19 is an exemplary view of a current block and neighboring blocks that can be employed in the image encoding method according to an embodiment of the present invention.
FIG. 20 is a block diagram illustrating a method of encoding an image according to an exemplary embodiment of the present invention. Referring to FIG. 20, when a temporal distance between a reference picture of a current block and a reference picture of a candidate block is greater than a predetermined distance, Candidate group, as shown in Fig.
FIG. 21 is an exemplary view for explaining a case where a current picture is added to a prediction candidate group when a picture referred to by a current block is a picture different from the current picture in the image coding method according to an embodiment of the present invention.
22 is an exemplary diagram for explaining a case where a current picture is added to a prediction candidate group when a picture referred to by a current block is a current picture in the image coding method according to an embodiment of the present invention.
23 is a flowchart illustrating a method of encoding an image according to another embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.

The terms first, second, A, B, etc. may be used to describe various elements, but the elements should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component. And / or < / RTI > includes any combination of a plurality of related listed items or any of a plurality of related listed items.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises ", or" having ", and the like, specify that the presence of stated features, integers, steps, operations, elements, parts, or combinations thereof, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

Unless otherwise defined, all terms used herein, including technical or scientific terms, are meant to be equivalent to those generally understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries are to be interpreted as being consistent with the meanings in the context of the relevant art and are not to be construed as ideal or overly formal meanings unless explicitly defined in the present application.

The moving picture may be generally composed of a series of pictures, and each picture may be divided into a predetermined area such as a frame or a block. In addition, the divided area includes not only a block but also a coding tree unit (CTU), a coding unit (CU), a prediction unit (PU), a transform unit (TU) As shown in FIG. Each unit may consist of one luminance block and two color difference blocks, which may be otherwise configured according to the color format. Further, the size of the luminance block and the color difference block can be determined according to the color format. For example, in the case of 4: 2: 0, the size of the color difference block may have a length of 1/2 of the luminance block. For these units and terms, terms such as conventional high efficiency video coding (HEVC) or advanced video coding (H.264 / AVC) can be referred to.

A reference picture, a reference block, or a reference pixel is referred to as a reference picture, a block, or a pixel to be referred to in encoding or decoding a current block or a current pixel. It is also to be understood that the term "picture" described below may be used in place of other terms having equivalent meanings such as image, frame, etc., I can understand if I am a child

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In order to facilitate the understanding of the present invention, the same reference numerals are used for the same constituent elements in the drawings and redundant explanations for the same constituent elements are omitted.

1 is a view for explaining a system using an image encoding apparatus and / or an image decoding apparatus according to the present invention.

1, a system using an image encoding apparatus and / or an image decoding apparatus includes a personal computer (PC), a notebook computer, a personal digital assistant (PDA), a portable multimedia player A user terminal 11 such as a PMP, a Playstation Portable (PSP), a wireless communication terminal, a smart phone, a TV or the like, or a server terminal such as an application server and a service server 12). Such a system may be referred to as a computing device.

In addition, the computing device may be a communication device such as a communication modem for performing communication with various devices or a wired / wireless communication network, an inter or intra prediction for encoding or decoding an image or encoding / A memory 18 for storing various programs and data for executing programs, a processor 14 for executing and calculating programs and the like, and the like.

In addition, the computing device can transmit a bitstream-encoded image by a video encoding device through a wired or wireless communication network such as the Internet, a short-range wireless communication network, a wireless LAN network, a WiBro network, a mobile communication network, (Universal Serial Bus), and the like, and can be decoded by the image decoding apparatus and reproduced as a reconstructed image. In addition, an image encoded by a video encoding apparatus by a bit stream may be transferred from a coding apparatus to a decoding apparatus via a computer-readable recording medium.

2 is a block diagram of an image encoding apparatus according to an embodiment of the present invention. 3 is a block diagram of an image decoding apparatus according to an embodiment of the present invention.

2, the image encoding apparatus 20 according to the present embodiment includes a predictor 200, a subtractor 205, a transformer 210, a quantizer 215, an inverse quantizer 220, An inverse transform unit 225, an adder 230, a filter unit 235, a decoded picture buffer (DPB) 240, and an entropy encoding unit 245. In addition, the image encoding apparatus 20 may further include a segmentation unit 190.

3, the image decoding apparatus 30 according to the present embodiment includes an entropy decoding unit 305, a predicting unit 310, an inverse quantizing unit 315, an inverse transforming unit 320, A decoded picture buffer 325, a filter unit 330, and a decoded picture buffer 335.

The image encoding apparatus 20 and the image decoding apparatus 30 may be separate apparatuses, but may be implemented as one image encoding / decoding apparatus according to the implementation. In this case, the prediction unit 200, the inverse quantization unit 220, the inverse transformation unit 225, the addition unit 230, the filter unit 235, and the decoding picture buffer 240 of the image coding apparatus 20 The inverse quantization unit 315, the inverse transform unit 320, the addition unit 325, the filter unit 330, and the memory 335 of the image decoding apparatus 30 as a technical element that is substantially the same as the prediction unit 310, the inverse quantization unit 315, But may at least include the same structure or may be implemented to perform at least the same function. In addition, the entropy coding unit 245 may correspond to the entropy decoding unit 305 when performing its function inversely. Therefore, a detailed description of the following technical elements and their operating principles will not be repeated.

Since the image decoding apparatus corresponds to a computing apparatus that applies the image encoding method performed in the image encoding apparatus to decoding, the following description will be made with reference to the image encoding apparatus.

The computing device may include a memory for storing a program or a software module implementing the image encoding method and / or the image decoding method, and a processor connected to the memory for executing the program. The video encoding apparatus may be referred to as an encoder, and the video decoding apparatus may be referred to as a decoder.

Each component of the image encoding apparatus of this embodiment will be described in more detail as follows.

The division unit 190 divides the input image into blocks (MxN) of a predetermined size. Here, M or N is an arbitrary natural number of 1 or more.

Specifically, the dividing unit 190 may be constituted by a picture dividing unit and a block dividing unit. The size or shape of the block can be determined according to the characteristics and resolution of the image, and the size or shape of the block supported through the picture partitioning unit is M × N square shape in which the length and the length are represented by an exponent of 2 (256 X 256, 128 x 128, 64 x 64, 32 x 32, 16 x 16, 8 x 8, 4 x 4, etc.), or an M x N rectangular shape. For example, an input image can be divided into 256 × 256 for an 8 k UHD image having a high resolution, 128 × 128 for a 1080p HD image, and 16 × 16 for a WVGA image.

Information on the size or shape of such a block can be set in units of a sequence, picture, slice, or the like, and related information can be transmitted to the decoder. That is, a sequence parameter set, a picture parameter set, a slice header, or a combination thereof.

Here, a sequence indicates a constituent unit formed by collecting a number of related scenes. A picture is a term that refers to a series of luminance (Y) components or luminance and chrominance (Y, Cb, Cr) components in one scene or picture. The range of one picture is one frame or one field .

A slice can refer to one independent slice segment and a number of dependent slice segments that are in the same access unit. An access unit means a set of network abstraction layer (NAL) units associated with a single coded picture. The NAL unit is a syntax structure composed of a video compression bitstream in a network-friendly format in the H.264 / AVC and HEVC standards. It is common to configure one slice unit as one NAL unit, and system standards generally regard a set of NALs or NALs constituting one frame as one access unit.

Returning to the description of the picture division section, information on the block size or type (M × N) can be made of an explicit flag, specifically, block type information, one length information when the block is square, The length information, or the difference value information between the horizontal and vertical lengths. For example, if M and N are composed of exponential powers of k (assuming k is 2) (M = 2 ^m , N = 2 ⁿ ), information on m and n can be unary binarization, Or the like, and transmit the related information to the decoding apparatus. Alternatively, if the division allowable minimum size (Minblksize) supported by the picture division unit is IxJ (assuming I = J for convenience of explanation, I = 2 ⁱ , J = 2 ^j ), information about mi or nj . As another example, when M and N are different, it is possible to convey the difference value (| mn |) between m and n. Alternatively, information about im or nj may be conveyed if the division allowable maximum size (Maxblksize) supported by the picture division unit is IxJ (assuming I = J for convenience of explanation, I = 2 ⁱ , J = 2 ^j ) .

In case of an implied situation, for example, if there is a syntax for the related information but can not be confirmed by the encoder / decoder, the encoder or decoder may follow the basic setting prepared in advance. For example, if the related syntax can not be confirmed at the step of confirming the block type information, the block type may be set to a square shape, which is the default setting. In addition, the step of checking the block size information may check the difference value related syntax in the step of checking the block size information through the difference value from the minimum allowable size (Minblksize) as in the above example, If the syntax related to the MinBlkSize is not available, it can be obtained from the default settings related to the MinBlkSize which is prepared in advance.

As such, the size or shape of the block in the picture partitioning unit can be implicitly determined by explicitly transmitting the related information in the encoder and / or the decoder or according to the characteristics and resolution of the image.

As described above, a block divided and determined through the picture division unit can be used as a basic encoding unit. The block divided and determined through the picture division unit may be a minimum unit that constitutes a high level unit such as a picture, a slice, and a tile, and may be a coding block, a prediction block, a quantization block, an entropy block, an inloopfiltering block, and the like, but some blocks are not limited to this and may be an exception. For example, some such as an in-loop filtering block may be applied in units larger than the block sizes described above.

The block division unit performs division on blocks such as coding, prediction, conversion, quantization, entropy, and in-loop filter. The partitioning unit 190 also functions in each configuration. For example, the transforming unit 210 may include a transform block dividing unit and the quantizing unit 215 may include a quantization block dividing unit. The size or shape of the initial block of the block partition may be determined by the result of division of the previous level or higher level block. For example, in the case of a coded block, a block obtained through a picture division unit which is a previous stage can be set as an initial block. Alternatively, in the case of a prediction block, a block obtained through a division process of a coding block which is a higher level of the prediction block may be set as an initial block. Alternatively, in the case of a transform block, a block obtained through a dividing process of an encoding block, which is a higher level of the transform block, may be set as an initial block. The condition for determining the size or type of the initial block is not always fixed, and there may be a case where a part is changed or an exception is made. In addition, it is also possible to set the setting level of the current level (for example, the size of the supported conversion block, the type of the conversion block, etc.) of the previous level or the higher level block (for example, the size of the encoding block, ) May affect the division operation of the current level (division possibility, divisible block type, etc.) according to a combination of at least one or more factors.

The block partitioning part can support a quad tree based partitioning method. That is, the block may be divided into four blocks each having a length of 1/2 in the horizontal and vertical directions in the block before division. This means that the block size limit (dep >> k, n >> k) means the allowable division depth limit (dep >> k, k >>) while the block size limit (dep_k) Can be repeatedly performed.

In addition, a binary tree-based partitioning scheme can be supported. This indicates that the length of one of the horizontal and vertical lengths can be divided into two blocks having a length of 1/2 as compared with the pre-division block. In the case of quad tree partitioning and binary tree partitioning, it may be a symmetric partition or an asymmetric partition, and it may be determined according to a partitioning scheme according to the setting of the encoder / decoder. In the image encoding method of the present invention, a symmetric division method will be mainly described.

The division flag (div_flag) indicates whether or not each block is divided. If the corresponding value is 1, the division is performed. If the value is 0, the division is not performed. Alternatively, if the value is 1, the segmentation is performed and further segmentation is possible. If the value is 0, the segmentation is not performed, and the segmentation is not allowed any more. Depending on the conditions such as the minimum allowable division size and the limit of the allowable division depth, the flag may be considered only for the division, and the additional division may not be considered.

The split flag can be used in a quadtree partition, and also in a binary tree partition. In binary tree segmentation, the dividing direction may be one of the block depth, encoding mode, prediction mode, size, type, and type (encoding, prediction, And may be determined depending on at least one of factors such as a slice type, a division allowable depth limit, a division allowable minimum / maximum size, or a combination thereof. In addition, it can be divided according to the division flag and / or according to the division direction, that is, divided into only 1/2 of the width of the block or 1/2.

For example, suppose that a block supports M × N (M> N) and M is greater than N, and that the current partition depth (dep_curr) is smaller than the partitioning allowable depth limit, The division flag is allocated with 1 bit. If the corresponding value is 1, the horizontal division is performed. If the division value is 0, the division flag can be no longer divided. The split depth can be one divide depth for quad tree and binary tree divisions, and one divide depth for quad tree and binary tree divisions. In addition, the allowable depth limit may have one partition allowable depth limit for the quadtree and the binary tree partition, or a respective partition allowable depth limit for the quadtree and binary tree partition.

 As another example, if the block is M × N (M> N) and N is equal to the predetermined minimum allowable division size, and the horizontal division is not supported, the division flag is allocated to 1 bit. If the value is 1, If it is 0, no division is performed.

Also, flags (div_h_flag, div_h_flag) for horizontal division or vertical division can be respectively supported, and binary division can be supported according to the flags. It is possible to indicate whether each block is horizontally or vertically divided through the horizontal division flag (div_h_flag) or the vertical division flag (div_v_flag). If the horizontal division flag (div_h_flag) or the vertical division flag (div_v_flag) If 0, no horizontal or vertical division is performed. Or, if each flag is 1, horizontal or vertical division is performed and horizontal or vertical division is possible. If the value is 0, horizontal division or vertical division is not performed and further division of horizontal or vertical division may not be allowed . The flag may be considered for the division depending on the conditions such as the division allowable minimum size and the division allowable depth limit, and the flag may not be considered. Alternatively, a flag (div_flag / h_v_flag) for horizontal division or vertical division may be supported, and binary division may be supported according to the flag. The division flag div_flag may indicate whether it is horizontally or vertically divided, and the division direction flag h_v_flag may indicate a horizontal or vertical division direction. If the division flag (div_flag) is 1, the division is performed and the horizontal or vertical division is performed according to the division direction flag (h_v_flag). If 0, the horizontal or vertical division is not performed. If the value is 1, horizontal or vertical division is performed according to the division direction flag (h_v_flag), and horizontal or vertical division is possible. If the value is 0, horizontal or vertical division is not performed. It can be regarded as not permitting. The flag may be considered for the division depending on the conditions such as the division allowable minimum size and the division allowable depth limit, and the flag may not be considered.

These division flags can also support for horizontal and vertical division, respectively, and can support binary tree division according to the flags. When the dividing direction is predetermined, only one of the two division flags may be used, or both of the two division flags may be used as in the above example.

For example, if the above flags are allowed, the possible block types can be divided into any one of M × N, M / 2 × N, M × N / 2, and M / 2 × N / 2. In this case, the flags can be encoded as 00, 10, 01, and 11 in the order of the horizontal division flag or the vertical division flag (div_h_flag / div_v_flag).

In the above case, the setting in which the division flag is used in the overlapping example is exemplified, and the setting in which the division flag can not be used overlapping is also possible. For example, the split block type may be divided into M × N, M / 2 × N, and M × N / 2. In this case, the above flags are encoded as 00, 01, Or the division flag div_flag and the horizontal-vertical flag h_v_flag, and the flags may be encoded as 0, 10, and 11 in the order of the horizontal or vertical division direction. Here, the meaning of the overlap may mean that the horizontal division and the vertical division are performed at the same time.

The quad tree division and the binary tree division described above can be used singly or in combination according to the setting of the encoder and / or the decoder. For example, a quadtree or binary tree partition may be determined depending on the size or type of the block. That is, when the block type is M x N and M is larger than N, the horizontal division, the block type is M x N, and if N is larger than M, the binary tree division can be supported according to the vertical division, Is M x N, and quad tree partitioning is supported if N and M are the same.

As another example, a binary tree partition may be supported if the size of the block (M x M) is greater than or equal to the block partition boundary value (thrblksize), and a quad tree partition may be supported if the size is smaller.

As another example, if M or N of the block (MxN) is equal to or smaller than the first allowable maximum size (Maxblksize1) and equal to or larger than the first allowable minimum size (Minblksize1), quad tree partitioning is supported, A binary tree segmentation may be supported if M or N of (M x N) is less than or equal to the second allowable maximum size Maxblksize2 and greater than or equal to the second allowable minimum size Minblksize2. If the first division supporting range and the second division supporting range, which can be defined as the division allowable maximum size and the division allowable minimum size, overlap, the priority of the first or second division method Rank can be given. In this example, quad tree partitioning is used for the first partitioning method, and binary tree partitioning is used for the second partitioning method.

For example, when the first division allowable minimum size (Minblksize1) is 16, the second division allowable maximum size (Maxblksize2) is 64, and the pre-division block is 64x64, the first division support range and the second division support range Quad tree partitioning and binary tree partitioning are possible because it belongs to all. If divide flag (div_flag) is 1, quad tree partitioning is performed and additional quadtree partitioning is possible, if the priority is given by the first partitioning method (quad tree partitioning in this example) according to the setting, Can be regarded as not performing a quadtree partition and no longer performing a quadtree partition.

Depending on the conditions such as the minimum allowable division size and the limit of the allowable division depth, the flag may be considered only for the division, and the additional division may not be considered. If the division flag div_flag is 1, it is divided into 4 blocks having a size of 32x32 and is larger than the first division allowable minimum size (Minblksize1), so that the quadtree partitioning can be continued.

If 0, no additional quadtree partitioning is performed, and the current block size (64x64) belongs to the second partitioning support range, so that binary tree partitioning can be performed. When the division flag (in the order of div_flag / h_v_flag) is 0, no further division is performed, and in the case of 10 or 11, a horizontal division or a vertical division can be performed.

In addition, when the block before division is 32x32 and the division flag (div_flag) is 0, the quad tree partition is not performed any more and the second block allowable maximum size (Maxblksize2) is 16, It does not belong to the second division support range and may not support further division. In the above description, the priority of the division method may be determined according to at least one of a slice type, a coding mode, a luminance / chrominance component, or a combination thereof.

As another example, various settings can be supported depending on the luminance and chrominance components. For example, the quadtree or binary tree partition structure determined by the luminance component can be used as it is without additional information / decoding in the color difference component. Alternatively, if independent division of the luminance component and the chrominance component is supported, quad tree + binary tree may be included in the luminance component, and quad tree division may be supported in the chrominance component. Alternatively, the quad tree + binary tree division is supported in the luminance and chrominance components, but the division support range may or may not be the same or proportional to the luminance and chrominance components. For example, when the color format is 4: 2: 0, the division support range of the chrominance component may be N / 2 of the division support range of the luminance component.

As another example, different settings can be made depending on the slice type. For example, I slices can support quadtree partitioning, P slices can support binary tree partitioning, and B slices can support quadtree partitioning and binary tree partitioning.

Quad tree partitioning and binary tree partitioning can be set and supported according to various conditions as in the above example. It should be noted that the above-mentioned examples are not limited to the above-mentioned cases and may include cases in which mutual conditions are reversed, and may include one or more factors mentioned in the above-mentioned examples or combinations thereof, It is possible. The above allowable division depth limit can be determined according to at least one factor or a combination of at least one of division method (quad tree, binary tree), slice type, luminance / chrominance component, coding mode and the like.

The division support range may be determined according to at least one factor or a combination thereof in a division mode (quad tree, binary tree), a slice type, a luminance / chrominance component, a coding mode, Maximum value, and minimum value.

When information on this is configured with an explicit flag, length information of each of the maximum value / minimum value, or difference value information between the minimum value and the maximum value can be expressed. For example, when the maximum value and the minimum value are composed of exponentiation powers of k (assuming k is 2), the exponential information of the maximum and minimum values can be encoded through various binarization and transmitted to the decoding apparatus. Alternatively, the difference value between the maximum value and the minimum value can be transmitted. The information transmitted at this time may be the difference information of the exponent information and the exponent of the minimum value

In accordance with the above description, information related to flags can be generated and transmitted in units of sequences, pictures, slices, tiles, blocks, and the like.

The block division information may be represented by a quad tree, a binary tree, or a combination of two tree schemes with the division flags shown in the above examples. The division flags may be encoded by various methods such as unary binarization and cutting type unary binarization, Device. The bitstream structure of the division flag for representing the division information of the block may be selected from one or more scanning methods.

For example, a bitstream of division flags can be configured based on the division depth order (from dep0 to dep_k), and a bitstream of division flags can be configured based on division. In the division depth order reference method, division information at the current level is acquired on the basis of the first block, and division information is acquired at the next level. In the division method, It means a method of preferentially acquiring additional division information, and other scanning methods not shown in the above example may be included and selected.

Also, according to the implementation, the block dividing unit may generate index information for a block candidate group of a predetermined type, which is not the division flag described above, and express it. M × N, M × N, M × N / 2, M / 4 × N, 3M / 4 × N, and M × N are the types of block candidates, M × N / 4, M × 3N / 4, M / 2 × N / 2, and the like.

When the candidate group of the divided block is determined as described above, the index information for the divided block type can be encoded by various methods such as fixed length binarization, truncated binarization, truncated binarization, and the like. At least one of factors such as the division depth of a block, a coding mode, a prediction mode, a size, a type, a type and a slice type, a division allowable depth limit, a division allowable minimum / maximum size, The divided block candidate group can be determined.

(M × N, M × N / 2, M / 2 × N / 2) as candidate lists 1 (list 1) (M × N, M / 2 × N, M × N / 2, M / N × 2) of the list 2 (list 2) 4 × N, 3M / 4 × N, M × N / 4, M × 3N / 4 and M / 2 × N / 2) are assumed to be the candidate list 4 (list4). For example, when describing M × N as a reference, a divided block candidate of candidate list 2 can be supported when (M = N), and a divided block candidate of candidate list 3 if (M ≠ N).

As another example, if M × N of M or N is greater than or equal to the boundary value blk_th, the divided block candidate of the candidate list 2 can be supported, and if it is smaller than that, the divided block candidate of the candidate list 4 can be supported. When M or N is equal to or greater than the first threshold value blk_th_1, if the divided block candidate of the candidate list 1 is smaller than the first threshold value blk_th_1 but equal to or greater than the second threshold value blk_th_2, If the divided block candidate of list2 is smaller than the second boundary value blk_th_2, the divided block candidate of the candidate list 4 can be supported.

As another example, it is possible to support a divided block candidate of the candidate list 2 when the coding mode is intra-picture prediction and a divided block candidate of the candidate list 4 when inter prediction is performed.

Even if the divided block candidates are supported, the bit configuration according to the binarization in each block may be the same or different. For example, if the divided block candidates supported by the block size or shape are limited as in the case of the division flag, the bit configuration according to the binarization of the corresponding block candidate may be changed. M × N, M × N / 2 and M / 2 × N / 2 in the case of (M> N) The binary bits of the index according to the M × N / 2 of the current condition and the M × N / 2 of the current condition may be different from each other.

Information on the segmentation and type of the block can be expressed using one of the segmentation flag or the segmentation index method according to the type of the block used for coding, prediction, conversion, quantization, entropy, in-loop filtering, and the like. In addition, block size limitation and division allowable depth limit for the partitioning and block type support may be different depending on each block type.

In the coding / decoding process on a block-by-block basis, after the coding block is determined, the coding / decoding can be performed according to the processes of the prediction block determination, the transform block determination, the quantization block determination, the entropy block determination and the in-loop filter determination. The order of the encoding / decoding process is not always fixed, and some order may be changed or excluded. The size and shape of each block are determined according to the coding cost of each candidate of the size and the shape of the block, and coding related information such as image data of each determined block and size and type of each determined block can be encoded.

The prediction unit 200 can be implemented using a prediction module, which is a software module, and can generate a prediction block using an intra-picture prediction scheme or an inter-picture prediction scheme for a block to be encoded. Here, the prediction block is a block which is understood to closely match the block to be encoded in terms of pixel difference, and can be determined by various methods including a sum of absolute difference (SAD) and a sum of square difference (SSD). Also, at this time, various syntaxes that can be used in decoding the image blocks may be generated. The prediction block can be classified into an in-picture block and an inter-picture block according to an encoding mode.

Intra prediction refers to a prediction technique that uses spatial correlation and predicts a current block using reference pixels of previously reconstructed and decoded blocks in the current picture. That is, the brightness value reconstructed by intra prediction and reconstruction can be used as reference pixels in the encoder and the decoder. Intra prediction can be effective for flat areas with continuity and areas with constant directionality, and it can be used for the purpose of guaranteeing random access and preventing error diffusion because it uses spatial correlation.

Inter prediction uses a compression technique that removes redundancy of data by using temporal correlation with reference to an image encoded in one or more past or future pictures. That is, the inter picture prediction can generate a prediction signal having high similarity by referring to one or more past or future pictures. In the encoder using inter-picture prediction, a block having a high degree of correlation with a block to be currently coded is searched in the reference picture, and the position information and residue signal of the selected block can be transmitted to the decoder. A reconstructed image can be constructed by generating the same prediction block as the encoder and compensating the transmitted residual signal.

FIG. 4 is a diagram illustrating an inter-picture prediction of a P slice in an image encoding and decoding method according to an embodiment of the present invention. FIG. 5 is a diagram illustrating an inter-picture prediction of a B slice in an image encoding and decoding method according to an embodiment of the present invention.

In the image coding method of this embodiment, the inter picture prediction generates a prediction block from a previously coded picture having a high temporal correlation, so that the coding efficiency can be increased. Current (t) may refer to the current picture to be coded and may be a first temporal distance (t-1) prior to the POC of the current picture based on temporal flow of the picture or POC (picture order count) And a second reference picture having a second temporal distance (t-2) before the first temporal distance.

4, inter-picture prediction that can be employed in the video coding method of this embodiment is the same as that of the current block of the current picture (current (t)) and that of the reference pictures (t-1, t-2) It is possible to perform motion estimation for finding an optimal prediction block from reference pictures (t-1, t-2) that have been previously encoded with a block having a high correlation through block matching of reference blocks. An interpolation process based on a structure in which at least one or more subpixels are arranged between adjacent two pixels as necessary for precise estimation is performed and then an optimal prediction block is found and motion compensation is performed, Can be found.

5, the inter-picture prediction that can be employed in the image coding method of the present embodiment is based on the already-coded reference pictures (" current (t) " t-1, t + 1). In addition, two prediction blocks can be generated from one or more reference pictures.

When encoding an image through inter-picture prediction, motion vector information for the optimal prediction block and information about the reference picture are encoded. In this embodiment, when a prediction block is generated in a unidirectional or bidirectional direction, a reference block list may be configured differently to generate a prediction block from the corresponding reference picture list. Basically, the reference pictures temporally present before the current picture are managed by being allocated to the list 0 (L0), and the reference pictures existing after the current picture can be allocated to the list 1 (L1) and managed.

The reference pictures existing after the current picture can be allocated when the reference picture list 0 can not be filled up to the reference picture list number 0 of the reference picture list 0. Similarly, when the reference picture list 1 is constructed, if the reference picture list 1 can not satisfy the reference picture list, the reference pictures existing before the current picture can be allocated.

FIG. 6 is a diagram illustrating an example of generating a prediction block in a unidirectional manner in the image encoding and decoding method according to an embodiment of the present invention. Referring to FIG.

Referring to FIG. 6, in the image encoding and decoding method according to the present embodiment, a prediction block can be found from a previously encoded reference picture t-1, t-2. In addition, a current picture t) from the region where coding has already been completed.

That is, in the image encoding and decoding method according to the present embodiment, not only a prediction block is generated from a previously coded picture (t-1, t-2) having a temporally high correlation, but also a prediction block having a high spatial correlation As shown in FIG. Finding such a spatially highly correlated prediction block can correspond to finding a prediction block in a manner of intra prediction. In order to perform block matching from an area where coding is completed in the current picture, the image coding method of the present embodiment can construct a syntax for information related to the prediction candidate in combination with the intra-picture prediction mode.

For example, when n (n is an arbitrary natural number) branch prediction mode is supported, one mode is added to the intra prediction prediction candidate to support n + 1 modes and ^{2M-1? N} + 1 ≪ 2 < ^M > can be used to encode the prediction mode. Also, it can be implemented to select among candidate groups of probable prediction modes, such as MPV (most probable mode) of HEVC. It is also possible to preferentially encode at a higher stage of prediction mode encoding.

When a prediction block is generated through block matching in the current picture, the image coding method of the present embodiment may form a syntax for information related to mixing with the inter-picture prediction mode. Additional associated prediction mode information may be motion or displacement related information. The motion or motion-related information may include optimal candidate information among various vector candidates, a difference value between an optimal candidate vector and an actual vector, a reference direction, reference picture information, and the like.

FIG. 7 is a diagram illustrating a reference picture list in the image encoding and decoding method according to an embodiment of the present invention. Referring to FIG. FIG. 8 is a diagram illustrating an example of performing inter-picture prediction from a reference picture list in an image encoding and decoding method according to an embodiment of the present invention.

7, a method of encoding an image according to an embodiment of the present invention includes a first reference picture list (L0) and a second reference picture list (L0) for a current block of a current picture (current (t) 1, < / RTI > L1).

Referring to FIGS. 7 and 8, the reference picture list 0 can be composed of reference pictures before the current picture t, where t-1 and t-2 are respectively the first , The temporal distance (t-1), and the second temporal distance (t-2). The reference picture list 1 can be composed of a reference picture after the current picture t and t + 1 and t + 2 respectively represent a first temporal distance t + 1 after the POC of the current picture t ) And a second temporal distance (t + 2).

Although the above-described examples of the reference picture list construction show an example in which the reference picture list is composed of the reference pictures having the difference of the temporal distance (POC standard in this example), the difference in temporal distance between the reference pictures is configured differently You may. That is, the index difference of the reference pictures and the temporal distance difference of the reference pictures may not be proportional. In addition, the list arrangement order may not be constituted based on temporal distance. This can be confirmed in an example of the reference picture list construction to be described later.

The prediction can be performed from the reference picture in the list according to the slice type (I, P, or B). When a prediction block is generated by block matching in the current picture (current (t)), a current picture is added to a reference picture list (reference list 0 and / or reference list 1) .

The current picture t can be added to the reference list 0 or the current picture current t can be added to the reference list 1 as shown in FIG. That is, the reference picture list 0 can be constructed by adding a reference picture which is a temporal distance (t) to the reference picture before the current picture (t), and the reference picture list 1 is composed of time A reference picture that is a distance t may be added.

For example, when the reference picture list 0 is constructed, the reference picture prior to the current picture can be assigned to the reference picture list 0, and the current picture t can be allocated. When the reference picture list 1 is constructed, The reference picture can be assigned to the reference picture list 1 and then the current picture t can be allocated. Alternatively, when the reference picture list 0 is constructed, the current picture t can be allocated and the reference picture prior to the current picture can be allocated. When the reference picture list 1 is constructed, the current picture t is allocated, It is possible to assign subsequent reference pictures.

Also, when the reference picture list 0 is constructed, the reference picture before the current picture can be allocated, the reference picture after the current picture can be allocated, and the current picture t can be allocated. Similarly, when constructing the reference picture list 1, the reference pictures after the current picture can be allocated, the reference pictures before the current picture can be allocated, and the current picture t can be allocated. The above examples are not specific to the above-mentioned case, but may include cases where the conditions of each other are reversed, and examples of other cases may also be modified.

Whether or not to include the current picture in each reference picture list (for example, not adding to any list, adding to list 0 only, or adding to list 1 or adding to lists 0 and 1) Information can be transmitted in units of sequences, pictures, slices, and the like. The information on this can be encoded by a method such as fixed-length binarization, truncation-type binarization, truncation-type binarization, and the like.

The image encoding and decoding method of this embodiment differs from the method of FIG. 7 in that block matching is performed on the current picture t to select a prediction block, and a reference picture list including related information about the prediction block is constructed , There is a difference in that such a reference picture list is used for image coding and decoding.

In the reference picture list configuration, it is possible to set the order and the rule of each list and the allowable number of the reference pictures of the respective lists differently. This includes whether the current picture is included in the list (whether the current picture is included as a reference picture in inter- (E.g., whether or not it is applicable to the list 0 or 1), a slice type, a list reconstruction parameter (which may be applied to lists 0 and 1 respectively, or may be applied to lists 0 and 1 respectively), a position in a group of pictures (GOP) At least one of the factors, or a combination thereof, and can explicitly transmit the related information in units of a sequence, picture, or the like.

For example, in the case of a P slice, the reference picture list 0 can follow the list construction rule A, regardless of whether the current picture is included in the list, and the reference picture list 0 that includes the current picture in the list in the case of B slice, Rule B and the reference picture list 1 can follow the list construction rule C, and the list construction rule D for the reference picture list 0 that does not include the current picture and the list construction rule E for the reference picture list 1, B and D, C and E may be the same. The list configuration rules may be configured in the same or modified manner as described in the above-described reference picture list configuration example.

As another example, when the current picture is included in the list, the first reference picture allowable number can be set, and when not included, the second reference picture allowable number can be set. The first reference picture allowable number and the second reference picture allowable number may be equal to or different from each other and the difference between the first reference picture allowable number and the second reference picture allowable number may be set to 1 by default.

As another example, if the current picture is included in the list and the list reconstruction parameter is applied, all the reference pictures in the slice A can be the list reconstruction candidates, and the slice B can include only some reference pictures in the list reconstruction candidate group. At this time, the slice A or B can be classified into whether or not the current picture is included in the list, temporal layer information, slice type, position in the GOP, and the like. The POC or reference picture index of the reference picture, (Before / after the current picture), whether or not the current picture is present, and the like.

According to the above-described structure, the reference block encoded with the inter-picture prediction can be used in the current picture, so that inter-picture prediction can be allowed or utilized even in the moving prediction of the I-slice.

Further, when constructing the reference picture list, the index allocation or the list construction order may be different according to the slice type. In the case of the I-slice, a lower index (for example, idx = 0, 1, 2) is used in the current picture (current (t)) with a higher priority as in the reference picture list structure example, It is possible to reduce the bit amount in the image encoding through the binarization (fixed length binarization, chopping-type binarization, truncation-type binarization, etc.) with the maximum allowable reference picture number C of the reference picture list. In the case of the P or B slice, when it is judged that the block matching is performed in the current picture and it is judged that it is lower than the probability of selecting the prediction candidate through the reference picture having a different probability of selecting the reference picture of the current block as the prediction candidate, The binarization of various methods of setting the priority for the block matching of the reference picture list to the maximum value using a higher index (for example, idx = C, C-1) The amount of bits in the image encoding can be reduced. In the above example, the priority setting of the current picture can be configured in the same or modified manner as that described in the reference picture list configuration example. In addition, it is possible to omit the information on the reference picture by not constructing the reference picture list according to the slice type (for example, I slice). For example, it is possible to generate a prediction block through the existing inter-view prediction, but to display the inter-view prediction information with the remainder excluding the reference picture information in the motion information in the inter-picture prediction mode.

The method of performing block matching in the current picture can determine whether or not to support according to the slice type. For example, block matching in the current block is supported by I slices, but not P slices or B slices, or other variations are possible. In addition, the method of supporting block matching in the current picture may determine whether to support in units of pictures, slices, tiles, etc., or may be determined according to the position in the GOP, temporal layer information, or the like. Such setting information may be transmitted in units of a sequence, a picture, a slice, or the like in an image encoding process or a decoder in an encoder.

Further, even when the above setting information or syntax exists in the upper level unit and the setting related operation is turned on, when the same setting information or syntax exists in the lower level unit, the setting information in the lower level unit is changed to the upper level The setting information in the unit can be prioritized. For example, if the same or similar setting information is processed in a sequence, a picture, or a slice, a picture unit may have priority rather than a sequence unit, and a slice unit may have priority rather than a picture unit.

9 is an exemplary diagram for explaining intra prediction in the image encoding method according to an embodiment of the present invention.

Referring to FIG. 9, the intra-frame prediction method according to the present embodiment includes a reference sample padding, a reference sample filtering, an intra predicting, and a boundary filtering. May comprise a series of steps.

The reference pixel filling step may be an example of the reference pixel forming step, the reference pixel filtering step may be executed by the reference pixel filter part, and the intra picture prediction may include a prediction block generating step and a prediction mode coding step, The filtering may be an example for one embodiment of the post-processing filter step.

That is, the intra-picture prediction performed in the image encoding method of this embodiment may include a reference pixel configuration step, a reference pixel filtering step, a prediction block generation step, a prediction mode encoding step and a post-processing filtering step. Depending on various environmental factors such as block size, block type, block location, prediction mode, prediction method, quantization parameter, etc., one or a part of the above processes may be omitted and other processes may be added, May be changed in a different order.

The reference pixel construction step, the reference pixel filtering step, the prediction block generation step, the prediction mode encoding step, and the post-processing filtering step may be implemented by a processor connected to a memory by executing software modules stored in the memory. Therefore, in the following description, for convenience of description, a functional unit generated by a combination of a software module that implements each step and a processor that executes the function, and / or a component that performs the function of the functional unit, A prediction block generation unit, a prediction mode encoding unit, and a post-processing filter unit will be referred to as an execution subject of each step.

More specifically, the reference pixel forming unit forms a reference pixel to be used for predicting the current block through reference pixel filling. If a reference pixel does not exist or is unavailable, reference pixel filling can be used for the reference pixel through a method such as copying a value from the nearest available pixel. A reconstructed picture buffer or a decoded picture buffer (DPB) may be used for copying a value or the like.

That is, the intra prediction uses the reference pixels of the blocks that have been encoded before the current picture to perform the prediction. To this end, adjacent pixels such as left, top, bottom left, top, and top right blocks of the current block are mainly used as reference pixels in the reference pixel configuration step.

However, the candidate group of the neighboring block for the reference pixel is only an example in which the coding order of a block is followed by a raster scan or a z-scan, Adjacent pixels such as right, lower right, and lower blocks may be used as reference pixels in addition to the above blocks.

Further, according to the implementation, additional pixels other than immediately adjacent pixels may be used as a substitute or mixed with an existing reference pixel according to the stepwise configuration of the intra prediction.

In addition, in the case of predicting a mode having a directionality among the modes of intra-picture prediction, it is possible to generate a reference pixel in decimal units through linear interpolation of reference pixels in integer units. The mode for performing the prediction through the reference pixel existing in the integer unit position includes some modes having vertical, horizontal, 45 degrees, and 135 degrees, and the process of generating reference pixels in decimal units for the above prediction modes It may not be necessary.

The reference pixels to be interpolated in the prediction modes having the direction other than the prediction mode are 1/2, 1/4, 1/8, 1/16, 1/32, 1/64, It may have interpolation precision, or it may have accuracy of a multiple of 1/2.

This is because the interpolation accuracy can be determined depending on the number of supported prediction modes or the prediction direction of the prediction mode and the like. A fixed interpolation precision may always be supported in a picture, a slice, a tile, a block, etc., and an adaptive interpolation precision may be supported depending on a block size, a block type, a prediction direction of a supported mode, At this time, the prediction direction of the mode can be expressed by the inclination information or the angle information of the direction indicated by the mode on a specific line reference (for example, x + axis of the positive + on the coordinate plane).

As an interpolation method, linear interpolation can be performed through adjacent integer pixels, but other interpolation methods can be supported. For interpolation, one or more filter types and the number of taps can be supported. For example, a 6-tap Wiener filter, an 8-tap Kalman filter, and so on can be determined. have. Further, the related information may be transmitted in units of a sequence, picture, slice, block, or the like.

The reference pixel filter unit can filter the reference pixels for the purpose of improving the prediction efficiency by reducing the deterioration in the encoding process after the reference pixels are constructed. The reference pixel filter unit can implicitly or explicitly determine the type of the filter and the application of the filtering according to the size, type, and prediction mode of the block. That is, even if the filter is the same tap, the filter coefficient can be determined differently depending on the filter type. For example, you can use a 3-tap filter such as [1,2,1] / 4, [1,6,1] / 8.

In addition, the reference pixel filter unit can determine whether to apply filtering by determining whether to send additional bits or not. For example, in an implied case, the reference pixel filter unit can determine whether to apply filtering according to the characteristics (dispersion, standard deviation, and the like) of the pixels in the peripheral reference block.

In addition, the reference pixel filter unit may determine whether or not the filtering is applied when the related flag satisfies a predetermined hiding condition such as a residual coefficient, an intra prediction mode, and the like. The number of taps of the filter is, for example, 3-tap such as [1,2,1] / 4 in a small block (blk), [2,3,6,3,2] / 16 in a large block The same 5-tap may be set, and the number of times of application may be determined by whether to perform filtering, whether to perform filtering once, or to perform filtering twice.

Also, the reference pixel filter unit can basically apply filtering to the nearest reference pixel of the current block. Additional reference pixels besides the nearest reference pixel may also be considered in the filtering process. For example, filtering may be applied to additional reference pixels by replacing the nearest reference pixel, or filtering may be applied by mixing additional reference pixels with the nearest reference pixel.

The filtering may be applied either fixedly or adaptively, depending on the size of the current block, the size of the neighboring block, the encoding mode of the current block or neighboring block, the block boundary property of the current block and the neighboring block (e.g., A prediction mode or direction of a current block or a neighboring block, a prediction method of a current block or a neighboring block, a quantization parameter, and the like, or a combination thereof. have. The decision can be made on the encoder / decoder with the same setting (implicit), or on the basis of the coding cost (explicit). Basically, the filter applied is a low pass filter. Depending on various factors described above, the number of filter taps, the filter coefficient, whether or not the filter flag is encoded, and the number of filter application can be determined. Slice, block, or the like, and can transmit the related information to the decoder.

The prediction block generation unit may perform an extrapolation or an extrapolation scheme through a reference pixel in intra-picture prediction, an interpolation scheme such as an average value (DC) or a planar mode of a reference pixel, a copy of a reference pixel ) Prediction block can be generated.

In the case of copying a reference pixel, one reference pixel may be copied to generate one or more prediction pixels, one or more reference pixels may be copied to generate one or more prediction pixels, and the number of copied reference pixels may be copied May be equal to or less than the number of prediction pixels.

In addition, according to the prediction method, it is possible to classify into a directional prediction method and a non-directional prediction method. Specifically, the directional prediction method can be classified into a linear directional method and a curved directional method. The linear direction method uses extrapolation or extrapolation but the pixels of the prediction block are generated through the reference pixels lying on the prediction direction line. The curved directional method uses an extrapolation or an extrapolation method, but the pixels of the prediction block are referred to as a reference Means a scheme in which a partial prediction direction of a pixel is allowed to be changed in consideration of a detailed directionality of a block (for example, edge <Edge>). In the case of the directional prediction mode in the image encoding and decoding method of the present invention, a linear directional method will be mainly described. In the case of the directional prediction method, the intervals between adjacent prediction modes may be equal or unequal, and this may be determined according to the size or type of the block. For example, when a block with the size and shape of M × N is acquired by the block division, if the M and N are the same, the intervals between the prediction modes may be uniform. If M and N are different The intervals between the prediction modes may be unequal. As another example, when M is greater than N, modes with vertical orientation may allocate finer intervals between prediction modes close to the vertical mode (90 degrees), and wider spacings may be allocated to far modes in the vertical mode. When N is greater than M, modes with horizontal orientation may allocate finer intervals between prediction modes close to the horizontal mode (180 degrees), and wide spacing may be allocated to horizontal modes with far prediction mode. The above examples are not specific to the above-mentioned case, but may include cases where the conditions of each other are reversed, and examples of other cases may also be modified. In this case, the interval between the prediction modes can be calculated based on numerical values indicating the directionality of each mode, and the directionality of the prediction mode can be numerically expressed as directional inclination information or angle information.

In addition, a prediction block including other methods using spatial correlation other than the above method can be generated. For example, a reference block using an inter prediction scheme such as motion search and compensation can be generated as a prediction block using the current picture as a reference picture.

The prediction block generating step may generate the prediction block using the reference pixel according to the prediction method. That is, it is possible to generate a prediction block through directional prediction or non-directional prediction such as extrapolation, interpolation, copying, and averaging of the existing intra-picture prediction method according to the prediction method, And other additional methods may be used.

The intra-picture prediction scheme may be supported under the same settings of the encoder / decoder, and may be determined according to a slice type, a block size, a block type, and the like. The intra prediction scheme can be supported according to at least one of the above-mentioned prediction schemes or a combination thereof. The intra prediction mode can be configured according to the supported prediction mode. The number of intra prediction modes supported can be determined according to the prediction method, slice type, block size, block type, and the like. The related information can be set and transmitted in units of a sequence, a picture, a slice, and a block.

The prediction mode encoding step for performing the prediction mode encoding can determine the mode with the best encoding cost according to each prediction mode as the prediction mode of the current block in terms of encoding cost.

For example, the prediction mode encoding unit may use one or more neighboring block modes for current block mode prediction for the purpose of reducing prediction mode bits. (MPM) candidates that are likely to be the same as the current block mode, and the modes of neighboring blocks may be included in the above candidate group. For example, the prediction mode of a block such as the upper left corner, the lower left corner, the upper right corner, and the upper right corner of the current block may be included in the above candidate group.

The candidate group of the prediction mode includes the position of the neighboring block, the priority of the neighboring block, the priority in the divided block, the size or type of the neighboring block, the preset specific mode, and the prediction mode of the luminance block And may be configured according to at least one or more factors or combinations thereof, and the related information may be transmitted in units of a sequence, a picture, a slice, a block, or the like.

For example, when a block adjacent to the current block is divided into two or more blocks, it is possible to determine which block mode among the divided blocks is included as a mode prediction candidate of the current block under the same setting of the encoder / decoder have. For example, the left block among the neighboring blocks of the current block (M × M) is divided into three blocks by performing a quadtree partition in the block division block, and M / 2 × M / 2 and M / 4 × M / 4 and M / 4 × M / 4, the prediction mode of the M / 2 × M / 2 block can be included as the mode prediction candidate of the current block on the basis of the block size. As another example, the upper block among the neighboring blocks of the current block (N × N) is divided into three blocks by performing the binary tree division in the block division block, and N / 4 × N and N / 4 × N, and N / 2 × N blocks, the prediction mode of the first N / 4 × N block on the left according to a predetermined order (priority assigned from left to right) is set as a mode prediction candidate of the current block .

As another example, if the prediction mode of the block adjacent to the current block is the directional prediction mode, the prediction mode adjacent to the prediction direction of the current mode (inclination information of the direction of the mode or angle information side) can do. In addition, the predetermined modes (planar, DC, vertical, horizontal, etc.) may be preferentially included according to the prediction mode configuration or combination of neighboring blocks. In addition, a prediction mode having a high occurrence frequency among the prediction modes of neighboring blocks can be preferentially included. The priority may be a possibility to be allocated not only to the possibility of being included in the mode prediction candidate group of the current block but also a higher priority or index in the candidate group configuration (that is, a higher probability of allocating fewer bits in the binarization process) have.

As another example, the maximum number of mode prediction candidates of the current block is k, the left block is composed of m blocks whose length is shorter than the length of the current block, and the upper block is n blocks whose length is shorter than the width of the current block (M + n) of neighboring blocks can be filled in according to a predetermined order (left to right, top to bottom) when the divided block sum (m + n) of neighboring blocks is larger than k, (E.g., lower left, upper left, upper right, etc.) other than the neighboring block position in the prediction mode of the neighboring block (left block, upper block) The prediction mode of the same block may be included in the mode prediction candidate group of the current block. The above examples are not specific to the above-mentioned case, but may include cases where the conditions of each other are reversed, and examples of other cases may also be modified.

As described above, the candidate block for predicting the mode of the current block is not limited to a specific block position, and prediction mode information can be utilized from at least one block among left, upper left, lower left, upper and upper right blocks, The prediction mode of the current block can be configured as a candidate group considering various factors as in the above example.

The prediction mode encoding unit can classify the candidate MPM candidate group (referred to as the candidate group 1 in this example) and the mode candidate group (referred to as the candidate group 2 in this example) having a high probability of being the same as the current block mode, The prediction mode encoding process may be changed depending on which candidate group among the candidate groups the prediction mode of the block belongs to. The total prediction mode may be composed of the sum of the prediction mode of the candidate group 1 and the prediction mode of the candidate group 2. The number of prediction modes of the candidate group 1 and the prediction group of the candidate group 2 may be the sum of the total number of prediction modes, The shape of the block, and the like, or a combination thereof. The same binarization may be applied or another binarization may be applied according to the candidate group. For example, fixed length binarization may be applied to candidate group 1, and truncation type binarization may be applied to candidate group 2. Although the number of candidate groups is exemplified in the above description, it is assumed that the number of the candidate groups is two, but the number of the candidate groups is not limited to the first candidate group having a high probability of being the same as the current block, the second candidate group having a high probability of being the same as that of the current block, It is expandable, and its variants are also possible.

The post-processing filtering step executed by the post-processing filter unit may include a step of performing a post-processing filtering step of performing a post- May be replaced with a value generated by filtering one or more reference pixels adjacent to the boundary and one or more prediction pixels, and a value obtained by quantifying characteristics between reference pixels adjacent to the boundary of the block (for example, Gradient information, etc.) may be applied to the filtering process to replace the predictive pixel with the generated value, and another method having a similar purpose (correcting a part of the predictive pixel of the predictive block through the reference pixel) .

In the post-processing filter unit, the kind of the filter and whether or not the filtering is applied can be implicitly or explicitly determined. The reference pixel used in the post-processing filter unit, the position and the number of the current pixel, And can be set in an encoder / decoder, and related information can be transmitted in units of a sequence, a picture, a slice, and the like.

Further, in the post-processing filtering step, an additional post-processing process can be performed after generating a prediction block such as boundary filtering. In addition, the reconstructed current block, which is obtained by adding the residual signal and the prediction signal obtained through the conversion / quantization process after the acquisition of the residual signal and the inverse process, is subjected to post-processing filtering considering the characteristics of the pixels of the reference block, .

Finally, the prediction block is selected or obtained through the above-described process. Information obtained in this process may include prediction mode related information, and after the prediction block is acquired, it is transmitted to the conversion unit 210 for encoding the residual signal .

10 is an exemplary diagram for explaining a prediction principle in a P slice or a B slice in the image encoding method according to an embodiment of the present invention. 11 is an exemplary diagram for explaining a case where interpolation is performed in the image encoding method of FIG.

Referring to FIG. 10, the image encoding method according to the present embodiment may include motion estimation module and interpolation steps. Information on the motion vector, the reference picture index, and the reference direction generated in the motion prediction step may be transmitted to the interpolation step. In the motion prediction step and the interpolation step, a value stored in the decoded picture buffer (DPB) can be used.

That is, the image encoding apparatus can perform motion estimation to find a block similar to the current block in the previous encoded pictures. In addition, the image encoding apparatus can perform interpolation of a reference picture for precise prediction with respect to precision in decimal units. Finally, the image coding apparatus acquires a prediction block through a predictor. The information generated in this process includes a motion vector, a reference picture index, a reference direction, ), And the residual signal coding can be proceeded.

In the present embodiment, in-picture prediction is performed in P-slice or B-slice, so that it is possible to implement a combination scheme as shown in FIG. 11 that supports inter-picture prediction and intra-picture prediction.

As shown in FIG. 11, the image encoding method according to the present embodiment includes reference sample padding, reference sample filtering, intra prediction, boundary filtering, Motion estimation, and interpolation. &Lt; RTI ID = 0.0 &gt;

When the image encoding apparatus supports block matching in the current picture, the prediction method in the I-slice can be implemented by the configuration shown in FIG. 11 instead of the configuration shown in FIG. That is, the image encoding apparatus can use information such as a motion vector, a reference picture index, and a reference direction, which are generated not only in the prediction mode in the I-slice but also in the P-slice or B-slice, for generating the prediction block. However, information that can be partially omitted may exist due to the characteristic that the reference picture is current. For example, when the reference picture is the current picture, the reference picture index and the reference direction can be omitted.

In addition, when interpolation is applied, the image encoding apparatus may not be required to perform block matching up to a decimal unit due to the characteristics of an artificial image such as a computer graphic due to characteristics of an image. And a unit such as a sequence, a picture, and a slice can be set.

For example, the image encoding apparatus may not perform interpolation of reference pictures used for inter-view prediction according to the setting of the encoder, and may perform various settings such as not performing interpolation only when block matching is performed in the current picture . That is, the image encoding apparatus of the present embodiment can set whether interpolation of reference pictures is performed or not. At this time, it is possible to determine whether to perform interpolation on all reference pictures or some reference pictures constituting the reference picture list.

For example, the image encoding apparatus does not perform interpolation when it is unnecessary to perform block matching on a fractional unit because the characteristic of an image in which a reference block exists in an existing block is an artificial image, Lt; RTI ID = 0.0 &gt; interpolation. &Lt; / RTI &gt;

In addition, the image encoding apparatus can set whether or not block matching is applied in a reference picture interpolated on a block-by-block basis. For example, when a natural image and an artificial image are mixed, interpolation is performed on a reference picture, but when an optimal motion vector can be obtained by searching a part of an artificial image, a predetermined unit (here assumed to be an integer unit) A motion vector can be represented. In addition, when an optimal motion vector can be obtained by selectively searching a portion of a natural image, a motion vector can be expressed by another constant unit (here, 1/4 unit is assumed).

FIG. 12 is a diagram for explaining a main procedure of an image coding method according to an embodiment of the present invention with a syntax in a coding unit.

Referring to FIG. 12, curr_pic_BM_enabled_flag denotes a flag that allows block matching in the current picture. The curr_pic_BM_enabled_flag can be defined and transmitted in units of a sequence and a picture. At this time, a process of generating a prediction block by performing block matching in the current picture, It can mean the case where it operates through inter prediction. It can be assumed that cu_skip_flag, which is an inter-screen technique that does not code the residual signal, is a flag that is supported only by the P slice or the B slice excluding the I slice. In this case, when curr_pic_BM_enabled_flag is turned on, block slicing (BM) can be supported in the inter-picture prediction mode also in the I-slice.

The image encoding method according to the present embodiment can support skipping in the case of generating a prediction block through block matching on the current picture and can support skipping in the case of an in-screen technique other than block matching. It may not support skipping on I slices depending on the condition. This skipping may be determined according to the encoder setting.

For example, when the I-slice supports skipping, the predicted block may be directly connected to prediction_unit () through a specific flag if (cu_skip_flag) to restore the prediction block to the restoring block through block matching without coding the residual signal . In addition, the image encoding apparatus classifies the method of using the prediction block through block matching in the current picture into the inter-picture prediction technique, and can process the classification through pred_mode_flag, which is a specific flag.

In addition, if the pred_mode_flag is 0, the image encoding apparatus according to the present embodiment sets the prediction mode to the inter-picture prediction mode (MODE_INTER), and if the pred_mode_flag is 1, it can set the intra prediction mode (MODE_INTRA). This is an in-screen technology similar to the existing one, but it can be classified as an inter-screen technique or an in-screen technique in an I-slice to distinguish it from an existing structure. In other words, the image encoding apparatus of the present embodiment can use a temporal correlation structure without using temporal correlation in the I-slice. part_mode indicates information on the size and type of the block divided in the encoding unit.

Some syntaxes used in connection with the present embodiment are as follows.

The syntax sps_curr_pic_ref_enabled_flag in the sequence parameter may be a flag for whether to use the IBC.

The syntax pps_curr_pic_ref_enabled_flag in the picture parameter may be a flag for whether IBC is used in units of pictures.

It is also possible to determine whether to increase the NumPicTotalCurr for the current picture according to the on / off of the syntax pps_curr_pic_ref_enabled_flag.

In the process of creating the reference picture list 0, it is possible to decide whether to insert the current picture into the reference picture according to the pps_curr_pic_ref_enabled_flag.

In the process of creating the reference picture list 1, it is possible to determine whether or not the reference picture list of the current picture is added according to the pps_curr_pic_ref_enabled_flag.

According to the above configuration, it can be determined whether or not the current picture is included in the reference picture list in the inter prediction according to the above flags.

13 is a diagram illustrating an example of supporting a symmetric type partition or an asymmetric type partition as in inter picture prediction when a prediction block is generated through block matching in the current picture used in FIG. Fig.

Referring to FIG. 13, in the case of generating a prediction block through block matching in the current picture, the image encoding method according to the present exemplary embodiment may use 2N × 2N, 2N × N, N × 2N, N, or support asymmetric partitioning such as nL x 2N, nR x 2N, 2N x nU, and 2N x nD.

FIG. 14 is an exemplary diagram for explaining that 2N × 2N and N × N can be supported in inter-picture prediction (Inter) as in intra-picture prediction (Intra) in FIG. It is possible to determine various block sizes and types according to the division method of the block dividing unit.

Referring to FIG. 14, the image encoding method according to the present embodiment can support 2N × 2N and N × N as well as a prediction block form used for intra-picture prediction. This is an example in which a square shape is supported through a quadtree division method in a block division unit or a division method according to a predefined block candidate group. In the case of intra-picture prediction, a binary tree division method or a predefined block Other types of blocks can be added by adding a type, and the setting for this can be set in the encoder.

It is also possible to determine whether skip is applied to only the case of performing block matching on the current picture (ref_idx = curr), whether the skip is applied to the existing intra-picture prediction, ) Can be set in the encoder as to whether or not the predictive block having the reference picture (see FIG. Information on this can be transmitted in units of a sequence, a picture, a slice, and the like.

The subtraction unit 205 (see FIG. 2) can generate the residual block by subtracting the pixel values of the prediction block generated from the prediction unit 200 from the pixel values of the current block to be encoded, thereby deriving the pixel difference value.

The transforming unit 210 (see FIG. 2) receives the residual block, which is a difference between the current block and a prediction block generated through intra-picture prediction or inter-picture prediction, in the subtracting unit 205 and converts the residual block into a frequency domain. Through the conversion process, each pixel of the residual block corresponds to the transform coefficient of the transform block. The size and shape of the transform block may be the same or smaller than the encoding unit. In addition, the size and shape of the conversion block may be the same as or smaller than the prediction unit. The image encoding apparatus can perform conversion processing by grouping various prediction units.

The size or type of the transform block can be determined through a block divider and can support square or rectangular transform according to the block division. The block dividing operation can be affected according to the conversion related setting (size, type, etc. of the supported conversion block) supported by the encoder / decoder.

The size and the shape of each transform block are determined according to the cost of coding the candidates of the size and the shape of the transform block and the division information such as the determined image data of each transform block and the determined size and type of each transform block can be encoded .

The transform can be transformed by a one-dimensional transform matrix. For example, discrete cosine transforms (DCT), discrete cosine transforms (DST), and horizontal and vertical units may be adaptively used for each transformation matrix. The adaptive use may include, for example, determining based on various factors such as the size of the block, the type of the block, the type of the block (luminance / chrominance), the encoding mode, the prediction mode information, the quantization parameter, have.

For example, in the case of intra-picture prediction, when the prediction mode is horizontal, a DCT-based transformation matrix may be used in the vertical direction, and a DST-based transformation matrix may be used in the horizontal direction. If the prediction mode is vertical, a DCT-based transformation matrix may be used in the horizontal direction and a DST-based transformation matrix may be used in the vertical direction.

The transformation matrix is not limited to that shown in the above description. Information on this may be determined using an implicit or explicit method, and may include one or more of the factors such as the size of the block, the type of the block, the encoding mode, the prediction mode, the quantization parameter, the encoding information of the neighboring block, And the related information may be transmitted in units of a sequence, a picture, a slice, a block, or the like.

Here, considering the case of using the explicit method, when two or more transformation matrices for the horizontal and vertical directions are used as candidate groups, information on which transformation matrix is used for each direction may be sent, It is also possible to transmit information on which conversion matrix is used in the horizontal and vertical directions by putting two or more pairs as a candidate group by grouping them into one pair each of which conversion matrix is used for the vertical direction.

In addition, the partial conversion or the entire conversion can be omitted in consideration of the characteristics of the image. For example, one or both of the horizontal and vertical components may be omitted. Since intra prediction or inter prediction is not performed well, a large difference between the current block and the prediction block, that is, when the residual component is large, the encoding loss due to the conversion can be increased. At least one factor among factors such as an encoding mode, a prediction mode, a block size, a block type, a block type (luminance / chrominance), a quantization parameter, a neighboring block coding information, . This can be expressed using an implicit or explicit method according to the above conditions, and information on this can be transmitted in units of sequence, picture, slice, and the like.

The quantization unit 215 (see FIG. 2) performs quantization of the residual component transformed by the transform unit 210. The quantization parameter is determined on a block-by-block basis, and the quantization parameter can be set in units of a sequence, a picture, a slice, and a block.

For example, the quantization unit 215 may predict the current quantization parameters using one or more quantization parameters derived from neighboring blocks such as left, top, top, right top, bottom bottom, etc. of the current block.

If the quantization parameter predicted from the neighboring block does not exist, that is, if the block is located at the boundary of a picture, a slice, or the like, the quantization unit 215 calculates a difference with respect to a basic parameter transmitted in units of a sequence, a picture, Value can be output or transmitted. When the quantization parameter predicted from the neighboring block exists, the differential value may be transmitted using the quantization parameter of the block.

The priority of the block from which the quantization parameter is to be derived may be set in advance or may be transmitted in units of a sequence, a picture, a slice, or the like. The residual block can be quantized through a dead zone uniform threshold quantization (DZUTQ), a quantization weighted matrix, or an improved technique. It can be set to one or more quantization schemes as candidates and can be determined by encoding mode, prediction mode information, and the like.

For example, the quantization unit 215 may set the quantization weight matrix to be applied to inter picture coding, intra picture coding unit, or the like, or may set another weighting matrix according to the intra picture prediction mode. Assuming that the size of the block is equal to the size of the quantization block and the size of the block is M × N, the quantization weight matrix can be configured by varying the quantization coefficient for each frequency component position. The quantization unit 215 may be one of various existing quantization methods and may be used under the same setting of the encoder / decoder. Information on this can be transmitted in units of a sequence, a picture, a slice, and the like.

The inverse quantization units 220 and 315 and the inverse transform units 225 and 320 shown in FIGS. 2 and 3 may be implemented by reversing the processes in the transform unit 210 and the quantization unit 215 described above. That is, the inverse quantization unit 220 can dequantize the quantized transform coefficients generated in the quantization unit 215, and the inverse transform unit 225 can inversely transform the inversely quantized transform coefficients to generate a reconstructed residual block have.

The adders 230 and 324 shown in FIGS. 2 and 3 may generate a restored block by adding the pixel values of the restored residual block to the pixel values of the predicted block generated from the predictor. The reconstruction block may be stored in the encoding / decoding picture buffer (240, 335) and provided to the prediction unit and the filter unit.

An in-loop filter such as a Deblocking Filter, a Sample Adaptive Offset (SAO), or an Adaptive Loop Filter (ALP) may be applied to the restoring block. The deblocking filter may filter the restoration block to remove distortion between block boundaries that occurs during the encoding and decoding processes. SAO is a filtering process that restores the difference between the original image and the reconstructed image by an offset, on a pixel-by-pixel basis, with respect to the residual block. The ALF can perform filtering to minimize the difference between the prediction block and the restoration block. The ALF can perform filtering based on the comparison value between the restored block and the current block through the deblocking filter.

The entropy encoding unit 245 (see FIG. 2) can entropy encode the quantized transform coefficients through the quantization unit 215. For example, be implemented using other coding schemes than context adaptive variable length coding (CAVLC), context adaptive binary arithmetic coding (CABAC), syntax based context adaptive binary arithmetic coding (SBAC), and probability interval partitioning entropy coding PIPE) coding can be performed.

The entropy encoding unit 245 may include various types of information necessary for decoding the quantized coefficient bit stream and the encoded bit stream in the encoded data. The encoded data may include a coded block type, a quantization coefficient, a bit string in which the quantization block is coded, and information necessary for prediction. In the case of the quantization coefficient, the two-dimensional quantization coefficient can be scanned in one dimension. The quantization factor may vary depending on the characteristics of the image. In particular, in the case of the intra prediction, since the distribution of coefficients can have a specific distribution according to the prediction mode, the scanning method can be set differently.

In addition, the entropy encoding unit 245 can be set differently according to the size of a block to be encoded. The scan pattern can be preset or candidate as at least one of various patterns such as zigzag, diagonal, and raster, and can be determined by the encoding mode, the prediction mode information, and the like, Can be used. Information on this can be transmitted in units of a sequence, a picture, a slice, and the like.

The size of a quantized block (hereinafter, a quantization block) input to the entropy encoding unit 245 may be equal to or smaller than the size of the transform block. In addition, the quantization block may be divided into two or more subblocks. In the case of division, the scan pattern in the divided block may be set to be the same as or different from that of the existing quantization block.

For example, when a scan pattern of an existing quantization block is a zigzag, zigzag may be applied to all subblocks, or a zigzag pattern may be applied to a subblock located at the upper left of a block including a DC component, A diagonal pattern can be applied to other blocks. May also be determined according to the encoding mode, the prediction mode information, and the like.

The start position of the scan pattern in the entropy encoding unit 245 basically starts from the upper left corner, but may start from the upper right corner, the lower right corner, or the lower left corner depending on the characteristics of the image. Information can be transmitted in units of a sequence, a picture, a slice, or the like. As the encoding technique, an entropy encoding technique may be used, but is not limited thereto.

The inverse quantization and inverse transformation of the inverse quantization unit 225 of the inverse quantization unit 220 shown in FIGS. 2 and 3 is performed by reversing the transforming structure of the quantization unit 215 and the transformation unit 210 And the basic filter units 235 and 330 are combined.

Next, interpolation that can be employed in the image encoding apparatus of the present invention will be briefly described as follows.

In order to improve the accuracy of the prediction through block matching, interpolation is performed with a finer resolution than that of an integer unit. The interpolation method includes a DCT-IF (discrete cosine transform based interpolation filter) technique. For example, a DCT-IF technique is used as an interpolation method in HEVC (high efficiency video coding). For example, pixels are generated in units of 1/2 and 1/4 between integers to interpolate reference pictures, Matching can be performed to generate a prediction block.

Table 1 and Table 2 show the filter coefficients used for the luminance component and the chrominance component, respectively. For the luminance components, 8-tap and 4-tap DCT-IF filters are used for chrominance components. For color difference components, filters can be applied differently according to the color format. In the case of 4: 2: 0 of YCbCr, the filter shown in Table 2 can be applied. In the case of 4: 4: 4, the filter shown in Table 1 and other filters can be applied. : In case of 2, a horizontal 1D 4-tap filter as shown in Table 2 and a vertical 1D 8-tap filter as shown in Table 1 can be applied.

FIG. 15 is a diagram for explaining a process of performing a horizontal 1D filter on pixels located at positions a, b, and c of an image (assuming x) in the image encoding method according to an embodiment of the present invention.

As shown in FIG. 15, a horizontal 1D filter can be applied to a sub-pixel located at a position (assumed to be x) of a, b, c between the first pixel G and a second pixel H adjacent thereto have. The equation is expressed as follows.

f + f2 * F + f3 * G + f4 * H + f5 * I + f6 * J + 32/64

Next, a vertical 1D filter can be applied to the sub-pixels in positions d, h, and n (assuming y). The equation is expressed as follows.

f / mo> 4 + M + f / mo> 5 R + f / mo> 6 + T / mo> 32 + 64 / mo>

A 2D separable filter can be applied to the center sub-pixels e, f, g, i, j, k, p, q and r. For example, in the case of the sub-pixel e, the pixels in the vertical direction with a are interpolated first, and then interpolated using the pixels. Then, a horizontal 1D filter is performed as interpolating a between G and H, and a vertical 1D filter is performed on the sub-pixels resulting from the horizontal 1D filter, thereby obtaining an e value. Similar operations can be performed on color difference signals.

The above description is only a partial explanation of interpolation. Filters other than DCT-IF can also be used, and the number of filters and the number of tabs to be applied per decimal unit can be different. For example, a coefficient or filter coefficient, such as 8-tap Kalman filter for 1/2, 6-tap Wiener filter for 1/4, 2-tap linear filter for 1/8, fixed like DCT-IF, Or may be encoded. As described above, one interpolation filter may be used for a picture, or an interpolation filter which is different for each region according to the characteristics may be used for an image. Alternatively, two or more reference pictures to which a plurality of interpolation filters are applied may be generated, have.

Different filters may be applied according to the encoding information such as the type of the reference picture, the temporal layer, the state of the reference picture (for example, whether it is the current picture) or the like. The above information can be set in units of sequence, picture, slice, etc., and can be transmitted in that unit.

Next, before describing in detail the improved techniques for motion estimation, motion compensation, and motion prediction that can be employed in the image encoding method according to the present embodiment, Is defined as follows.

Motion estimation refers to a process of estimating motion of a current block to be encoded by dividing an image frame into small blocks and estimating which blocks are temporally shifted from previous or later base-delayed frames (reference frames). That is, the motion estimation is a process of finding the block most similar to the target block at the time of coding the current block to be compressed. Block-based motion estimation is a process of estimating the position of a video object or screen processing unit block (macro block, etc.) in time.

The motion compensation includes motion information (motion vector, reference picture index) for the optimal prediction block found in the motion estimation process in order to predict at least a part of the previously encoded reference image to encode the current image and to predict the current image Which means that a prediction block of the current block is generated. That is, the motion compensation is a process of making an error block as a difference between a current block to be encoded and a reference block that is found to be the most similar block.

Motion prediction means finding a motion vector at the time of coding for motion compensation. Skipping, temporal prediction, spatial prediction, and the like are main techniques of motion prediction. Skipping is because the motion of the screen is constant and the size of the motion vector predicted by the image coding apparatus is zero or the residual is small enough If it can be ignored, it means that the coding of the corresponding image block is skipped. Temporal prediction can be mainly used for inter-picture prediction, and spatial prediction or inter-view prediction can be mainly used for intra-picture prediction.

Information derived from inter-picture prediction may include information (unidirectional (L0, L1), bidirectional) for distinguishing reference picture list directions, an index for distinguishing reference pictures in the reference picture list, a motion vector, and the like. Since temporal correlation is used, motion information can be efficiently coded when the motion vector of the current block is the same as or similar to the motion vector of the neighboring block.

16 is an exemplary view of a current block and a neighboring block according to a comparative example. 17 is an exemplary view of a current block and neighboring blocks according to another comparative example. 18 is an exemplary view of a current block and a neighboring block according to yet another comparative example.

As shown in FIG. 16, the current block PU and the neighboring blocks A, B, C, D, and E are examples. At least one block among the spatially adjacent blocks of the current block, that is, the upper left block (A), the upper block (B), the upper right block (C), the left block (D) And can utilize the information.

In addition, a block F within a co-located PU located at a position corresponding to the current block in the selected reference picture or its neighboring block G can be used as a candidate, and the information can be utilized. 16 shows only two candidates, but at least one block among the upper left corner, the upper corner, the upper right corner, the left corner, the lower left corner, the lower right corner, and the lower right corner, As shown in Fig. At this time, the position included in the candidate group may be determined according to the picture type, the size of the current block, correlation with spatially adjacent candidate blocks, and the like. And the selected reference picture may mean a picture having a distance of 1 or more between pictures existing before or after the current picture.

Also, as shown in FIGS. 17 and 18, information of a specific block may be utilized according to the size and position of the current PU. A term prediction unit (PU) may refer to a basic unit for performing the prediction task of the prediction unit.

In the image coding method of the present embodiment, the area of a block adjacent spatially and temporally can be expanded as follows.

FIG. 19 is an exemplary view of a current block and neighboring blocks that can be employed in the image encoding method according to an embodiment of the present invention.

Referring to FIG. 19, in the method of selecting a motion information prediction candidate according to the present embodiment, neighboring encoding blocks of a current block, blocks co co-located or close to the current block in the reference image, -located block), or motion information of blocks located in the same space that are not adjacent to each other. Here, the blocks located in the same space but not adjacent to each other are candidates for a block including at least one other block between the current block determined based on information such as an encoding mode, a reference picture index, You can do it.

In order to reduce the amount of motion information, the image encoding method may encode motion information using a motion vector copy (MVC) or a motion vector prediction (MVP).

The motion vector copy is a technique for deriving a motion vector. Instead of using a motion vector difference or a reference picture index, the motion vector predictor is used as it is. This is somewhat similar to the merge mode of the HEVC, but if the mixed mode is to merge the motion vectors of spatially and temporally adjacent blocks, then the motion vector copy of the present embodiment can be used to generate motion vectors that are spatially and temporally adjacent (For example, H, I, J) are used with motion vectors. Considering this meaning, the motion vector copy is a different concept from the mixed mode and is in a higher concept. And the purpose of the motion vector prediction may be to create a minimum motion vector difference in generating the prediction block.

The motion vector copy can derive a reference direction, a reference picture index, and a motion vector predicted value from various candidate groups of blocks as shown in Fig. It is possible to select n candidate blocks from the blocks of various candidate groups and transmit the motion vector by encoding the index information of the optimal block among the candidate blocks. Here, when n is 1, the motion vector copy can obtain the motion information of the block having the highest priority according to a predetermined criterion. When n is 2 or more, the motion vector copy can be determined for each candidate group in consideration of encoding cost, which is optimal.

The information on the above-described n may be fixed to an encoder or a decoder, and may be transmitted in units of separate sequences, pictures, slices, and the like. For example, it may be determined whether n is 1 or 2 based on information of neighboring blocks. In this case, if the motion vectors obtained from the candidate block are similar to the predetermined reference, it is judged that it is useless to make two or more candidate groups, and a method such as average or median (mvA, mvB, ..., mvJ) is used A motion vector prediction value can be generated with a fixed value, and if it is not determined to be meaningless, it is possible to generate an optimal motion vector prediction value with two or more candidate groups.

Motion vector prediction can also work similar to motion vector copy. However, the criterion for priority may be the same as or different from that of the motion vector copy. In addition, the number n of candidate blocks may be the same or different. The information on this may be fixed to the encoder and / or the decoder, or may be transmitted in units of a sequence, a picture, a slice, or the like.

The setting of reference to the candidate group may be determined according to information such as the type of the current picture, temporal id, etc. Information on the candidate may be fixed or transmitted in units of a sequence, a picture, a slice, or the like . Here, the reference to the candidate group may be set using only spatially adjacent blocks A, B, C, D, and E, or may be set to be temporally adjacent to spatially adjacent blocks A, B, C, D, May include settings using blocks H, I, J or spatially separated blocks F, G from spatially adjacent blocks A, B, C, D,

The following is a description of the setting of "block matching is applicable to the current picture".

First, in the case of an I picture, for example, a spatially adjacent block may be placed in a priority order and other blocks may be set as a candidate group. For example, the availability of reference blocks can be checked in the order of E → D → C → B → A → H → I → J. The availability can be determined by the encoding mode of the candidate block, the motion information, the position of the candidate block, and the like. The motion information may include a motion vector, a reference direction, a reference picture index, and the like.

Since the current picture is an I picture, motion information exists only when the coding mode is intra prediction (hereinafter referred to as INTER) of the present embodiment. Therefore, when you see it in priority order, it is confirmed that it is INTER first. For example, if n is 3 and E is encoded as INTER, E is excluded from the candidate group and D is checked. If D is coded as INTER, it carries motion information because it performs block matching in the current picture, and adds D to the candidate group based on the motion information. Then n remains two. Then, the image encoding apparatus can check the priority again. So when the last three candidates are filled, we stop looking for candidates.

The availability can be used not only in the encoding mode but also in the case of boudnary of pictures, slices, tiles, and the like. If it is a border, availability is checked as not available. If it is determined that the candidate is the same as or similar to the already filled candidate, the corresponding block is excluded from the candidate and the reference pixel unit checks the availability of the next candidate.

Here, INTER differs from the existing inter-picture prediction (inter). That is, the INTER mode of the present embodiment differs from inter-picture prediction in which a predictive block is generated in a reference picture because a predictive block is generated in the current picture only by utilizing an inter-picture prediction (inter) structure. That is, in the encoding mode of the present embodiment, a method of block matching in the current picture can be classified into INTER mode and intra (same as existing intra).

On the other hand, the first filled candidate may not check for availability. It is also possible to check according to the situation. That is, it is possible to apply such that the availability is not checked when the range of the motion vector of the candidate block exceeds a certain reference or exceeds the boundary of a picture, slice, tile, or the like.

In the following description, the availability is checked in the order of E? D? C? B? A? H? I? J for reference blocks (see FIG. 19) The following will describe various embodiments of the process of obtaining the candidate block of FIG.

&Lt; Example 1 >

E (not coded) → D → C → B → A → H → I → J: Exclude E

E (Delete) → D (Inter) → C → B → A → H → I → J: Include D

E (delete) → D (included) → C (Inter) → B → A → H → I → J: C included

(Block C is said to be different after a similarity check.)

E (delete) → D (included) → C (included) → B (Intra) → A → H → I → J: B excluded

E (delete) → D (included) → C (included) → B (delete) → A (Inter) → H → I → J: A (Block A is different after checking similarity)

The three blocks D, C, and A can be selected as described above.

&Lt; Example 2 >

E (Inter) → D → C → B → A → H → I → J: E

E (included) → D (Inter) → C → B → A → H → I → J: Include D

(Block D is checked for similarity and is not equal or similar to the current block.)

E (included) → D (included) → C (Inter) → B → A → H → I → J: C excluded

(Block C is the same or similar after similarity check.)

E (included) → D (included) → C (delete) → B (boundary) → A → H → I → J: B excluded

E (excluded) → D (included) → C (delete) → B (delete) → A (Intra) → H → I → J: A excluded

E (included) D (included) C (delete) B (delete) A (delete) H (Intra) I I J: H (Block H is different after checking similarity)

The three blocks E, D and H can be selected as described above.

&Lt; Example 3 >

E (Intra) → D → C → B → A → H → I → J: Exclude E

E (delete) → D (Intra) → C → B → A → H → I → J: D excluded

E (Delete) → D (Delete) → C (Intra) → B → A → H → I → J: Exclude C

E (delete) → D (delete) → C (delete) → B (boundary) → A → H → I → J: B excluded

E (delete) → D (delete) → C (delete) → B (delete) → A (Intra) → H → I → J: A excluded

E (delete) → D (delete) → C (delete) → B (delete) → A (delete) → H (Intra) → I → J: H

E (delete) → D (delete) → C (delete) → B (delete) → A (delete) → H (delete) → I (Inter) → J: I

E (delete) → D (delete) → C (delete) → B (delete) → A (delete) → H (delete) → I (included) → J Come out.)

E (deletion) → D (delete) → C (deleted) → B (deleted) → A (removed) H → (deleted) → I (included) → J (included): - including (a, 0)

Since only two reference blocks are selected as the candidate blocks in the availability check of the eight reference blocks as described above, the fixed coordinates (-a, 0) set in advance to fill the number (n = 3) Can be added.

<Example 4>

Since only one reference block A is selected as the candidate block in the availability check for the eight reference blocks as described above, two preset fixed coordinates (-a, -a, 0), (0, -b) can be added as candidates.

The third embodiment and the fourth embodiment will be described as follows.

If n candidates are not found in the candidate group determined in the third embodiment, a constant candidate having fixed coordinates can be added as a motion vector candidate. That is, in the inter-picture prediction mode of the HEVC, a zero vector may be added. However, in the case of block matching in the current picture as in the present invention, since there is no (0,0) vector in the current picture, (-A, 0) can be added as shown in Fig.

In addition, in the fourth embodiment, even if one virtual zero vector is added, if the candidate number of the motion vector can not be filled through the availability check, another preset virtual zero vector (0, -b) can be added to the candidate group . Here, a of (-a, 0) and b of (0, -b) can be variously set in consideration of the horizontal or vertical size of the block.

In the above description, the similarity is checked and the candidate group is excluded. However, the similarity check may be set not to be performed. In that case, since it is not necessary to send information about which to send, if n is 2, it can be automatically determined according to a predetermined setting whether to use the average of the two values or either of them , The index information on which of the two is to be used may be omitted. That is, the motion vector prediction value index can be omitted.

In the above description, all candidates are grouped into one group to check availability. However, the present invention is not limited to such a configuration, and the candidates can be determined by dividing into two or more groups and checking the availability of candidates in each candidate group. Herein, assuming that the number n of candidate blocks is 2, the maximum number of candidates in each group can be determined to be the same or individually. For example, as in the following embodiment, one in group 1_1 and one in group 1_2 can be respectively defined. Of course, the priority of each group can also exist separately. In addition, if the availability check is performed on the group 1_1 and the group 1_2, the group 2 can be applied to fill the number of candidates if the criterion for the number of candidates of the motion vector is not satisfied. Group 2 may include a constant candidate with fixed coordinates.

For example, the candidates for group 1_1, group 1_2, and group 2, the availability check order or priority of each group, and the group classification criteria are shown below in order.

Group 1_1 = {A, B, C, I, J}, (C? B? A? I? J)

Group 1_2 = {D, E, H}, (E? D? H), left and lower blocks

Group 2 = {fixed candidate}, (-a, 0)? (0, -b), (-a, 0)

In summary, the image encoding apparatus of the present embodiment can spatially search motion vector copy candidates or motion vector prediction candidates and generate candidate groups in a predetermined fixed candidate order according to the search results.

Hereinafter, the motion vector copy (MVC) and the motion vector prediction (MVP) will be separately described. This is because whether or not the scaling process is included is different.

The motion vector prediction (MVP) will be described first.

P Picture I  B Picture  Occation

The temporal candidates (F, G) will be described in addition to the above-mentioned candidates (A, B, C, D, E, F, G, H, I and J). In this embodiment, it is assumed that candidates are spatially searched, temporally searched, mixed lists are constructed and searched, and fixed candidates are searched in the described order.

First, the candidates are prioritized and the availability is checked accordingly. It is assumed that the number of candidates (n) of the motion vectors is set to 2, and the priority is as described in parentheses.

For example, when searching spatially, you can group them into the following groups.

Group 1_1 = {A, B, C, I, J}, (C? B? A? I? J)

Group 1_2 = {D, E, H}, (D? E? H)

In this embodiment, the group 1_1 of the two groups includes the blocks immediately above, the upper left, and the upper right with respect to the current block, and the group 1_2 includes blocks immediately adjacent to the current block, , And the lower left blocks.

In another embodiment, candidate blocks of motion vectors can be spatially searched by dividing the motion vectors into three groups. The three groups can be categorized as follows.

Group 1_1 = {A, B, C}, (C? B? A)

Group 1_2 = {D, E}, (D? E)

Group 1_3 = {H, I, J}, (J? I? H)

In the present embodiment, the group 1_1 includes blocks immediately adjacent to the current block, adjacent upper left, and upper right adjacent to the current block. Group 1_2 includes blocks immediately adjacent to the immediately adjacent block on the basis of the current block And the group 1_3 includes the current block and non-contiguous blocks having one or more block intervals.

In another embodiment, the candidate blocks of the motion vector can be spatially searched by dividing the motion vectors into three groups in another manner. The three groups can be categorized as follows.

Group 1_1 = {B}

Group 1_2 = {D}

Group 1_3 = {A, C, E}, (E? C? A)

In this embodiment, the group 1_1 includes blocks located in the vertical direction with respect to the current block, the group 1_2 includes adjacent blocks positioned in the horizontal direction with respect to the current block, And includes the remaining adjacent blocks with respect to the block.

As discussed above, since there are many referenceable information such as a reference direction and a reference picture in the P picture or the B picture, the candidate group can be set accordingly. A candidate block having a current block different from that of a current block may be included in a candidate block having a different reference block from the current block or a picture of count (POC) between a reference block of a current block and a reference block of a candidate block The vector of the corresponding block may be scaled and added to the candidate group. It is also possible to add a scaled candidate group according to which picture the reference picture of the current block is. When the temporal distance between the reference picture of the current block and the reference picture of the candidate block exceeds a predetermined distance, the candidate group is excluded. If the temporal distance is less than the predetermined distance, the scaled block may be included in the candidate group.

The similarity check described above is a process of comparing and determining how similar a motion vector already included in a prediction candidate group is and a motion vector to be newly added. Depending on the definition, it can be set to be true when the x and y components are perfectly matched, or to be true when there is a difference below the theshold value range.

FIG. 20 is a block diagram illustrating a method of encoding an image according to an exemplary embodiment of the present invention. Referring to FIG. 20, when a temporal distance between a reference picture of a current block and a reference picture of a candidate block is greater than a predetermined distance, Candidate group, as shown in Fig. FIG. 21 is an exemplary view for explaining a case where a current picture is added to a prediction candidate group when a picture referred to by a current block is a picture different from the current picture in the image coding method according to an embodiment of the present invention. FIG. 22 is a diagram illustrating an example of adding a current picture to a prediction candidate group when a picture referred to by a current block is a current picture in the image coding method according to an embodiment of the present invention.

The image encoding method according to the present embodiment is characterized in that encoding is performed prior to a predetermined time (t-1, t-2, t-3, t-4, t-5) temporally with respect to a current picture of a reference time The motion information prediction candidate of the current block can be selected from the reference pictures rf1, rf2, rf3, rf4, rf5 and the current intra-picture candidate blocks encoded prior to the current block of the current picture.

Referring to FIG. 20, a method of encoding an image according to an embodiment of the present invention includes moving pictures of a current block from reference pictures rf1, rf2, rf3, rf4, and rf5 temporally encoded with respect to a current picture current (t) Information prediction candidates can be selected.

For example, in the case where the current block B_t is coded with reference to the second reference picture rf1, the left upper block referring to a picture having a difference of 3 or more from the temporal distance with reference to the second reference picture rf1, (A), scaling is not allowed and the corresponding moving vector (mvA_x ', mvA_y') is not included in the candidate group. However, the other blocks have a time difference of less than 3 centered on t-2 So it is included in the candidate group through scaling. The set of motion vectors (MVS) included in the candidate group is as follows.

MVV = mvB_x ', mvB_y', mvC_x ', mvC_y', mvD_x ', mvD_y', mvE_x ', mvE_y'

In other words, the reference whether the motion vector is included in the candidate group or not includes the reference picture of the current block or the current picture. In this way, it is possible to determine whether to add a motion vector to a candidate group through a distance between pictures, and information on the information, such as a threshold or a certain picture, can be transmitted through a sequence, a picture, a slice, have. In FIG. 20, a solid line indicates a motion vector of each block, and a dotted line indicates a scaling corresponding to a picture referred to by the current block. This can be expressed by the following equation (1).

[Equation 1]

scaled factor = (POCcurr - POCcurr_ref) / (POCcan - POCcan_ref)

In Equation (1), POCcurr indicates the POC of the current picture, POCcurr_ref indicates the POC of the picture referred to by the current block, POCcan indicates the POC of the candidate block, and POCcan_ref indicates the POC of the picture referred to by the candidate block. Here, the POC of the candidate block may indicate a POC that is the same as the current candidate if it is a spatially neighboring candidate block, and a POC that is different from the current POC when the candidate block is a co-located block temporally coincident with the same position in the frame.

For example, as shown in FIG. 21, when scaling is performed when the current block B_t and the reference picture are different, the motion vectors scaled by the motion vector, for example, {(mvA_x ', mvA_y' (mvC_x ', mvC_y'), (mvD_x ', mvD_y'), (mvE_x ', mvE_y')} may be included in the candidate group. This indicates that even if the reference picture index is different, such as the HEVC, the constraint can be made depending on what the reference picture is, unlike the general case of the candidate group.

21, when the picture referred to by the current block B_t is a picture t-1, t-2, or t-3 different from the current picture t, that is, (MvA_x, mvA_y), (mvC_x), and (mvC_x) only when the current picture is a picture different from the current picture as the current block B_t except for the candidate group when the picture of the candidate group indicates the current picture t , mvC_y), (mvD_x, mvD_y), (mvE_x, mvE_y)} can be included in the candidate group.

22, the motion vectors {(mvB_x, vB_y), (mvC_x, mvC_y)} in the current picture t can be obtained only when the picture referred to by the current block B_t is the current picture t, Lt; / RTI &gt; In other words, when pointing to the same reference picture, that is, when referring to the same picture through comparison of the reference picture indexes, the corresponding motion vector can be added to the candidate group. The reason is that since the block matching is performed in the current picture t, the motion information in the existing intra-picture prediction inter referring to another picture is not necessary. In addition, even when block matching is performed in the current picture t, when the reference picture index is different as in H.264 / AVC, it can be implemented by excluding it from the candidate group.

In this embodiment, although the condition that the reference picture indicates the current picture is taken as an example, the reference picture can be expanded to the case where the current picture is not the current picture. For example, you can use a setting such as 'Exclude blocks that use pictures that are farther than the current block'. 16, some blocks A, B, and E are excluded from the candidate group because they are farther than the temporal distance t-2.

In the present embodiment, even if the current block and the reference picture are different, they can be included in the candidate group through scaling. Assume that the number of candidates (n) of motion vectors is 2. In the following groups 1_1 and 2_2, the reference picture of the current block is not the current picture.

Group 1_1 includes blocks (A, B, C, I, J) which are not adjacent to or proximate to the current block, and the priorities for the availability check of these blocks are as described below.

Priority: C → B → A → I → J → C_s → B_s → A_s → I_s → J_s

Group 1_2 includes left and lower blocks (D, E, H) that are not adjacent to or proximate to the current block, and the priorities for the availability check of these blocks are as follows.

Priority: D → E → H → D_s → E_s → H_s

In the above priority, a block indicated by attaching subscript (s) to a block symbol refers to a block subjected to scaling.

On the other hand, in the modified example of this embodiment, the availability is checked first in a predetermined order with respect to some of the blocks A, B, C, I, J of the group 1_1 and some blocks D, E, H of the group 1_2 Next, if it is judged that it is not available to the candidate group because it is coded as Inter but is different from the reference picture of the current block, the scaling is performed according to the distance between the current picture t and the reference picture of the candidate block, It is possible to additionally check availability in accordance with a predetermined priority order with respect to (C_s, B_s, A_s, I_s, J_s, D_s, E_s, H_s). According to the availability check described above, the video encoding apparatus can determine two candidates in total by one candidate in each group, and use the motion vector of the best candidate block among the candidate groups as the motion vector prediction value of the current block.

On the other hand, when one candidate block is not obtained in one group, for example, group 1_1, it is also possible to find two candidate blocks in another group 1_2. That is, if two candidates from a spatially adjacent block are not filled, a candidate block can be found in the temporal candidate group.

For example, when the reference picture of the current block is the current picture, a co-located block located at the same position in temporally adjacent reference pictures and blocks scaled by the block may be used. These blocks G, F, G_s and F_s may have the following priority (G? F? G_s? F_s).

As with the processing of spatially contiguous candidate blocks, availability is checked to match the priority. When the distance between the reference picture of the candidate block and the reference picture of the candidate block is different from the distance between the current picture and the reference picture of the current block, the availability can be confirmed for the candidate block subjected to the scaling process. It may not be performed if the reference picture of the current block indicates the current picture.

Mixed list

It is assumed that bidirectional prediction of the current block is performed and each motion information is present in a reference picture existing in the reference picture list (L0, L1). And checks the availability according to the priority order of the preset candidates. In this case, the priority is first confirmed as being encoded in bidirectional prediction.

If each reference picture is different, scaling is performed. In the case where only the bidirectional predicted blocks are included in the candidate group and the maximum number of candidate blocks is not exceeded, the blocks coded by unidirectional prediction among the candidate blocks previously performed are put into the preliminary candidate groups, You can create a candidate for bidirectional prediction.

First, it is assumed that the motion information of the bidirectional prediction of the current block is referred to in the reference picture at L0 and at the reference picture at L1. In the case of Table 5 (a), the motion information of the first candidate block is (mvA ₁ , ref1), (mvA ₂ , ref0), and the motion information of the second candidate block is mvB ₁ ', ref1 ) And (mvB ₂ ', ref0). Where the apostrophe (') is the scaled vector. Assuming that the number of candidates after the spatial and temporal search is two, if n is 5, the unidirectionally predicted blocks can be put into the preliminary candidates according to the preset priority in the preceding step.

In Table 3 (a), since the maximum number of candidates has not been filled yet, new candidates can be added by combining scaled unidirectional candidates using the remaining motion vectors mvC, mvD, and mvE.

In Table 3 (b), the motion information of the unidirectionally predicted block is scaled according to the reference picture of the current block. Here, an example of creating a new combination with unidirectional candidates is presented, but it is possible to combine new candidates with motion information of each of the bidirectional reference pictures L0 and L1 already added. This part is not performed in situations such as unidirectional prediction. It is also not performed when the reference picture of the current block is the current picture.

Constant candidate

If a candidate block of n maximum candidates (assumed to be 2 in this embodiment) can not be constructed through the above process, a constant candidate having a preset fixed coordinate may be added. (0, 0), (-a, 0), (-2 * a, 0), (0, -b), and the number of fixed candidates Can be set.

The above fixed coordinates can be set. Through the above process, it is possible to add as a fixed candidate through processes such as average, weighted average, and median value of at least two motion vectors included in the candidate group. If n is 5 and thus far three are registered as candidates {(mvA_x, mvA_y), (mvB_x, mvB_y), (mvC_x, mvC_y)}, the fixed candidates having a predetermined priority to fill the remaining two candidates You can add a fixed candidate based on the priority of the candidate to be included. (MvA_x + mvB_x) / 2, (mvA_y + mvB_y) / 2), ((mvA_x + mvB_x + mvC_x) / 3, (mvA_y + mvB_y + mvC_y) / 3) , mvB_x, mvC_x), median (mvA_y, mvB_y, mvC_y)), and the like.

In addition, the fixed candidate can be set differently according to the reference picture of the current block. For example, when the current picture is a reference picture, fixed candidates may be set as (-a, 0), (0, -b), (-2 * a, 0), and when the current picture is not a reference picture (0,0), (-a, 0), (average (mvA_x, ...), average (mvA_y, ...)). The information can be preset in an encoder or a decoder, or can be transmitted in units of a sequence, picture, slice, or the like.

Hereinafter, embodiments using fixed candidates will be described by way of example. In the following embodiments, n is assumed to be 3.

&Lt; Example 1: n = 3 >

It is assumed that the reference picture of the current block is the current picture and the reference picture of the current block B_t is the two reference pictures rf1 and rf0. The numbers 0, 1, 2, 3, 4, and 5 in rf0, rf1, rf2, rf3, rf4, and rf5 representing the reference picture and the other picture are examples that are applied only to the present embodiment and do not specify any meaning.

(Spatial search: E → D → A → B → C → E_s → D_s → A_s → B_s → C_s)

E (Inter, rf0) → D → A → B → C → E_s → D_s → A_s → B_s → C_s: E (the reference picture of the current block is the current picture)

E (deletion) → D (Inter, rf2) → A → B → C → E_s → D_s → A_s → B_s → C_s:

E (delete) → D (delete) → A (Inter, rf1) → B → C → E_s → D_s → A_s → B_s → C_s:

E (delete) → D (delete) → A (contains) → B (Inter, rf1) → C → E_s → D_s → A_s → B_s → C_s:

E (Delete) → D (Delete) → A (Include) → B (Include) → C (Intra) → E_s → D_s → A_s → B_s → C_s: Exclude C

E (excluded) → D (delete) → A (included) → B (included) → C (delete) → E_s → D_s → A_s → B_s → C_s: E_s Not necessary)

E (delete) → D (delete) → A (contains) → B (contains) → C (delete) → E_s (delete) → D_s → A_s → B_s → C_s: D_s There is no need to compare whether the reference picture is different because the reference picture is excluded because it is scaled in this case)

&Lt; Example 2: n = 3 >

The reference picture of the current block is the current picture and the reference picture of the current block is one reference picture rf0.

(Spatial search: E → D → A → B → C → E_s → D_s → A_s → B_s → C_s)

E (Inter, rf1) → D → A → B → C → E_s → D_s → A_s → B_s → C_s: Excluding E

E (deletion) → D (Intra) → A → B → C → E_s → D_s → A_s → B_s → C_s: Excluding D

E (delete) → D (delete) → A (Inter, rf0) → B → C → E_s → D_s → A_s → B_s → C_s:

E (delete) → D (delete) → A (contains) → B (Inter, rf0) → C → E_s → D_s → A_s → B_s → C_s:

E (Delete) → D (Delete) → A (Include) → B (Include) → C (Intra) → E_s → D_s → A_s → B_s → C_s: Exclude C

E (Delete) → D (Delete) → A (Include) → B (Include) → C (Delete) → E_s → D_s → A_s → B_s → C_s: Excluding C_s from E_s

Since there are not yet three, you can apply a fixed candidate, or you can check additional candidates H, I, J, and so on.

&Lt; Example 3: n = 3 >

When the reference picture of the current block is not the current picture, the reference picture of the current block is one reference picture rf1, and the reference picture whose distance from the reference picture of the current block is 2 or more is added to the candidate group . Assuming that the number x of (t-x) of the reference picture rfx-1 is a distance from the current picture t, scaling may not be supported from (t-2).

(Spatial search: E → D → A → B → C → E_s → D_s → A_s → B_s → C_s)

E (Intra) → D → A → B → C → E_s → D_s → A_s → B_s → C_s: Excluding E

E (deletion) → D (Inter, rf2) → A → B → C → E_s → D_s → A_s → B_s → C_s:

E (delete) → D (delete) → A (Inter, rf1) → B → C → E_s → D_s → A_s → B_s → C_s:

E (delete) → D (delete) → A (included) → B (Inter, rf3) → C → E_s → D_s → A_s → B_s → C_s:

E (delete) → D (delete) → A (include) → B (delete) → C (Inter, rf2) → E_s → D_s → A_s → B_s → C_s: C excluded

E (delete) → D (delete) → A (include) → B (delete) → C (delete) → E_s (Intra) → D_s → A_s → B_s → C_s: E_s excluded

E (Delete) → D (Delete) → A (Included) → B (Delete) → C (Delete) → E_s (Delete) → D_s → A_s → B_s → C_s: D_s including similarity check and scaled motion vectors )

E (delete) → D (delete) → A (include) → B (delete) → C (delete) → E_s (delete) → D_s (included) → A_s → B_s → C_s:

E (Delete) → D (Delete) → A (Include) → B (Delete) → C (Delete) → E_s (Delete) → D_s (Included) → A_s (Delete) → B_s → C_s: 2 or more)

E (delete) → D (delete) → A (include) → B (delete) → C (delete) → E_s (delete) → D_s (included) → A_s (delete) → B_s (delete) → C_s : C_s Check similarity, use scaled motion vector)

&Lt; Example 4: n = 3 >

When the reference picture of the current block is not the current picture, it is assumed that the current picture is the specific reference picture rf1.

(Spatial search: two motion vector candidates are selected through a spatial search process)

(Temporal search: G? F? G? S? F_s)

G (Intra) → F → G_s → F_s: Except G

The scaled factor is not 1. Assume that the distance between the current picture and the reference picture of the current block is co- the distance between the picture of the located block and the reference picture of the corresponding block is different.)

G (Delete) → F (Delete) → G_s → F_s: Exclude G_s (because G is Intra)

G (deletion) → F (delete) → G_s (delete) → F_s: F_s (because it is different after checking the similarity)

Hereinafter, the motion vector copy (MVC) will be described in more detail.

P Picture I  B In picture  Explanation for

In this embodiment, it is assumed that temporal candidates (F, G) are also included. The candidate group includes A, B, C, D, E, F, G, H, I, Although the search order is not fixed, it is assumed that the MVC candidate is searched spatially, temporally searched, a mixed list is searched, and a constant candidate is added.

In other words, the above-described portion arbitrarily sets the search order like this, and does not mean that a predetermined order is used. Priority is set and availability is checked accordingly. Let n be 5, and the priority is assumed to be the same as in parentheses.

In the following description, only differences in the motion vector prediction (MVP) will be described. The MVP part can be written with the following content, with the exception of the part for scaling in the previous section. Spatial candidates can be checked for availability without omitting the scaling process. However, similar to MVC, the type of the reference picture, the distance between the current picture and the reference picture of the current block, and the like may be excluded from the candidate group.

If there is a mixed list, candidates for bidirectional prediction can be created with a combination of the candidates added up to now as shown in Table 4 below.

As shown in Table 4 (a), a candidate using the reference list LO and a candidate using the reference list L1 can be combined to add a new candidate to the motion vector candidate group. If it is not possible to fill five motion vectors in a predetermined number, as shown in Table 4 (b), the next candidate of L0 and the candidate using L1 may be combined and added to the candidate.

The above-described motion vector candidate selection process will be described as an example. In the example below, n is assumed to be 3.

&Lt; Example 5: n = 3 >

When the reference picture of the current block is not the current picture, the reference picture ref = 1 of the current block. (Spatial search: B → D → C → E → A → J → I → H)

B (Inter, rf3) → D → C → E → A → J → I → H: B included

B (inclusive) → D (Inter, rf1) → C → E → A → J → I → H: D (similarity check)

B (included) → D (included) → C (Intra) → E → A → J → I → H: Exclude C

B (included) → D (included) → C (delete) → E (Inter, rf0) → A → J → I → H:

B (inclusion) → D (inclusion) → C (delete) → E (delete) → A (Inter, rf3) → J → I → H: A similarity check

&Lt; Example 6: n = 3 >

The reference picture of the current block is the current picture, for example, as follows.

(Spatial search: B → D → C → E → A → J → I → H)

B (Inter, rf1) → D → C → E → A → J → I → H: Excluding B

B (delete) → D (Intra) → C → E → A → J → I → H: D excluded

B (delete) → D (delete) → C (Inter, rf0) → E → A → J → I → H: C included

B (delete) → D (delete) → C (included) → E (Inter, rf0) → A → J → I → H:

B (delete) → D (delete) → C (included) → E (included) → A (Inter, rf2) → J → I → H: A excluded

B (delete) → D (delete) → C (contains) → E (contains) → A (delete) → J (Inter, rf0) → I → H: J similarity check

MVP, MVC, and the like, which find candidates of the optimal motion information as described above.

In the case of the skip mode, encoding can be performed using MVC. That is, information on the optimal motion vector candidate can be encoded after the skip flag process. If there is one candidate, this part can be omitted. It is possible to encode residual components, which are the difference values between the current block and the prediction block, through a process such as transformation and quantization without encoding the motion vector difference values separately.

If it is not a skip, it is first checked whether to process motion information through MVC with a priority order, and if it is correct, information about a candidate group of an optimal motion vector can be encoded. If MVP does not handle motion information, MVP can handle motion information. In the case of MVP, information on the optimal motion vector candidates can be encoded. Here, if there is one candidate, motion information processing can be omitted. Then, information such as a difference value with respect to a motion vector of the current block, a reference direction, a reference picture index, and the like can be coded and the residual component can be obtained and then encoded through transformation and quantization.

The subsequent codec such as entropy and post-processing filtering is omitted in order to avoid redundancy with the above description.

A motion vector selection method used in the above-described image coding method is summarized as follows.

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

Referring to FIG. 23, the motion vector selection method of the image encoding method according to the present embodiment basically constructs a spatial motion vector candidate (S231), determines whether a reference picture of the current block blk exists in the current picture (S232). If the determination result is yes (YES), the spatial motion vector candidate of the current picture is added (S233). The adding step S233 indicates that an additional spatial motion vector candidate is confirmed and added to the current picture.

That is, in the present embodiment, a method of adding a motion vector to a candidate group includes: first, a motion vector of a spatially adjacent block is composed of a candidate group through MVC, MVP, and the like, and when the reference picture of the current block is a current picture, The motion information of the block in the current picture is added. Blocks such as I, J, H in Fig. 19 may correspond to this. The definitions of these blocks are not immediately adjacent to the current block, but may be referred to as blocks recently coded as INTER. These blocks can be expressed as 'additional' or 'additional' since they mean that the candidate blocks are composed of blocks having different spatial positions from the blocks obtained through MVC and MVP.

If the reference picture of the current block is not the current picture, a candidate group can be set from a block of temporally adjacent pictures. Then, a mixed list is constructed, and candidates for bidirectional prediction can be configured by a combination of the candidates added so far. In addition, a constant candidate may be included according to the reference picture of the current picture. If the current picture is a reference picture, a fixed candidate having a predetermined fixed coordinate may be added. If the current picture is not a reference picture, a fixed candidate such as (0, 0) may be included in the candidate group of the motion vector.

On the other hand, if the determination result in step S232 is NO, the temporal motion vector candidate may be searched for and added (S234).

Next, the image encoding apparatus implementing the motion vector selection method can construct a mixed list candidate including the motion vector candidates configured in at least one of the above steps (S231, S233, S234) (S235).

Next, the image encoding apparatus confirms whether the current picture is a reference picture (S236). If the current picture is a reference picture and the number of mixed list candidates is smaller than the preset number of motion vectors, The candidate can be added to the candidate group (S237).

If it is determined in step S236 that the current picture is not a reference picture, the image encoding apparatus may add (0, 0) to the candidate group as a fixed candidate (S238).

When the reference pixel forming process is completed by the motion vector selecting method, the reference pixel filtering process of FIG. 9 may be performed, but the present invention is not limited thereto. For example, when a reference pixel is constructed by a motion vector selection method and the motion prediction process is completed based on the reference pixel, the process can proceed to the interpolation step of FIG.

In the above-described embodiment, when the above-described 'method of constructing a list of neighboring blocks for deriving motion information', that is, 'motion vector candidate selection method', is used to decode a coded image, Device. &Lt; / RTI &gt; It is needless to say that the image encoding apparatus may be implemented by an image processing apparatus or an image encoding and decoding apparatus having at least one means for encoding and decoding or a component performing a function corresponding to such means.

According to the embodiments described above, it is possible to use an international codec such as MPEG-2, MPEG-4, or H.264 or other codecs in which the intra prediction technique is used, a medium using these codecs, A high-efficiency image encoding / decoding technique can be provided. In future, it is expected to be applied to the current high efficiency image coding technology (HEVC) and standard codec such as H.264 / AVC and the image processing field using intra prediction.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the present invention as defined by the following claims It can be understood that

Claims (20)

delete delete delete delete delete delete delete delete delete delete delete delete delete delete Decoding information from a picture parameter set indicating whether or not a current picture including a current block can be used as a reference picture;
And generating a reference picture list, wherein the reference picture list includes a previously decoded picture decoded prior to the current picture, and when the information indicates that the current picture can be used as a reference picture, Further comprising a picture;
Selecting a reference picture of the current block from the reference picture list based on a reference picture index; And
And obtaining a prediction sample of the current block from the reference picture based on the motion vector of the current block. 16. The method of claim 15,
Wherein the maximum allowable number of reference pictures that the reference picture list can include is determined according to whether or not the current picture is used as a reference picture. 16. The method of claim 15,
Wherein the step of generating the reference picture list includes a step of generating a reference picture list 0, wherein the reference picture list 0 includes a base-decoded picture having temporal order earlier than the current picture, And the current picture are generated in the order of the first decoded picture and the second decoded picture. 16. The method of claim 15,
Wherein the step of generating the reference picture list includes a step of generating a reference picture list 1, wherein the reference picture list 1 includes a first decoded picture having temporal order later than the current picture, And the current picture are generated in the order of the first decoded picture and the second decoded picture. delete delete

Images