Method of encoding video and apparatus for encoding video using prediction units based on encoding units defined according to tree structure, and method of decoding video and apparatus for decoding video using prediction units based on encoding units defined according to tree structure

FIELD: physics, video.

SUBSTANCE: invention relates to video encoding and decoding, which provides conversion between a spatial region and a conversion region. The method of encoding video includes breaking down a video image into encoding units having a maximum size. The images are encoded based encoding units according to depths obtained by hierarchically breaking down each maximum depth encoding unit, and based on the type of partition defined according to depths of the depth encoding units. The type of partition includes a data unit having the same size as the current encoding unit, and a partial data unit obtained by breaking down the height or width of the current encoding unit. Encoding units are determined according to encoding depths relative to each of the depth encoding units, and encoding units having a tree structure are therefore determined. The encoded data are output.

EFFECT: high efficiency of image compression and, as a result, high efficiency of encoding and decoding video.

15 cl, 15 dwg, 1 tbl

 

The technical FIELD TO WHICH the INVENTION RELATES.

The present invention relates to video encoding and decoding video that perform the conversion between spatial region and transformation.

The LEVEL of TECHNOLOGY

While continuing to be developed and delivered hardware for playback and storage of high definition video content or high quality, continues to grow the need for codec for efficient encoding or decoding of high definition video content or high quality. In traditional video codec, video is encoded according to a limited encoding method based on the macroblock having a predetermined size. Moreover, the traditional codec performs the transformation and inverse transformation on the macroblock by using a block having the same size, and thus encodes and decodes video.

DETAILED description of the INVENTION

TECHNICAL TASK

The present invention provides a video encoding and decoding video that perform the conversion between spatial region and conversion through the use of hierarchical division predictions.

TECHNICAL SOLUTION

Saglanacakt of the present invention, provides a method for decoding video by using block prediction based coding blocks having a tree structure, and the method includes the operations of: receiving a bit stream in respect of the encoded video and parsing of the bitstream; extracting information about the structure of the coding block, which specifies the size and the variable depth of the coding block, i.e. the block of data for decoding the video picture, information about a coded depth and the encoding mode in relation to the coding blocks having a tree structure mentioned pictures from the bitstream; and determining the coding blocks having a tree structure, on the basis of information about the structure of the coding block and information about the coded depth and the encoding mode, determine the type of partition on the basis of the depth of the current coding block, and decoding the said image on the basis of the coding blocks and partition type.

BENEFICIAL EFFECTS

The compression efficiency of the image can be increased, because the coding block is governed hierarchically along with consideration of the characteristics of the image when you increase the maximum size of the coding block, along with consideration of the size of the image. Since the encoder transmits the encoded video data with information about the code glubiny the encoding mode, the decoder can decode each piece of coded image data after determining at least one coded depth according to the coding blocks having a tree structure, so that may be the improved efficiency of the encoding and decoding of the image.

Description of the DRAWINGS

Fig. 1 is a structural diagram of an apparatus for encoding video by using block prediction based coding blocks having a tree structure according to a variant implementation of the present invention.

Fig. 2 is a structural diagram of an apparatus for decoding video by using block prediction based coding blocks having a tree structure according to a variant implementation of the present invention.

Fig. 3 is a diagram for describing the concept of block coding according to a variant implementation of the present invention.

Fig. 4 is a structural diagram of the encoder of the image based on the units of encoding, according to a variant implementation of the present invention.

Fig. 5 is a block diagram of the decoder of the image based on the units of encoding, according to a variant implementation of the present invention.

Fig. 6 is a diagram illustrating a deeper blocks of the coding depths and sections under option exercise nastojasih the invention.

Fig. 7 is a diagram for describing the relationship between the coding block and blocks conversion according to a variant implementation of the present invention.

Fig. 8 is a diagram for describing information about the coding units of the coding, the appropriate code depth, according to a variant implementation of the present invention.

Fig. 9 is a diagram deeper blocks of the coding depths according to a variant implementation of the present invention.

Fig. 10-12 diagram for describing the relationship between coding blocks, the blocks of prediction and blocks conversion according to a variant implementation of the present invention.

Fig. 13 is a diagram for describing the relationship between the coding block, the block prediction or section, and a conversion unit according to the information about the encoding mode according to table 1.

Fig. 14 is a flowchart illustrating a method of encoding a video by using block prediction based coding blocks having a tree structure according to a variant implementation of the present invention.

Fig. 15 is a flowchart illustrating a method of decoding a video by using block prediction based coding blocks having a tree structure according to a variant implementation of the present invention.

The BEST OPTION IMPLEMENTATION

p> According to the aspect of the present invention, a method of decoding a video by using block prediction based coding blocks having a tree structure, and the method includes the operations of: receiving a bit stream in respect of the encoded video and parsing of the bitstream; extracting information about the structure of the coding block, which specifies the size and the variable depth of the coding block, i.e. the block of data for decoding the video picture, information about a coded depth and the encoding mode in relation to the coding blocks having a tree structure mentioned pictures from the bitstream; and determining the coding blocks having a tree structure, on the basis of information about the block structure and encoding information about the coded depth and the encoding mode, determine the type of partition on the basis of the depth of the current coding block, and decoding the said image on the basis of the coding blocks and partition type.

The partition type can include the data block having the same size as the current coding block, and a partial data block obtained by partitioning one of the height and width of the current coding block.

According to another aspect of the present invention provides a method for video encoding means and the use of block prediction based on block coding having a tree structure, and the method includes the operations of: partitioning the video into one or more of the maximum coding blocks, which are blocks of coding, having a maximum size; encoding the above-mentioned image based on block coding according to the depths, which is obtained by the hierarchical partitioning each of the one or more maximum block coding depths in each of the one or more maximum coding blocks and based on the partition type, specified according to depths of the coding blocks by depth, determination block coding according to coded depths for each of the blocks of the coding depths, and thus, determination block coding having a tree structure; and output data that is encoded based on the partition type and coding blocks having a tree structure, information about a coded depth and the encoding mode, and information about the structure of the coding block, which specifies the size and the variable depth of the coding block.

According to another aspect of the present invention provides a video decoding device that includes a processor decoding the video and use the block prediction based coding blocks having a tree structure, and the device is in the decoding video includes a receiver for receiving a bit stream in respect of the encoded video, and then parse the bitstream; an extraction module for extracting information about the structure of the coding block, which specifies the size and the variable depth of the coding block, i.e. the block of data for decoding the video picture, information about a coded depth and the encoding mode in relation to the coding blocks having a tree structure mentioned pictures from the bitstream; and a decoder for determining the coding blocks having a tree structure, on the basis of information about the structure of the coding block and information about a coded depth and the encoding mode, determine the type of partition on the basis of the depth of the current coding block, and decoding the said image on the basis of the coding blocks and partition type associated with the processor for video decoding.

According to another aspect of the present invention provides a video encoding device, comprising a processor, a video coding using block prediction based coding blocks having a tree structure, with the video encoding device includes a module splitting the maximum coding blocks to split the video into one or more of the maximum coding blocks, which are blocks of coding, having a maximum size; determine the ü block coding to encode the said image on the basis of block coding according to the depths, obtained by the hierarchical partitioning each of the one or more maximum block coding depths in each of the one or more maximum coding blocks and based on the partition type, specified according to depths of the coding blocks by depth, determination block coding according to coded depths for each of the blocks of the coding depths, and thus, determine the coding blocks having a tree structure associated with the processor for video encoding; and an output module for outputting data that is encoded based on the partition type and coding blocks having a tree structure, information about a coded depth and the encoding mode, and information on the structure of the coding block, which specifies the size and the variable depth of the coding block.

According to another aspect of the present invention provides a computer-readable recording medium containing recorded thereon a program for executing the method of encoding a video by using a computer. According to another aspect of the present invention provides a computer-readable recording medium containing recorded thereon a program for executing the method of decoding a video by using a computer.

Embodiments of the INVENTIONS

p> Further, the present invention will be more fully described with reference to the accompanying drawings showing exemplary embodiments of the invention. In embodiments of the invention, the term 'image' may together indicate not only a still image but also a moving picture, such as video.

Further, the devices and methods of encoding and decoding video by using block prediction based coding blocks having a tree structure, will be described in detail with reference to Fig. 1-15.

Fig. 1 is a structural diagram of an apparatus for encoding video by using block prediction based coding blocks having a tree-like structure 100, according to a variant implementation of the present invention.

Device for encoding video by using block prediction based coding blocks having a tree structure 100 includes a module 110 splitting the maximum coding blocks, the determiner 120 of the coding block and the module 130 output. Further, for convenience of description, the device for video coding using block prediction based coding blocks having a tree-like structure 100, is indicated by reference as 'device 100 Kadirova the Oia video'.

The module 110 splitting the maximum coding blocks can break the current picture based on the maximum coding block for the current picture image. If the current image is greater than the maximum block coding, the image data of the current picture may be divided into at least one maximum coding block. The maximum coding block according to a variant implementation of the present invention may be a data block having a size of 32x32, 64x64, 128x128, 256x256 etc, while the shape of the data block is a square with width and height as 2 squares. The image data may be generated in the specifier 120 coding block according to at least one maximum coding block.

The coding block according to a variant implementation of the present invention can be characterized by a maximum size and depth. Depth refers to the number of times the coding block subjected to spatially split from the maximum coding block and, as the depth increases, the deeper coding blocks according to depths may be split from the maximum coding block to the minimum coding block. The depth of the maximum coding block is the top depth, and the depth of the minimum block coding is of the lowest depth. Since the size of the coding block corresponding to each depth, decreases as increases the depth of the maximum coding block, the coding block corresponding to the upper depth, may include multiple blocks of coding corresponding to the lower depths.

As described above, the image data of the current picture is divided into a maximum coding blocks according to the maximum size of the coding block, and each of the maximum coding blocks may include more in-depth coding blocks, which are divided into the depths. Since the maximum coding block according to a variant implementation of the present invention is divided by depth, image data of a spatial area, included in the maximum coding block can be classified hierarchically according to the depths.

Can be pre-determined maximum depth and the maximum size of the coding block, which limit the total number of times that hierarchically broken the height and width of the maximum coding block.

The identifier 120 block coding encodes at least one region of the partition obtained by splitting a region of the maximum coding block relative to depth, and determines the depth to the conclusion eventually the coded image data in at least one area of the break.

In other words, the determiner 120 of the coding block determines a coded depth, by encoding the image data in the deeper units of the coding depths, according to the maximum coding block of the current picture, and selecting a depth of less encoding error. Thus, in conclusion, the displayed encoded image data of the coding block corresponding to a particular code depth. In addition, the coding blocks corresponding to the coded depth may be considered as a coded block encoding.

Certain coded depth and the coded image data according to a coded depth are given in the module 130 output.

Image data in the maximum coding block is encoded based on deeper coding blocks corresponding to at least one depth equal to or below the maximum depth, and encodes the image data are compared based on each deeper coding blocks. Depth, having smaller error coding may be chosen after comparison of error coding deeper coding blocks. At least one coded depth may be chosen for each maximum coding block.

The maximum size of the coding block is divided as to the to the coding block is hierarchically divided by depth, and as an increasing number of coding blocks. Moreover, even if the coding blocks correspond to the same depth in one maximum coding block, whether to split each of the coding blocks corresponding to the same depth to a lower depth is determined by measuring the error coding image data for each coding block separately. Accordingly, even when the image data included in one maximum coding block, the image data are divided into areas by depth, and encoding errors may vary according to the fields in one maximum coding block, and thus, the coded depth may vary by regions in the image data. Thus, one or more of the coded depths can be defined in one maximum coding block, and the image data of the maximum coding block may be split by block encoding at least one coded depth.

Accordingly, the determiner 120 of the coding block may determine the coding blocks having a tree structure, included in the maximum coding block. 'The coding blocks having a tree structure', according to a variant implementation of the present invention include block coding, for the matter of depth, specified to be coded depth, from among all of the deeper coding blocks included in the maximum coding block. The block code encoding depth can hierarchically be defined according to the depths in the same region of the maximum coding block and can independently be determined in different areas. Similarly, code depth in the current scope can be independently determined by the code depth in another area.

The maximum depth according to a variant implementation of the present invention is an index related to the number of times the split from the maximum coding block to the minimum coding block. The maximum depth according to a variant implementation of the present invention may denote the total number of times the split from the maximum coding block to the minimum coding block. For example, when the depth of the maximum coding block is set to 0, the depth of the coding block, on which the maximum coding block is divided once, can be set to 1, and the depth of the coding block, in which the maximum coding block is divided twice, can be set to 2. Here, if the maximum block coding is a coding block, in which the maximum coding block is divided four times), is there is 5 levels deep, at depths of 0, 1, 2, 3 and 4, and thus, the maximum depth may be set to 4.

Coding with prediction and the transformation can be performed according to the maximum coding block. Coding with prediction and conversion is also performed based on deeper coding blocks according to the depth equal to, or depth that is less than the maximum depth, according to the maximum coding block. The conversion can be performed according to the method of frequency conversion, orthogonal transformations or integer conversion.

As the number of deeper coding blocks is incremented whenever the maximum coding block is split according to depths, the encoding includes encoding the prediction and the transform is performed over all of the deeper coding blocks as deepening depth. For convenience of description, coding with prediction and the transformation will hereinafter be described on the basis of the coding block of the current depth, maximum block encoding.

The device 100 video encoding can choose the size or shape of the data unit for encoding the image data. In order to encode the image data, performs operations such as encoding with preds is showing, transformation and entropy encoding, and, at this time, the same data block can be used for all operations, or different blocks of data can be used for each operation.

For example, the device 100 video encoding can choose not only the coding block for encoding the image data, but also data block different from the block coding in order to perform coding with prediction over the image data in the coding block.

In order to perform coding with prediction of the maximum block coding, coding with prediction may be performed based on the coding block corresponding to the coded depth, i.e. on the basis of the coding block, which is no longer divided into coding blocks corresponding to a lower depth. Further, the coding block, which is no longer breaks up and becomes the basic unit for coding with prediction, next will be noted by reference as 'block prediction'. The partition obtained by splitting a block of prediction may include block prediction or data unit obtained by splitting at least one of the height and width of the block prediction.

For example, when the coding block 2Nx2N (where N is a positive integer) is no longer breaks up and becomes a block prediction 2Nx2N, R is Merom section can be 2Nx2N, 2NxN or Nx2N. Examples of the type of partition include symmetrical sections, which received a symmetric splitting the height or width of the block prediction, the sections obtained asymmetric splitting the height or width of the block prediction, such as 1:n or n:1, the sections that received a geometric partitioning of the block prediction, and partitions having arbitrary shapes.

The size of the type section or block of the prediction block of the coding can be determined according to whether the break above the current block coding with the current depth or lower depth.

When the partition type of the current coding block is a symmetric partition type, symmetric partition type of the current coding block may include a section having the same size as the current coding block, and a section obtained by dividing the height or width of the current coding block into two. That is, a symmetric partition type of the coding block having the size of 2Nx2N, may include sections 2Nx2N, 2NxN or Nx2N.

When the current coding block is no longer partitioned into blocks of coding a lower depth, symmetric partition type of the current coding block may include sections having the same size as the units of encoding a lower depth. That is, when the current coding block is minimalimpact encoding, which cannot be split into blocks of coding a lower depth, and which from the current maximum coding blocks, symmetric partition type of the current coding block may include not only the sections of 2Nx2N, 2NxN and Nx2N, but may also include a partition with size NxN.

Similarly, when the current coding block is a block encoding the lowest depth of the current maximum coding blocks, symmetric type splitting of the current coding block may include not only the sections of 2Nx2N, 2NxN and Nx2N, but may also include a partition with size NxN.

For example, when the coding block having the current depth and the size of 2Nx2N, broken once and, thus, divided into coding blocks having a lower depth and size NxN, intra-frame prediction and inter prediction may be performed on a coding block with size NxN, through the partition with size NxN. Thus, in order to avoid unnecessary repetition of the sequence of operations, in the structure of the hierarchical block coding according to the present variant implementation, the partition type with size NxN, may not be set for coding block having the size of 2Nx2N.

However, when the current coding block having the size of 2Nx2N is minimum is determined as being the coding block, the current coding block is no longer divided into coding blocks having a size NxN, so intra-frame prediction or inter-frame prediction can be performed on the current block coding through the use of partitions with size NxN. Thus, the partition type of the minimum coding block having the size of 2Nx2N, may include sections 2Nx2N, 2NxN, Nx2N and NxN.

The mode of the prediction unit, a prediction may be at least one of intra-frame mode, inter mode and skip mode. For example, coding with prediction in intra-frame mode and inter mode may be performed on a section of 2Nx2N, 2NxN or Nx2N.

That is, in at least one case in which the current coding block is the minimum unit of encoding, in which the current coding block is divided into blocks of coding a lower depth, and in which the current coding block is not a block encoding the lowest depth of the number of current maximum block coding intra-frame prediction and interframe prediction performed by using the NxN, may be skipped.

However, when the current coding block is the minimum unit of coding, since intra-frame prediction and inter-frame prediction can not is to be performed on the coding block of the lower depths, intra-frame prediction and inter prediction may be performed on a minimum coding block by using sections 2Nx2N, 2NxN, Nx2N and NxN.

In addition, the skip mode may be performed on a section of 2Nx2N. The coding is performed independently on the same block prediction in the coding block, thereby selecting a prediction mode that causes the least encoding error.

The device 100 video encoding can also perform the conversion on image data in the coding block not only on the basis of the coding block for encoding the image data, but also on the basis of the data block that is different from the block encoding.

In order to perform the conversion in block coding, the transform may be performed on the basis of the conversion unit, having a size less than or equal to the coding block. For example, a conversion unit for converting may include a conversion unit for intra-frame mode and a conversion unit for interframe mode.

Like the coding block having a tree structure, the conversion unit in the coding block can accordingly be split into a region with a smaller size. Thus, the residual data in the coding block can be divided according to the transformation, the ima is the current tree structure by depth conversion.

Depth conversion, indicating the number of times the split in order to achieve conversion unit by dividing the height and width of block encoding can also be set in the conversion unit. For example, in the current coding block 2Nx2N, depth conversion can be set to 0 when the size of the transform block is also 2Nx2N, can have a value of 1 when the block size conversion, therefore, is NxN, and may have a value of 2 when the size of the transform block, therefore, is N/2xN/2. That is, the conversion unit may be set according to a hierarchical tree structure according to the hierarchical characteristics of the depths of the conversion.

Information about coding according to the coding blocks corresponding to the coded depth, requires not only information about a coded depth, but also about information relevant to coding with prediction and transformation. Accordingly, the determiner 120 block coding not only determines the coded depth having the minimum encoding error, but also determines the type of partition in the block prediction mode of the prediction blocks according to the predictions and the size conversion unit to convert.

The blocks of the coding according to the tree structure of the maximum coding block and methods for the determination section, according to variants of implementation of the present invention will later be described in detail with reference to Fig. 3-13.

The identifier 120 block coding can measure the error coding deeper block coding according to depth, using random distortion optimization based on the Lagrange coefficients.

The module 130 produces output image data of the maximum coding block, which is encoded based on the at least one coded depth determined by the determiner 120 of the coding block, and information about the encoding mode according to the coded depth, bit streams.

The coded image data can be obtained by encoding the residual image data.

Information about the encoding mode according to the coded depth may include information about a coded depth, about the partition type in the block prediction, the prediction mode and the block size of the transform.

Information about the coded depth may be determined by using information about the breaking depth, which indicates whether the coding unit coding a lower depth instead of the current depth. If the current depth of the current coding block is coded depth image data in the current coding block are encoded and issued, and thus, information is provided about splitting can be determined, in order not to break the current coding block to the lower depths. Alternatively, if the current depth of the current coding block is not coded depth and the encoding is performed on the coding block lower depth and, thus, information about the partitioning can be defined in order to break the current coding block to obtain a block encoding a lower depth.

If the current depth is not coded depth and the encoding is performed on the coding block, which is divided into the coding block of the lower depths. Because at least one coding block lower depths exist in the same block as the encoding of the current depth, the coding is performed repeatedly on each coding block lower depth and, thus, the encoding can be performed recursively for coding blocks having the same depth.

Because the coding blocks having a tree structure, defined for one maximum coding block, and information about at least one coding mode determined for the coding block coded depth, information about at least one encoding mode may be determined for one maximum coding block. In addition, the depth of the image data of the maximum coding block which may be different by location, since the image data is hierarchically split according to depths, and thus, information about a coded depth and the encoding mode may be set for the image data.

Accordingly, the module 130 output may appoint encoding information about the respective coded depth and the encoding mode for at least one coding block, the block prediction and the minimum unit included in the maximum coding block. The module 130 conclusion can insert information about the respective coded depth and the corresponding encoding mode in the header of the bitstream for transmission of the coded video data, parameter set sequence (SPS), or a set of image parameters (PPS), and can output them.

The minimum unit according to a variant implementation of the present invention may be a rectangular data unit obtained by splitting the minimum coding block, which is at the bottom of the depth 4. The minimum unit according to a variant implementation of the present invention can be maximum rectangular block of data that can be included in all of the blocks in the coding blocks of the prediction blocks of the split and conversion blocks included in the maximum coding block.

For example, coding information, issued by the via module 130 o can be classified on the information about the coding units of the coding and coding by block prediction. Information about the coding units of the coding includes information on the prediction mode and information on the size of the partitions. Coding information according to the blocks of prediction may include information about the estimated direction of the interframe mode, about the index of the reference image frame-to-frame mode, the motion vector component of the chroma intra-frame mode and the interpolation method intraframe mode.

In addition, information about the structure of the coding block about the size and the variable depth of the coding block, according to certain sequences, pictures, slices, or GOP, can be inserted in the SPS, PPS, or the header of the bitstream.

Variable depth may indicate not only the maximum depth of the current coding blocks having a tree structure, but may also indicate the lowest depth of the coding block having the minimum size, the number of levels of depth or depth change.

The number of levels of depth can specify the number of levels of depth or deeper block coding according to the depths that may exist in current coding blocks having drawoval the second structure. Depth change may indicate the magnitude of the change is more profound block coding according to the depths that may exist in current coding blocks having a tree structure.

Information about the variable depth can be set according to the sequences, pictures, slices, or GOP. That is, information about the variable depth and information about the maximum size or minimum size of the coding block from among the current coding blocks having a tree structure, can be set for each of the blocks of data sequences, pictures, slices, or GOP.

Thus, the module 130 conclusion can consider the encoding information includes, as information about the structure of the coding block, at least two of information about the variable depth information about the maximum size of the coding block and information about the minimum size of the coding block may insert the information about the encoding in the header of the bitstream, i.e., SPS or PPS, and then may output the bitstream. Variable depth, maximum size and the minimum size of the coding block is determined according to the sequences, pictures, slices, or GOP, respectively. In addition, coding information, derived from the module 130 conclusion might include is the index conversion. Information about the index conversion can provide information about the structure of the transform block, which is used to convert the current coding block. Information about the index conversion may specify, is broken if the current transform block on the lower-level blocks the conversion.

In the device 100 video encoding, the deeper coding block may be a coding block obtained by dividing by two the height or width of the block encoding higher depth, which is one level higher. In other words, when the size of the coding block of the current depth is set to 2Nx2N, the size of the coding block of the lower depth is set to NxN. In addition, the coding block of the current depth, with the size of 2Nx2N, may include a maximum of 4 of the coding block of the lower depths.

Accordingly, the device 100 video encoding can generate the coding blocks having a tree structure, determining the coding blocks having the optimal shape and the optimum size for each maximum coding block based on the maximum size of the coding block and the maximum depth defined based on the characteristics of the current picture. Moreover, since the encoding can be performed on each maximal block to the financing by using any one of various prediction modes and transformations, the optimal encoding mode may be determined based on the characteristics of the coding block of the various dimensions of the image.

Thus, if an image having a high resolution or a large amount of data, encoded in the traditional macroblock, the number of macroblocks per picture excessively increases. Accordingly, increasing the number of slices of compressed data generated for each macroblock, and thus, difficult to transmit the compressed information, and the effectiveness of data compression is reduced. However, through the use of device 100 video encoding, the compression efficiency of the image can be increased, because the coding block is adjustable along with consideration of the characteristics of the image when you increase the maximum size of the coding block, along with consideration of the size of the image.

Fig. 2 is a structural diagram of an apparatus for decoding video by using block prediction based on block coding tree structure 200, according to a variant implementation of the present invention.

A device for decoding video by using block prediction based on block coding tree structure 200 includes a receiver 210, the module 220 retrieve the image data and information about codero the years and the decoder 230 of the image data. Further, for convenience of description, the device for decoding video by using block prediction based on block coding tree structure 200, indicated by reference as 'device 200 video decoding'.

Definitions of various terms, such as block coding, depth, block prediction, the conversion unit and information about the different encoding modes for various operations of the device 200 video decoding is identical to that described with reference to Fig. 1 and the device 100 video encoding.

The receiver 210 receives and parses the bitstream of the encoded video. Module 220 retrieve the image data and information about coding extracts encoded image data for each coding block of syntactically analyzed bitstream, the coding blocks are arranged hierarchically according to each maximum coding block, and outputs the extracted image data to the decoder 230 of the image data. Module 220 retrieve the image data and information about coding can extract information about the structure of the coding block about the size and the variable depth of the coding block of the current picture, information about a coded depth and the encoding mode of the at least one header, the SPS and PPS in relation t the tabernacles pictures from a received bit stream.

Module 220 retrieve the image data and information about coding can extract information about the variable depth and from the information about the maximum permissible size and information about the minimum allowable size of the coding block from among the blocks of the encoding that has a tree structure, for each of the blocks of data sequences, pictures, slices, or GOP, from the information about the encoding. The decoder 230, the image data can determine the maximum size and minimum size of the coding block from among the blocks of the encoding that has a tree structure, for each of the blocks of data sequences, pictures, slices, or GOP, through the use of at least two fragments of information about the variable depth information about the maximum size of the coding block and information about the minimum size of the coding block.

At least two pieces of information about the variable depth information about the maximum size of the coding block and information about the minimum size of the coding block, which is defined for each of pictures, clippings, or GOP, can be extracted from the information about the encoding, and the maximum size and minimum size of the current data block can be determined on the basis of the read information. Furthermore, Module 220 retrieve the image data and information about coding retrieves information about the coded depth and the encoding mode for coding blocks having a tree structure, according to each maximum coding block of syntactically analyzed bitstream. The extracted information about the coded depth and the encoding mode is issued to the decoder 230 of the image data. In other words, image data in a bit stream is divided into a maximum coding block, so that the decoder 230 decodes the image data the image data for each maximum coding block.

Information about the coded depth and the encoding mode according to maximum coding block may be set for information about at least one coding block corresponding to the coded depth and the encoding mode may include information about the partition type of the corresponding coding block corresponding to the coded depth, the prediction mode and the block size of the transform. In addition, information about breaking depths can be removed as the information about the coded depth.

In addition, the decoder 230, the image data can read information about the indexes conversion of information encoding, which is extracted from the syntactically analizirue the CSOs bit stream. The decoder 230, the image data may configure the processing block of the current coding block based on the image data and information about the index conversion retrieved by module 220 retrieve the image data and information about coding, can perform the inverse transform of the current coding block based on the transform block and, thus, can decode the coded data. In the decoding block coding can be restored current image.

Information about the coded depth and the encoding mode according to each maximum coding block extracted by the module 220 retrieve the image data and information about coding, is the information about the coded depth and the encoding mode determined for the formation of the minimum encoding errors when the encoder, such as the device 100 video encoding, re-performs coding for each deeper coding block according to the depth of each maximum coding block. Accordingly, the device 200 video decoding can restore the image decoding image data according to the coded depth and the encoding mode, which generate the minimum encoding error.

Because coding information about a coded depth and directed the mA coding can be assigned to a predefined block of data from among the corresponding coding block, block prediction and the minimum block module 220 retrieve the image data and information about coding can extract the information about the coded depth and the encoding mode according to the predefined data blocks. Predefined data blocks that are assigned to the same information about the coded depth and the encoding mode may be implied by the data blocks included in the same maximum coding block.

The decoder 230, the image data restores the current picture by decoding the image data in each maximum coding block based on the information about the coded depth and the encoding mode according to maximum coding blocks. In other words, the decoder 230, the image data can decode the coded image data based on the extracted information about the partition type, a prediction mode and a conversion unit for each coding block from among the blocks of the encoding that has a tree structure included in each maximum coding block. The sequence of decoding operations may include prediction, including intra-frame prediction and motion compensation, and inverse transform. The inverse transformation can be performed according to the method of the inverse orthogonal transformations the Finance or the inverse integer transform.

The decoder 230, the image data can perform intra-frame prediction or motion compensation according to the section, and the prediction mode of each coding block based on the information about the partition type and a prediction mode block prediction of the coding block according to coded depths.

In addition, the decoder 230, the image data may perform inverse transformation according to each processing block in the coding block, reading the conversion unit according to a tree structure, and information about the block size of the transform block coding on the coded depths, in order to perform the inverse transform according to the maximum coding blocks.

The decoder 230, the image data may define at least one code maximum depth of the current coding block by using information about breaking depths. If the information on splitting indicates that the image data is no longer broken at the current depth, current depth is coded depth. Accordingly, the decoder 230, the image data can decode the coded data of at least one coding block corresponding to each code the depth of the current maximum coding block using the information about the partition type of the block prediction mode preds the speeds and the size conversion unit for each coding block, the corresponding code depth, and generates image data of the current maximum coding block.

In other words, the blocks of data containing coded information, including the same information about the split, can be combined by detecting the set of information about the encoding assigned to a predetermined data unit from among the coding block, the block prediction and the minimum unit, and the combined data blocks can be considered as one block of data that must be decoded by the decoder 230 of the image data in the same encoding mode.

The device 200 video decoding can obtain information about at least one coding block, which generates the minimum encoding error when encoding is performed recursively for each maximum coding block, and may use the information for decoding the current picture. In other words, can be decoded by the coding blocks having a tree structure defined which is the optimal coding blocks in each maximum coding block. In addition, the maximum size of a coding block is determined based on the resolution and amount of data of the image.

Accordingly, even if image data has high resolution and large to the number of data the image data can effectively be decoded and restored through the use of a block size, coding and encoding method, which adaptively determined according to the characteristics of the image data using the information about the optimal encoding mode adopted of the encoder.

The method of determining the coding blocks having a tree structure, the block prediction and the transform block, according to a variant implementation of the present invention, will be described hereinafter with reference to Fig. 3-13.

Fig. 3 is a diagram for describing the concept of block coding according to a variant implementation of the present invention.

The size of the coding block can be expressed in terms of the width to the height, and can be set to 64x64, 32x32, 16x16 and 8x8.

In the video data 310, the resolution is set to 1920x1080, the maximum size of a coding block is set to 64, and the maximum depth is set to 2. In the video data 320, the resolution is set to 1920x1080, the maximum size of a coding block is set to 64, and the maximum depth is set to 3. In the video data 330, the resolution is set to 352x288, the maximum size of a coding block is set to 16, and the maximum depth is set to 1. The maximum depth shown in Fig. 3, indicates the total number of splits from the maximum unit to the investments to the minimum coding block.

If the resolution is high, and the amount of data is large, the maximum size of a coding block may be large, in order not only to increase the coding efficiency, but also to accurately reflect the characteristics of the image. Accordingly, the maximum size of the coding block of video data 310 and 320 having a higher resolution than video data 330 may be 64.

Since the maximum depth of the video data 310 has a value of 2, the blocks 315 encoding video data 310 may include a maximum coding block having the size of the long axis 64, and the coding blocks having a size of long axis 32 and 16, because the depth is deeper at two levels by splitting the maximum coding block twice. Meanwhile, since the maximum depth of the video data 330 is set to 1, the blocks 335 encoding video data 330 may include a maximum coding block having the size of the long axis 16, and the coding blocks having the size of the long axis 8, because the depth is deeper by one level by splitting the maximum coding block once.

Since the maximum depth of the video data 320 has a value of 3, the blocks 325 encoding video data 320 may include a maximum coding block having the size of the long axis 64, and the coding blocks having a size of long axis 32, 16 and 8, which since the depth delve into 3 levels by splitting the maximum coding block three times. As depth increases, can accurately detailed information.

Types partition having a size of 64x64, 64x32, 32x64, can be installed for coding blocks having a size of 64x64. Because the coding block having a size of 64x64, is not the minimum coding block in respect of multiple fragments of video data 310, 320 and 330, the partition type, with a size of 32x32, may not be installed.

Types partition having a size of 32x32, 32x16, and 16x32, can be installed for coding blocks having a size of 32x32. Because the coding block having a size of 32x32, is not the minimum coding block in respect of multiple fragments of video data 310, 320 and 330, the partition type, with a size of 16x16, may not be installed.

Types of partition sizes 16x16, 16x8, 8x16, can be installed for coding blocks that have a size of 16x16. Because the coding block having a size of 16x16, is the minimum block decoding in relation to the video data 310 may set the partition type, having a size of 8x8. However, the coding block having a size of 16x16, is not the minimum block decoding in respect of multiple fragments of video data 320 and 330, the partition type, having a size of 8x8, may not be installed.

In this regard, as the coding block of size 8x8 is a minimum unit on the encoding in respect of multiple fragments of video data 310, 320 and 330 may also be installed not only the types of the partition with size 8x8, 8x4 and 4x8, but, in addition, the partition type, with a size of 4x4.

Fig. 4 is a structural diagram of the encoder 400 images using block-based coding, according to a variant implementation of the present invention.

The encoder 400 image performs the operation identifier 120 block coding device 100 video encoding to encode the image data. In other words, the intra-frame predictor 410 performs intra-frame prediction on the coding blocks in intra-frame mode from among the current frame 405, and the estimator 420, the motion compensator 425 movement perform frame-to-frame estimation and motion compensation on the blocks of the interframe coding mode from among the current frame 405 by using the current frame 405 and the reference frame 495.

Data issued from the intra-frame predictor 410, the estimator 420, the motion compensator 425 movements are as quantized transform coefficients through a transformer 430, the quantizer 440. Quantized conversion coefficient restored as the data in the spatial domain through the inverse quantizer 460, and a backward Converter 470, and the recovered data in the spatial domain are displayed as the reference frame 495 p the following post-processing by module 480 release and module 490 contour filtering. Quantized conversion coefficient can be outputted as a bit stream 455 through entropy encoder 450.

In order for the encoder 400 images were used in the device 100 video encoding, all of the elements of the encoder 400 of the image, i.e. intra-frame predictor 410, the estimator 420, the motion compensator 425 motion Converter 430, the quantizer 440, the entropy encoder 450, the inverse quantizer 460, the inverse Converter 470, module 480 release and module 490 contour filtering operations on the basis of each coding block from among the blocks of the encoding that has a tree structure along with the maximum depth of each maximum coding block.

More precisely, intraframe predictor 410, the estimator 420, the motion compensator 425 motion detect the partitions and the prediction mode of each coding block from among the blocks of the encoding that has a tree structure, along with the maximum size and the maximum depth of the current maximum coding block and the inverter 430 determines the size of the transform block in each coding block from among the blocks of the encoding that has a tree structure.

Fig. 5 is a block diagram of the decoder 500 images using block-based coding, according to a variant of implementation of this image is the shadow.

The parser 510 parses the encoded image data which should be decoded, and the encoding information required for decoding from the bit stream 505. The coded image data are given as the converted quantized data by entropy decoder 520 and turned quantizer 530, and converts the quantized data is restored to the image data in the spatial domain through the inverse transformer 540.

Intraframe predictor 550 performs intra-frame prediction on the coding blocks in intra-frame mode with respect to image data in the spatial domain, and the compensator 560 movement performs motion compensation on the blocks of the interframe coding mode by using the reference frame 585.

Image data in the spatial domain, which are passed through the intra-frame predictor 550 and the compensator 560 movement, may be issued in the quality of the reconstructed frame 595 after post-processing by module 570 release and module 580 contour filtering. In addition, image data which is subjected to post-processing by module 570 release and module 580 contour filter, may be issued as a reference frame 585.

For the s to decode the image data decoder 230 of the image data of the device 200 video decoding, the decoder 500 of the image may perform operations that are performed after the parser 510.

In order for the decoder 500 images were used in the device 200 video decoding, all of the elements of the decoder 500 images, that is, the parser 510, the entropy decoder 520, the inverse quantizer 530, the inverse Converter 540, the intra-frame predictor 530, the compensator 560 motion module 570 release and module 580 contour filtering performed the operation on the basis of the coding blocks having a tree structure, for each maximum coding block.

More precisely, intraframe predictor 550 and the compensator 560 movement perform operations on the basis of sections and prediction mode for each of the coding blocks having a tree structure, and inverter 540 performs operations on the basis of the size conversion unit for each coding block.

Fig. 6 is a diagram illustrating a deeper blocks of the coding depths and sections according to a variant implementation of the present invention.

The device 100 video encoding and device 200 decodes the video using hierarchical coding blocks in order to accommodate the characteristics of the image. Maximum height, maximum width and maximum depth is @ encoding can adaptively determined according to the characteristics of the image, or can be installed by the user. The dimensions of the deeper coding blocks according to depths may be determined according to a predefined maximum size of the coding block.

In a hierarchical structure 600 block coding according to a variant implementation of the present invention, each of the maximum height and maximum width of the blocks of the encoding is set to 64, and the maximum depth is set to 3. Here, the maximum depth indicates the total number of partitions of the coding block according to the depth of the maximum coding block to the minimum coding block. When the depth increases along the vertical axis of the hierarchical structure 600, splits each of the height and width of the deeper coding block.

In addition, the block prediction and sections, which are the bases for coding with prediction of each deeper coding block shown along the horizontal axis of the hierarchical structure 600.

In other words, the block 610 coding is the maximum coding block in a hierarchical structure 600, while the depth is set to 0, and the size, i.e. the height to the width is set to 64x64. The depth increases along the vertical axis, and there are block 620 encoding with a size of 32x32 and a depth of 1, nl is to 630 encoding, have a size of 16x16 pixels and a depth of 2, block 640 coding with size 8x8 and a depth of 3. Block 640 coding with size 8x8 and a depth of 3, is the minimum coding block.

The block prediction and the sections of the coding block are arranged along the horizontal axis according to each depth. In other words, if the block 610 coding, have a size of 64x64 and depth 0, is a block to the prediction block of the prediction can be split into partitions included in block 610 encoding, then there is a section 610 having a size of 64x64, sections 612, having a size 64x32, and sections 614 having a size 32x64. Because the block 610 coding, have a size of 64x64, is not the minimum coding block, the partition having a size of 32x32, not installed.

Similarly, the block prediction unit 620 encoding with a size of 32x32 and a depth of 1 may be split into partitions included in block 620 encoding, that is, section 620, having a size of 32x32, sections 622, having a size 32x16, and sections 624, having a size 16x32. Because the block 620 encoding with a size of 32x32, is not the minimum coding block, the sections that have a size of 16x16, not installed.

Similarly, the block prediction unit 630 coding, have a size of 16x16 pixels and a depth of 2, can be split into partitions included in block 630 encoding, that is, the partition having a size of 16x16 pixels included in the block 630 is tiravanija, sections 632, having a size 16x8 and sections 634, having a size 8x16. Because the block 630 coding, have a size of 16x16, is not the minimum coding block, the partition having a size of 8x8, not installed.

In conclusion, the power prediction unit 640 coding with size 8x8 and a depth of 3, is the minimum coding block and has the lowest depth, and thus, may be split into partitions included in block 640 encoding, i.e. partition with size 8x8 included in block 640 coding sections 642 having dimensions 8x4, sections 644, having a size 4x8, and sections 646 having a size of 4x4.

In order to determine at least one coded depth coding blocks constituting the maximum block 610 encoding determiner 120 block coding device 100 performs video encoding encoding encoding for blocks corresponding to each depth, included in the maximum block 610 encoding.

A number of deeper coding blocks according to depths, which includes the data in the same range and the same size, increases as increases the depth. For example, four coding block corresponding to the depth of 2, are required to cover the data included in one coding block corresponding to the depth of 1. Accordingly, for th is would be to compare the results of coding the same data according to depths, coded each of the coding block corresponding to the depth of 1, and four blocks of the encoding corresponding to the depth of 2.

In order to perform the encoding for the current depth from a number of depths, the minimum encoding error can be selected for the current depth by performing encoding for each block prediction in the coding blocks corresponding to the current depth, along the horizontal axis of the hierarchical structure 600. Alternatively, the minimum encoding error can be found by comparing the minimum error coding depths, performing encoding for each depth as the depth deepens along the vertical axis of the hierarchical structure 600. The depth and the section having the minimum encoding error in block 610 encoding can be chosen as a coded depth and a partition type of block 610 encoding.

Fig. 7 is a diagram for describing the relationship between block 710 coding and blocks 720 conversion according to a variant implementation of the present invention.

The device 100 or 200 video encoding encodes or decodes the image coding blocks having dimensions smaller than or equal to the maximum block coding, for each maximum coding block. The block size of the transform to convert what about the encoding time can be selected on the basis of data blocks, which are not greater than the corresponding coding block.

For example, device 100 or 200 video encoding, if the block size 710 coding is 64x64, transformation can be performed through the use of blocks 720 conversion with a size of 32x32.

Moreover, the data block 710 coding, have a size of 64x64, can be encoded by performing transformations on each of the conversion blocks that have a size of 32x32, 16x16, 8x8 and 4x4, which are smaller than 64x64, and then can get a conversion unit having the least encoding error.

Fig. 8 is a diagram for describing information about the coding units of the coding, the appropriate code depth, according to a variant implementation of the present invention.

The module 130 output device 100 video encoding can encode and transmit information 800 about the partition type, information 810 about a prediction mode, and information about 820 size conversion unit for each coding block corresponding to the coded depth, information about the encoding mode.

Information 800 indicates information about the shape of the partition obtained by splitting a block of prediction of the current coding block, in this case, the partition is a data unit for encoding the prediction of the current block is encoded. For example, the current b is OK CU_0 encoding, with the size of 2Nx2N, can be broken down into any one of section 802 of size 2Nx2N, section 804, having a size 2NxN and section 806 having a size Nx2N. Here, information 800 about the partition type is set to indicate one of section 804, having a size 2NxN, and section 806 having a size Nx2N. However, when the current block CU_0 encoding of size 2Nx2N, is the minimum unit of encoding information 800 about the type section may include a section 808 of size NxN.

Information 810 indicates the prediction mode of each section. For example, information 810 may indicate the encoding mode with the prediction performed on the partition specified information 800, i.e. intra-frame mode 812, interframe mode 814 or 816 passes.

Information 820 specifies a conversion unit, which should be based, when you transform on the current coding block. For example, the conversion unit may be the first block 822 intraframe transform, the second block 824 intraframe transform, the first block 826 interframe transformation or the second block 828 interframe transformation.

Module 220 retrieve the image data and information about the encoding device 200 video decoding can extract and use information 800, 810 and 820 for decoding according to b the CMOS encoding each depth.

Fig. 9 is a diagram deeper blocks of the coding depths according to a variant implementation of the present invention.

Information about splitting can be used to specify the depth change. Information about splitting indicates, is broken if the coding block of the current depth in units of encoding a lower depth.

Block 910 prediction for coding with prediction block 900 encoding, which has a depth of 0 and the size 2N_0x2N_0, may include partitions type 912 partition with size 2N_0x2N_0, type 914 partition with size 2N_0xN_0, and type 916 partition with size N_0x2N_0. Fig. 9 illustrates only the types 912-916 section, which is obtained by a symmetric splitting unit 910 predictions, but the type section is not limited to such, and the sections of the unit 910 predictions may include asymmetric sections, the sections having an arbitrary shape, and the sections having the geometrical form.

Coding with prediction is performed again on the same partition with size 2N_0x2N_0, two partitions having a size 2N_0xN_0, and two partitions having a size N_0x2N_0, according to each type of partition. Coding with prediction in intra-frame mode and inter mode may be performed on sections having dimensions 2N_0x2N_0, N_0x2N_0 and 2N_0xN_0. Coding with prediction in the skip mode is performed only on the partition with the size 2N_0x2N_0.

If an encoding error is the smallest in one of the types 912-916 section having dimensions 2N_0x2N_0, N_0x2N_0 and 2N_0xN_0, block 910 predictions may not be split to lower depths. However, if an encoding error is the smallest in blocks 930 encoding sized N_0xN_0, the depth varies from 0 to 1, to perform the split (operation 920), and the coding is performed repeatedly over blocks 930 coding, with a depth of 1 and the size N_0xN_0 to find the minimum encoding error.

Block 940 prediction for coding with prediction block 930 coding, with a depth of 1 and the size 2N_1x2N_1 (=N_0xN_0), may include categories such 942 partition with size 2N_1x2N_1, type 944 partition with size 2N_1xN_1, and type 946 partition with size N_1x2N_1.

If an encoding error is the smallest in blocks 960 encoding sized N_2xN_2, compared to encoding errors in the types 942-946 section having dimensions 2N_1xN_1, N_1x2N_1 and 2N_1xN_1, the depth varies from 1 to 2, to perform the split (operation 950), and the coding is performed repeatedly over blocks 960 coding with depth 2 and size N_2xN_2 to find the minimum encoding error.

When the maximum depth is set to d-1, block coding according to each depth can be broken down to when the depth becomes d-1, and information on the situation of splitting can be encoded until when the depth is one of from 0 to d-2. In other words, when the coding is performed until when the depth is set to d-1, after the coding block corresponding to the depth d-2, broken in operation 970, the block 990 prediction for coding with prediction block 980 coding with depth d-1 and a size of 2N_(d-1 )x2N_(d-1)may include sections like section 922 having a size of 2N_(d-1)x2N_(d-1), type 994 partition having a size of 2N_(d-1)xN_(d-1), type 996 partition having a size of N_(d-1)x2N_(d-1)and type 998 partition having a size of N_(d-1)xN_(d-1). Coding with prediction can be performed repeatedly on the same topic, have a size of 2N_(d-1)x2N_(d-1), two partitions having a size of 2N_(d-1)xN_(d-1), two partitions having a size of N_(d-1)x2N_(d-1), four sections having a size of N_(d-1)xN_(d-1) from among the types 992-998 section search technique for partition type, having a minimum encoding error.

Even when the type 998 section has a minimum encoding error, since the maximum depth is set to d-1, block CU_(d-1) encoding, which has a depth d-1, no longer split to a lower depth, and code depth encoding for blocks constituting the current maximum block 900 encoding is defined to have the value d-1, and the partition type of the current maximum of 900 block coding can be defined with a value of N_(d-1)xN_(d-1). Yet the, since the maximum depth is set to d-1, and the minimum block 980 encoding having the lowest depth of d-1 is no longer broken down to a lower depth, information about the split for the minimum block 980 encoding is not set.

Because the 900 block coding, which has a depth of 0 and the size 2N_0x2N_0 and block 930 coding, with a depth of 1 and the size 2N_1x2N_1 are not minimal blocks of the coding unit 910 predictions for 900 block coding with size 2N_0x2N_0 may not include the partition type, having a size N_0xN_0 and block 940 prediction for a block 930 encoding sized 2N_1x2N_1 may not include the partition type, having a size N_1xN_1.

However, the block 990 prediction for coding with prediction block 980 coding with depth d-1, which is the minimum unit of encoding may include the type 998 partition having a size of N_(d-1)xN_(d-1).

Unit 999 data can be 'minimal unit' for the current maximum coding block. The minimum unit according to a variant implementation of the present invention may be a rectangular data unit obtained by splitting the minimum block 980 coding 4. Through repeated execution of the encoding device 100 video encoding can choose the depth, with a minimum encoding error, compare the first encoding errors according to depths of 900 block coding to determine a coded depth and install the appropriate type of break and the prediction mode as the encoding mode code depth.

Essentially, the minimum encoding errors according to depths compared to all of depths from 1 to d, and the depth having the least encoding error may be defined as a coded depth. Code depth, the type of partition block prediction and the prediction mode can be encoded and transmitted as information about the encoding mode. Moreover, since the coding block is divided from a depth of 0 to a coded depth, information partitioning code depth is set to 0, and the information about the splitting of depths, including the coded depth is set to 1.

Module 220 retrieve the image data and information about the encoding device 200 video decoding can extract and use the information about the coded depth and the block prediction unit 900 encoding to decoding section 912. The device 200 video decoding can determine the depth at which information about the partition is set to 0, as a coded depth, using the information on the split according to depths, and to use the information about the encoding mode code depth for decoding coding block corresponding to the coded depth.

Fig. 0-12 - diagram for describing the relationship between blocks 1010 coding blocks 1060 predictions and blocks 1070 conversion according to a variant implementation of the present invention.

Blocks 1010 coding is the coding blocks having a tree structure corresponding to the coded depth, a certain device 100 video encoding, the maximum coding block. Blocks 1060 predictions are sections of blocks of prediction of each block 1010 coding, and blocks 1070 conversion are blocks conversion of each block 1010 encoding.

When the depth of the maximum coding block has a value of 0 in block 1010 coding, depth of blocks 1012 and 1054 encoding are set to 1, the depth of blocks 1014, 1016, 1018, 1028, 1050 and 1052 encoding are set to 2, the depth of the blocks 1020, 1022, 1024, 1026, 1030, 1032 and 1048 encoding are set to 3, and the depth of the blocks 1040, 1042, 1044 and 1046 encoding are set to 4.

In block 1060 predictions, some blocks 1014, 1016, 1022, 1032, 1048, 1050, 1052 and 1054 coding are obtained by splitting blocks of coding among blocks 1010 coding. In other words, the types of the partition in blocks 1014, 1022, 1050 and 1054 coding have size 2NxN, the types of the partition in blocks 1016, 1048 and 1052 coding have size Nx2N, and the type of partition block 1032 encoding has a size of NxN. The partition type with size NxN, can ustanavlivat is, only when the unit 1032 coding is minimal coding block. The blocks of prediction blocks and sections 1010 encoding are less than or equal to each coding block.

The transformation or inverse transformation is performed on image data block 1052 coding in blocks 1070 transformation of data blocks that are smaller than the block 1052 coding. In addition, blocks 1014, 1016, 1022, 1032, 1048, 1050 and 1052 coding in blocks 1070 conversion are different from those in blocks 1060 predictions in terms of sizes and shapes. In other words, the devices 100 and 200 encoding and decoding video can perform intra-frame prediction, motion estimation, motion compensation, transform and inverse transform individually on the data block in the same block encoding.

Accordingly, the coding is performed recursively on each of the coding blocks having a hierarchical structure in each region of the maximum coding block, to determine the optimal coding block and, thus, can get the coding blocks having a recursive tree structure. Coding information may include one of the partition information regarding the block encoding type information section, information about a prediction mode, and information about the size is d conversion unit. Table 1 shows the encoding information, which can be mounted devices 100 and 200 encoding and decoding video.

Table 1
Information 0 about splitting
(Encoding on coding block having the size of 2Nx2N and +current depth of d)
Information 1 about splitting
Mode predictionsPartition typeThe block size conversionRe-encode the blocks Kadirova of having a lower depth
d+1
Intra-frame
Interframe
Pass
(Only 2Nx2N)
Symmetric partition typeAsymmetric partition typeInformation 0 about splitting unit conversionInformation 1 about splitting unit
conversion
2Nx2N
2NxN
Nx2N
NxN
2NxnU
2NxnD
nLx2N
nRx2N
2Nx2NNxN (symmetric partition type)
N/2xN/2 (asymmetric-tion type Radel)

The module 130 pickup the device 100 and the video encoding can give information about coding units of the coding having a tree structure, and module 220 retrieve the image data and information about the encoding device 200 video decoding can extract information about coding units of the coding that has a tree structure, from the received bit stream.

Information about splitting points, divided whether the current block encoding units encoding a lower depth. If information about the breaking of the current depth d is set to 0, the depth at which the current coding block is no longer divided into a lower depth is coded depth, and thus, the information about the partition type, the prediction mode and the block size of the transform can be defined for code depth. If the current coding block is additionally partitioned according to the partition information, the coding is performed independently on the four blocks of encoding the split lower depths.

The prediction mode can be one of the intraframe mode, inter mode and skip mode. Intraframe mode and inter mode may be determined for all types of partition, and the skip mode is determined only in the type section, having a size 2Nx2N.

Information about the partition type may specify a symmetric types partition having a size of 2Nx2N, 2NxN, Nx2N and NxN, which received a symmetric splitting high the s or the width of the block prediction, and asymmetric types section having dimensions 2NxnU, 2NxnD, nLx2N and nRx2N, which received an asymmetric splitting the height or width of the block prediction. Asymmetric types of break sizes 2NxnU and 2NxnD, can accordingly be obtained by dividing the height of the block prediction in 1:3 and 3:1, and asymmetric types of break sizes nLx2N and nRx2N, can accordingly be obtained by splitting the width of the block prediction in 1:3 and 3:1. Symmetric partition type NxN can be set only when the current block 2Nx2N coding is minimal coding block.

The size of the transform block may be set to be two types in intra-frame mode, and two types in interframe mode. In other words, if the information about the breaking of the conversion unit is set to 0, the block size conversion may be a 2Nx2N, which is the size of the current coding block. If information about the breaking of the conversion unit is set to 1, units conversion can be obtained by splitting the current coding block. Moreover, if the partition type of the current coding block having the size of 2Nx2N, is a symmetric partition type, block size conversions can be NxN, and if the partition type of the current coding block is an asymmetric partition type, block size conversions can the t to be N/2xN/2.

Coding information regarding the coding blocks having a tree structure may include at least one coding block corresponding to the coded depth, block prediction and the minimum unit. The coding block corresponding to the coded depth may contain at least one block prediction and the minimum unit that includes the same information about the encoding.

Accordingly, determines whether the adjacent data blocks in the same block encoding, the corresponding code depth by comparing information about the coding related data blocks. In addition, the corresponding data block corresponding to the coded depth, is determined by using information about the encoding of the data block and, thus, can determine the distribution of the coded depth of the maximum coding block.

Accordingly, if the current coding block is predicted based on the information about the coding related data blocks, information about the encoding of data blocks in the deeper coding blocks adjacent to the current coding block, can directly be treated and used.

Alternatively, if the current coding block is predicted based on the information about the encoding adjacent the @ data, data blocks adjacent to the current coding block, found using the coded information data blocks, and found adjacent coding blocks may be subjected to treatment to predict the current coding block.

Fig. 13 is a diagram for describing the relationship between the coding block, the block prediction or section, and a conversion unit according to the information about the encoding mode according to table 1.

The maximum 1300 block coding includes blocks 1302, 1304, 1306, 1312, 1314, 1316 and 1318 coding coded depths. Here, since the block 1318 coding is a coding block coded depth, information about the partitioning can be set to 0. Information about the partition type block 1318 encoding of size 2Nx2N, can be set to be one of the type section 1322 of size 2Nx2N, type 1324 partition with size 2NxN, type 1326 partition with size Nx2N, type 1332 partition with size 2NxnU, type 1334 partition with size 2NxnD, type of 1336 section, having a size nLx2N, and type 1338 partition with size nRx2N. When the block 1318 encoding of size 2Nx2N, is the minimum unit of encoding information about the partition type can be installed in type 1328 partition with size NxN.

Information about the split (flag size TU) conversion unit is the index type conversion and the size of the transform block, which corresponds to the index of conversion may vary according to the type of block prediction, or the partition type of the coding block.

When the partition type is set symmetric, i.e. type 1322, 1324, 1326, or 1328 section, block 1342 conversion of size 2Nx2N, set if information about the breaking of the conversion unit is set to 0, and sets the block 1344 conversion with size NxN, if the flag of the TU size is set to 1.

When the partition type is set to be asymmetric, i.e. the type 1332, 1334, 1336 or section 1338, block 1352 conversion of size 2Nx2N, is set if the flag of the TU size is set to 0, and sets the block 1354 transform having a size N/2xN/2, if the flag of the TU size is set to 1.

With reference to Fig. 13, the flag of the TU size is a flag having a value of either 0 or 1, but the flag of the TU size is not limited to 1 bit, and the conversion unit may be hierarchically split, having a tree-like structure, while the flag of the size of the TU increases from 0. Information about the breaking of the conversion unit can be used as an example of the index conversion.

In this case, if the information about the breaking of the conversion unit is used together with the maximum size of the transform block and the minimum block size conversion can be expressed by the block size ol the education, used in fact. The device 100 video encoding can encode information about the maximum size of the transform block, information about the minimum size conversion unit and the information about the maximum splitting of the conversion unit. Encoded information about the maximum size of the transform block, encoded information about the minimum size of the transform block and the encoded data on the maximum splitting of the conversion unit can be inserted in the SPS. The device 200 video decoding can decode video by using information on the maximum size of the transform block, information about the minimum size of the transform block and information about the maximum splitting of the transform block.

For example, (a) if the size of the current coding block is set to 64x64, and the maximum block size of the transform is set to 32x32, (a-1) the size of the transform block may be set to 32x32, when information about the breaking of the conversion unit is set to 0, (a-2) size conversion unit can be installed in 16x16, when information about the breaking of the conversion unit is set to 1, and (a-3) size conversion unit can be installed in 8x8, when information about the breaking of the conversion unit is set to 2.

In yet another example, (b if the size of the current coding block is set to 32x32, and the minimum size of the transform block is set to 32x32, (b-1) the size of the transform block may be set to 32x32, when information about the breaking of the conversion unit is set to 0, and since the size of the transform block can not be less than the size of 32x32, information about the breaking of the conversion unit may not be installed in addition.

In yet another example, (c) if the size of the current coding block is set to 64x64, and information about the maximum splitting of the conversion unit is set to 1, information about the breaking of the conversion unit may have a value of 0 or 1, and other information about the breaking of the conversion unit may not be installed.

Thus, if it is determined that the flag of the maximum TU size is 'MaxTransformSizeIndex', the minimum block size conversion is 'MinTransformSize', and the size of the transform block is 'RootTuSize'when the flag of the TU size is set to 0, then the current minimum size 'CurrMinTuSize' conversion unit, which can be defined in the current coding block may be determined according to equation (1): CurrMinTuSize = max(MinTransformSize, RootTuSize/(2MaxTransformSizeIndex)) (1)

Compared with the current size 'CurrMinTuSize' minimum conversion unit, which can be defined in the current coding block, the size of the 'RootTuSize' conversion unit, when the flag of the TU size is set to 0, may denote the maximum size of a transform block, which can be selected in the system. In equation (1), 'RootTuSize/(2MaxTransformSizeIndex)' denotes the size of the transform block, when the size of the 'RootTuSize' conversion unit, when the flag of the TU size is set to 0, laid out a number of times corresponding to the maximum flag size TU and 'MinTransformSize' denotes the minimum size of the transform. Thus, a smaller value from among the 'RootTuSize/(2MaxTransformSizeIndex)' and 'MinTransformSize' may be the current size 'CurrMinTuSize' minimum conversion unit, which can be defined in the current coding block.

According to a variant implementation of the present invention, the maximum size RootTuSize conversion unit may vary according to the type of prediction mode.

For example, if the current prediction mode is inter mode, 'RootTuSize' can be determined by using equation (2)below. In equation (2), 'MaxTransformSize' denotes the maximum size of a transform block, and 'PUSize' denotes the current block size prediction.

RootTuSize = min(MaxTransformSize, PUSize) (2)

That is, if the current prediction mode is the inter mode, the size of the 'RootTuSize' conversion unit, when the flag of the TU size is set to 0, may be a smaller value from among the max is th block size of the prediction and the current block size prediction.

If the prediction mode of the current block partitioning is intraframe mode, 'RootTuSize' can be determined by using equation (3)below. In equation (3), 'PartitionSize' denotes the size of the current block partitioning:

RootTuSize = min(MaxTransformSize, PartitionSize) (3)

That is, if the current prediction mode is the intra-frame mode, the size of the 'RootTuSize' conversion unit, when the flag of the TU size is set to 0, may be a smaller value from among the maximum size of the transform block and the size of the current block partitioning.

However, the size of 'RootTuSize' current maximum conversion unit, which varies according to the type of prediction mode in block section is only an example, and the present invention is not limited to such.

Fig. 14 is a flowchart illustrating a method of encoding a video by using block prediction based coding blocks having a tree structure according to a variant implementation of the present invention.

At operation 1210, the current picture is divided into at least one maximum coding block. The maximum depth that indicates the total number of times of possible partitioning, can be pre-specified.

At operation 1220, a coded depth to output a final result of the coding is according to at least one split pane, which is obtained by splitting a region of each maximum coding block relative to depth, is determined by the encoding at least one split pane, and identifies blocks of the coding according to the tree structure.

The maximum spatial coding block is broken whenever deepens the depth, and thus is divided into blocks of coding a lower depth. Each coding block can be split into blocks of encoding to another, lower depth, as the spatial divide independently of adjacent blocks of the encoding. The re-encoding is performed on each coding block according to the depths.

In addition, the conversion unit according to the types section having the least encoding error is determined for each deeper coding block. In order to determine a coded depth having the minimum encoding error in each maximum coding block, encoding errors can be measured and compared in all the deeper blocks of the coding depths.

When the processing block is determined, there may be a conversion unit for converting a coding block. The conversion unit according to the present variant exercise can be defined as a block of data to minimize error,nalezeno a conversion unit for converting a coding block.

In each maximum coding block, the picture is encoded based on the partition type, which is defined on the basis of the blocks of the coding depths, and the depths of the blocks of the coding depths, and the coding blocks of the coded depths are determined independently for each of the blocks of the coding depths, so that you can determine the coding blocks having a tree structure.

In at least one case in which the current coding block is no longer partitioned into blocks of coding a lower depth, in which the current coding block is the maximum coding block from the current maximum coding blocks, and in which the current coding block is a block encoding the lowest depth of the current maximum coding blocks, the partition type of the current coding block may further include a partition with the same size as the block coding a lower depth. The partition type can include symmetrical sections that are symmetrical splitting the height or width of the current coding block, the sections obtained asymmetric splitting the height or width of the current coding block, the sections that received a geometric partitioning of the current coding block, or partitions having arbitrary shapes. The coding is improving with the prediction may be performed based on the partition type and a prediction unit predicting the against the current coding block.

Accordingly, in at least one case in which the current coding block is no longer partitioned into blocks of coding a lower depth, in which the current coding block is the minimum unit of the encoding of the number of current maximum coding blocks, and in which the current coding block is a block encoding the lowest depth of the number of current maximum block coding, coding with prediction can be performed by using not only the symmetric sections that are symmetrical splitting the height or width of the current coding block, sections, obtained asymmetric splitting the height or width of the current coding block, partition, which is obtained geometric the splitting of the current coding block, and partitions having arbitrary shapes, but also through the use of the partition have the same size as the block coding lower depths.

In addition, in at least one case in which the current coding block may be divided into blocks of coding a lower depth, in which the current coding block is not a block encoding the lowest depth of the current maximum coding blocks, and in which the current block coding is a coding block itself is a low depth of the current maximum coding blocks, intra-frame prediction and interframe prediction performed by using the partition with the same size as the block coding lower depth may be skipped.

In operation 1230, the coded image data constituting the final result of the coding according to the coded depth, are given for each maximum coding block with information about the coding on the coded depth and encoding mode. Information about the encoding mode may include information about a coded depth, or information stored, the information about the partition type of the block prediction, information about the prediction mode, information about the size of a transform block, information about the index conversion, and the like.

Information about the structure of the coding block about the size and the variable depth of the coding block defined according to the block of data, such as sequences, pictures, clippings, or GOP, can be inserted in the header of the bitstream, SPS or PPS, and then may be output.

Encoded information about the encoding mode, and information about the structure of the coding block about the size and the variable depth of the coding block can be inserted in the header of the bitstream, SPS or PPS, and then can be transmitted to the decoder with the coded data is mi image.

Fig. 15 is a flowchart illustrating a method of decoding a video by using block prediction based coding blocks having a tree structure according to a variant implementation of the present invention.

At operation 1310, the bitstream of the encoded video is received and syntactically parsed.

In operation 1320, the coded image data of the current picture assigned to a maximum coding block, and information about a coded depth and the encoding mode according to maximum coding blocks, and information about the structure of the coding block about the size and the variable depth of the coding block is retrieved from syntactically analyzed bitstream. Information about the coded depth and the encoding mode information about the structure of the coding block about the size and the variable depth of the coding block, and information partitioning can be extracted from the header of the bitstream, SPS or PPS.

Code depth of each maximum coding block is deep, with a minimum encoding error in each maximum coding block. When encoding each maximum coding block, the image data is encoded based on the at least one data block received by the hierarchical partitioning each max is th code block according to the depths.

According to the information about the coded depth and the encoding mode, the maximum coding block can be divided into coding blocks having a tree structure. Each of the coding blocks having a tree structure, defined as the coding block corresponding to the coded depth, and optimally encoded in granting the minimum encoding errors. Accordingly, the efficiency of encoding and decoding image can be improved by decoding each piece of the coded image data in units of encoding after determining at least one coded depth according to block encoding.

The maximum size and the minimum size of the coding block from among the current coding blocks having a tree structure, can be determined on the basis of the information about the encoding includes at least two of information about the variable depth of the current coding blocks having a tree structure, information about the maximum size of the coding block and information about the minimum size of the coding block.

In addition, blocks conversion according to the tree structure in the coding blocks can be determined on the basis of the index of the transformation of the information about the encoding.

The partition type can be determined on the basis of opertion what are the current coding block, and decoding the prediction is performed based on the coding blocks and partition type, so that the picture can be decoded. Type breaker according to the present variant implementation may include at least one of the sections having the same size as the current coding block, partition, which is obtained by dividing the height or width of the current coding block into two symmetrical sections, which received a symmetric splitting the height or width of the current coding block, sections, obtained asymmetric splitting the height or width of the current coding block, partition, which received a geometric partitioning of the current coding block, and partitions having arbitrary shapes.

In at least one case in which the current coding block is no longer partitioned into blocks of coding a lower depth, in which the current coding block is the minimum unit of the encoding of the number of current maximum coding blocks, and in which the current coding block is a block encoding the lowest depth of the current maximum coding blocks, the partition type of the current coding block may further include a partition with the same size as the block coding lower depths.

In operation 1330, the data image is agenia each maximum coding block are decoded on the basis of the information about the coded depth and the encoding mode according to maximum coding blocks.

The maximum size and minimum block size coding can be read on the basis of information about the structure of the coding block and partition information, and thus, can be defined units having a tree structure. Information about the partition type and a prediction mode block prediction of the coding block can be read from the information about the encoding mode, and decoding a prediction can be made for block coding on the basis of the information about the partition type and a prediction mode, so that they can be decoded current block encoding.

For example, in at least one case in which the current coding block may be divided into blocks of coding a lower depth, in which the current coding block is not a block encoding the lowest depth of the current maximum coding blocks, and in which the current coding block is a block encoding the lowest depth of the number of current maximum block coding intra-frame prediction or inter-frame prediction/compensation is performed by using the partition with the same size as the block coding lower depth may be skipped.

However, in at least one case in which the current coding block bol is better not split into blocks of coding a lower depth, in which the current coding block is the minimum unit of the encoding of the number of current maximum coding blocks, and in which the current coding block is a block encoding the lowest depth of the number of current maximum block coding intra-frame prediction or inter-frame prediction/compensation can be performed by using not only the sections that have the same size as the current coding block, the sections obtained by dividing the height or width of the current coding block into two, and sections obtained asymmetric splitting the height or width of the current coding block, but also through the use of the partition have the same size as the block coding having a depth lower than the current coding block.

The playback device can play back the decoded image data stored on a storage medium or transmitted through a network.

Embodiments of the present invention can be written as computer programs and can be implemented in digital General purpose computers that execute the programs using a computer readable recording media. Examples of computer-readable recording media include magnetic storage media (n is an example, ROM, floppy disks, hard disks and so on) and optical recording media (e.g. CD-ROM or DVD).

Although this invention has been shown and described in detail with reference to its preferred embodiments of the specialists in the art should understand that it can be made various changes in form and content, without departing from the essence and scope of the invention defined by the attached claims. Preferred embodiments of should only be considered in a descriptive sense and not for purposes of limitation. Therefore, the scope of the invention is defined not by the detailed description of the invention and the attached claims, and all differences within the scope will be construed as being included in the present invention.

1. A method of decoding a video, comprising stages, which are:
receive bit stream in respect of the encoded video and syntactically analyze the bitstream;
retrieve information about the structure of the coding block, which specifies the size and the variable depth of the coding block, i.e. the block of data for decoding the video picture, information about a coded depth and the encoding mode in relation to the coding blocks having a tree structure mentioned pictures from the bitstream; and
define Blo and coding having a tree structure, on the basis of information about the structure of the coding block and information about a coded depth and the encoding mode, determine the type of partition on the basis of the depth of the current coding block, and decode the image on the basis of the coding blocks and partition type,
the type section contains the data block having the same size as the current coding block, and a partial data block obtained by partitioning one of the height and width of the current coding block.

2. The method according to claim 1, in which in at least one case in which the current coding block is no longer partitioned into blocks of coding a lower depth, in which the current coding block is the minimum unit of the encoding of the number of current maximum coding blocks, and in which the current coding block is a block encoding the lowest depth of the current maximum coding blocks, the partition type of the current coding block further comprises a section having the same size as the block coding lower depths.

3. The method according to claim 1, in which in at least one case in which the current coding block is no longer partitioned into blocks of coding a lower depth, in which the current coding block is the minimum unit of the encoding and the number of the current maximum coding blocks, and in which the current coding block is a block encoding the lowest depth of the current maximum coding blocks, the partition type of the current coding block further comprises a section having the same size as the coding block of the lower depths, and
the decoding includes a stage on which to perform intra-frame prediction or inter-frame compensation through the use of partition type.

4. The method according to claim 3, in which in at least one case in which the current coding block is divided into blocks of coding a lower depth, in which the current coding block is not the minimum coding block from the current maximum coding blocks, and in which the current coding block is not a block encoding the lowest depth of the maximum number of current blocks of the encoding, decoding further comprises a stage on which miss intra-frame prediction or inter-frame compensation, which is performed through the use of the partition have the same size as the block coding with lower depth of the current block encoding.

5. The method according to claim 1, wherein the removing includes a step in which extract the information about the partition type and a prediction mode block prediction, that is, the data block is La perform decoding with the prediction of the coding block, from the information about the encoding mode,
the decoding includes a stage on which to perform the decoding with the prediction of the coding block based on the extracted information about the partition type and a prediction mode,
the type of partition contains at least one symmetric sections, which received a symmetric splitting the height or width of the current coding block, partition, which received an asymmetric splitting the height or width of the current coding block, partition, which received a geometric splitting the height or width of the current coding block, and partitions having arbitrary shapes, and
the prediction mode includes at least one of intra-frame mode, inter mode and skip mode.

6. The method according to claim 1, in which the blocks of the coding according to the depths obtained by splitting the maximum coding block on the total number of spatial partitions according to each of the depths
the coding block is obtained from the number of blocks of the coding according to the depths, according to the coded depth,
the depth of the maximum coding block is the highest depth, the size of each block coding according to the depths of the quotient of the height and width in half a block coding according to the depth of the upper depth,
the picture is hierarchically split into Maxim the local block coding and block coding according to the depths, and the blocks of the coding independently are split according to coded depths, respectively, and
at least one coded depth determined in respect of the maximum coding block, and the maximum coding block contains at least one coding block according to at least one coded depth.

7. A method of encoding video, comprising stages, which are:
break picture video on one or more of the maximum coding blocks, which are blocks of coding, having a maximum size;
encode said picture based on block coding according to the depths, which is obtained by the hierarchical partitioning each of the one or more maximum block coding depths in each of the one or more maximum coding blocks and based on the partition type, defined according to the depth of the blocks of the coding depths, determine the units of encoding according to coded depths for each of the blocks of the coding depths, and thus, determine the coding blocks having a tree structure; and
output data that is encoded based on the partition type and coding blocks having a tree structure, information about a coded depth and the encoding mode, and information about the structure of the coding block, indicating the size and the variable depth of the coding block,
the type section contains the data block having the same size as the current coding block, and a partial data block obtained by partitioning one of the height and width of the current coding block.

8. The method according to claim 7, in which in at least one case in which the current coding block is no longer partitioned into blocks of coding a lower depth, in which the current coding block is the minimum unit of the encoding of the number of current maximum coding blocks, and in which the current coding block is a block encoding the lowest depth of the current maximum coding blocks, the partition type of the current coding block further comprises a section having the same size as the block coding lower depths.

9. The method according to claim 7, in which in at least one case in which the current coding block is no longer partitioned into blocks of coding a lower depth, in which the current coding block is the minimum unit of the encoding of the number of current maximum coding blocks, and in which the current coding block is a block encoding the lowest depth of the current maximum coding blocks, the partition type of the current coding block further comprises a section having the same size as the block Cody the Finance lower depths and
the definition block encoding contains a stage on which to perform intra-frame prediction or inter-frame compensation through the use of partition type.

10. The method according to claim 9, in which in at least one case in which the current coding block is divided into blocks of coding a lower depth, in which the current coding block is not the minimum coding block from the current maximum coding blocks, and in which the current coding block is not a block encoding the lowest depth of the current maximum coding blocks, the definition of units of encoding further comprises the step on which miss intra-frame prediction or inter-frame compensation, which perform through the use of the partition have the same size as the block coding with lower depth from the current coding block.

11. The method according to claim 7, in which the encoding contains the stage at which coding with prediction based on the information about the partition type and a prediction mode block prediction, i.e. the data block to perform coding with prediction of the current block, encoding,
at the conclusion paste information about the partition type and a prediction mode of the current coding block in the bit stream and output bi the new thread as the information about the encoding mode,
the type of partition contains at least one symmetric sections, which received a symmetric splitting the height or width of the current coding block, partition, which received an asymmetric splitting the height or width of the current coding block, partition, which received a geometric splitting the height or width of the current coding block, and partitions having arbitrary shapes, and
the prediction mode includes at least one of intra-frame mode, inter mode and skip mode.

12. The video decoding device containing the processor to decode the video, and the video decoding device includes:
a receiver for receiving a bit stream in respect of the encoded video, and then parse the bitstream;
an extraction module to extract information about the structure of the coding block, which specifies the size and the variable depth of the coding block, i.e. the block of data for decoding the video picture, information about a coded depth and the encoding mode in relation to the coding blocks having a tree structure mentioned pictures from the bitstream; and
a decoder for determining the coding blocks having a tree structure, on the basis of information about the structure of the coding block and information about a coded depth and the coded mode the I, determine the type of partition on the basis of the depth of the current coding block, and decoding the said image on the basis of the coding blocks and partition type associated with the processor decoding video,
the type section contains the data block having the same size as the current coding block, and a partial data block obtained by partitioning one of the height and width of the current coding block.

13. The video encoding device containing the processor, video encoding, the video encoding device includes:
module splitting the maximum coding blocks to split the video into one or more of the maximum coding blocks, which are blocks of coding, having a maximum size;
the determinant of block coding for encoding the said image on the basis of block coding according to the depths, which is obtained by the hierarchical partitioning each of the one or more maximum block coding depths in each of the one or more maximum coding blocks and based on the partition type, specified according to depths of the coding blocks by depth, determination block coding according to coded depths for each of the blocks of the coding depths, and thus, the identification of the units of the coding region have the tree structure, associated with the processor for video encoding; and
an output module to output data that is encoded based on the partition type and coding blocks having a tree structure, information about a coded depth and the encoding mode, and information about the structure of the coding block, which specifies the size and the variable depth of the coding block,
the type section contains the data block having the same size as the current coding block, and a partial data block obtained by partitioning one of the height and width of the current coding block.

14. Computer-readable recording medium containing recorded thereon a program for executing the method of decoding video according to claim 1 by using a computer.

15. Computer-readable recording medium having recorded thereon a program for executing the method of coding video according to claim 7, through the use of a computer.



 

Same patents:

FIELD: physics, video.

SUBSTANCE: invention relates to encoding three-dimensional video signals, and specifically to a transport format used to transport three-dimensional content. The technical result is achieved using a device which is characterised by that it includes a means of generating a stream which is structured into multiple levels: level 0, having two independent layers: a base layer containing video data of a right-side image, and a level 0 extension layer containing video data of a left-side image, or vice versa; level 1, having two independent extension layers: a level 1 first extension layer containing a depth map relating to the image of the base layer, a level 1 second extension layer containing a depth map relating to the image of the level 0 extension layer; level 2, having a level 2 extension layer containing overlapping data relating to the image of the base layer.

EFFECT: high quality of three-dimensional images with a large number of presentations used.

6 cl, 2 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to computer engineering. The image processing device includes an extraction means which performs motion compensation using as a reference frame a frame formed from a decoded image, and using a motion vector in an image that was encoded, and for extracting a motion compensation image corresponding to a predicted image from the reference frame; a means of generating an image with intra-frame prediction which performs intra-frame prediction for the current frame for which the predicted image is to be generated, and which generates an image with intra-frame prediction, which corresponds to the predicted image from a portion of the decoded image; and a means of generating a predicted image, which generates a predicted image by performing filter processing to compensate for high-frequency component shortcomings in the motion compensation image extracted by the extraction means, and an image with intra-frame prediction generated by generating an image with intra-frame prediction using correlation in a temporal direction which is included in the motion compensation image and the image with intra-frame prediction.

EFFECT: providing high encoding efficiency without increasing the volume of transmission of motion vectors in a stream.

13 cl, 32 dwg

FIELD: physics, video.

SUBSTANCE: invention relates to video encoding technology. Disclosed is a method of controlling video encoding, which encodes an input video signal by controlling the generated bit rate to prevent failure of a hypothetical buffer in a decoder. The method includes a step of successively encoding each image in a group of images in an encoding queue in accordance with a predefined encoding parameter. The group of images in the encoding queue includes a predefined number of images and is a set of successive images in the encoding queue. Further, the method includes calculating quantisation statistic of each image based on information about the quantisation parameter used to encode each image every time each image is encoded, and checking if the quantisation statistic exceeds a predefined threshold.

EFFECT: high efficiency of encoding images.

14 cl, 26 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to computer engineering. A multi-view video encoding method comprising generating a projection-synthesised image on which a projection-synthesised image is synthesised from an already encoded base projection frame, which corresponds to a target encoding frame in a target encoding image; searching for a base portion on the already encoded base frame in the target encoding image which corresponds to the projection-synthesised image, for each processing element portion, having a predefined size; calculating a correction parameter for correcting mismatch between cameras based on the projection-synthesised image for the processing element portion and the base frame for the base portion; correcting the projection-synthesised image for the processing element portion using the calculated correction parameter; and performing predictive video encoding in the target encoding projection using the corrected projection-synthesised image.

EFFECT: high efficiency of encoding/decoding multi-view video without further encoding/decoding correction parameters.

23 cl, 7 dwg

FIELD: radio engineering, communication.

SUBSTANCE: invention relates to video conferencing. A video conference server may receive video streams from each client in a video conference and may receive subscription requests from each client. The subscription requests may include requests to view video streams from specific other clients with a given resolution and/or frame rate. The video conference server may match up the received video streams with the subscription requests in order to send the subscribing clients their desired video streams. The server may also be able to request different versions of video streams from participants (e.g. different resolutions) and/or alter the video streams in order to better comply with the subscription request.

EFFECT: providing video conferencing subscriptions using multiple bit rate streams.

20 cl, 6 dwg

FIELD: physics, video.

SUBSTANCE: invention relates to video encoding and decoding. The video decoding method includes receiving and parsing a bit stream of an encoded video and decoding encoded image data for a maximum size of a coding unit based on information on coding depth and coding mode for the maximum coding unit based on a raster scanning order for the maximum coding unit and a zigzag scanning order for coding units of the maximum depth coding unit.

EFFECT: high efficiency of encoding or decoding high-resolution or high-quality video content.

15 cl, 29 dwg, 1 tbl

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to image encoding. The image predictive coding apparatus includes: a partitioning module for dividing an input image into a plurality of units; a prediction signal generating module for generating a prediction signal relative to a pixel signal which is included in the unit under consideration, which is to be processed, from a plurality of units; a residual signal generating module for generating a residual signal between the pixel signal of the unit under consideration and the generated prediction signal; a signal encoding module for generating a compressed signal by encoding the residual signal; and a storage module for decompressing the compressed signal and storing the decompressed signal as a reconstructed pixel signal.

EFFECT: high encoding efficiency by increasing pixel prediction accuracy.

16 cl, 16 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to computer engineering. A method for network-wide storage and distribution of data for internet protocol television (IPTV) comprising adding, by a live broadcast media code stream sending server of a content delivery network, an identifier of a program to which a media code stream data packet belongs and a storage identifier of the media code stream data packet into the media code stream data packet, where the identifier of the program to which the media code stream data packet belongs is a program label and the storage identifier of the media code stream data is a storage offset label; transmitting, by the live broadcast media code stream sending server, the media code stream data packet to a recording node; and storing, by the recording node, the media code stream data as a recording file according to the identifier of the program and the storage identifier; distributing and requesting, by the client terminal, the media code stream data from the edge node or the recording node; and transmitting the media code stream data to the client terminal.

EFFECT: providing seamless distribution of media code stream data.

8 cl, 5 dwg

FIELD: physics, video.

SUBSTANCE: invention relates to multiview image encoding/decoding. The technical result is increasing efficiency of encoding multiview images in which there is localised mismatch of illumination and colour between cameras, a well as reducing the amount of code. The invention provides, when dividing an encoded and decoded frame and encoding/decoding each region, generation of a prediction image not only for the processed region, but for already encoded/decoded regions adjacent to the processed region. Prediction images are generated using the same prediction method. A correction parameter is then estimated for correcting mismatch of illumination and colour from the predicted images and the decoded images of adjacent regions. An estimated correction parameter can be found even at the decoding side. Encoding is therefore unnecessary.

EFFECT: corrected prediction image is generated using estimated correction parameters in order to correct a prediction image that was generated for a processed region.

16 cl, 8 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to a recording device which stores a basic image stream and an extended image stream, obtained by encoding multiview video. The technical result is that data on the medium using the disclosed record encoding may be reproduced in a device which is incompatible with reproduction of multiview video. In an Access Unit containing basic video display, MVC header encoding is prohibited. For a display component contained in an Access Unit without a MVC header, determination is carried out such that the "view_id" parameter thereof is recognised as 0.

EFFECT: present invention can be applied to a reproducing device which is compatible with the BD-ROM standard.

7 cl, 48 dwg

FIELD: physics, video.

SUBSTANCE: invention relates to encoding three-dimensional video signals, and specifically to a transport format used to transport three-dimensional content. The technical result is achieved using a device which is characterised by that it includes a means of generating a stream which is structured into multiple levels: level 0, having two independent layers: a base layer containing video data of a right-side image, and a level 0 extension layer containing video data of a left-side image, or vice versa; level 1, having two independent extension layers: a level 1 first extension layer containing a depth map relating to the image of the base layer, a level 1 second extension layer containing a depth map relating to the image of the level 0 extension layer; level 2, having a level 2 extension layer containing overlapping data relating to the image of the base layer.

EFFECT: high quality of three-dimensional images with a large number of presentations used.

6 cl, 2 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to means of encoding and decoding images. In the method, the motion vector of a reference section has the same shape as the current section and belongs to a reference image which is different from the current image and is broken down in advance as a result of encoding with subsequent decoding into a plurality of sections. When a reference section overlaps a set of reference sections from said plurality of sections of the reference image, said motion vector of the current image section is determined based on a reference motion vector function belonging to a set of reference motion vectors associated with k overlapped reference sections.

EFFECT: high accuracy of predicting the motion vector of an image section.

15 cl, 6 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to predictive motion vector predictive encoding/decoding of moving pictures. The moving picture encoding apparatus includes a primary candidate reference motion vector determination unit which sets N primary candidate reference motion vectors, a degree of reliability calculation unit which calculates the reliability of each primary candidate reference motion vector which quantitatively represents effectiveness in motion vector prediction of the block to be decoded, using encoded or decoded picture information, a reference motion vector determination unit selects M (M<N) secondary candidate reference motion vectors in accordance with the degree of reliability of N primary candidate reference motion vectors, a motion vector encoding unit calculates a predictive motion vector of the block to be encoded using M secondary candidate reference motion vectors with high reliability.

EFFECT: improved efficiency of predicting and encoding moving pictures.

16 cl, 14 dwg

FIELD: information technology.

SUBSTANCE: method for alphabetical representation of images includes a step for primary conversion of an input image to a multi-centre scanning (MCS) format, constructed according to rules of a plane-filling curve (PFC). The initial MSC cell is a discrete square consisting of nine cells (3×3=9), having its own centre and its own four faces (sides). Scanning of the initial MSC cell is performed from the centre to the edge of the square while bypassing the other cells on a circle. The path with a bypass direction to the left from the centre of the square and then on a circle, clockwise, is the priority path for scanning and displaying images.

EFFECT: high efficiency of encoding images.

3 cl, 5 dwg

FIELD: radio engineering, communication.

SUBSTANCE: invention relates to video monitoring means. The method involves a mobile client sending a request, at the request of an external control device, to a multimedia transcoder; the multimedia transcoder receiving said request; requesting an encoded multimedia stream of said external control device from a fixed streaming media network server; transcoding the obtained encoded multimedia stream; the multimedia transcoder outputting a transcoded encoded multimedia stream to the mobile client or mobile streaming media network server, where the multimedia transcoder sets video transcoding parameters corresponding to various mobile network standards.

EFFECT: enabling a user to perform video monitoring of a control point using a mobile terminal.

10 cl, 4 dwg

FIELD: physics, computation hardware.

SUBSTANCE: invention relates to coding/decoding of picture signals. Method for variation of reference block (RFBL) with reference pixels in reference picture (I_REF) converts (TRF) reference block to first set of factors (REF (u, v,)). It changes the first set of factors (REF (u, v,)) with the help of one or several weights (TR (u, v,)) and executes the inversion (ITR) of changed. Note here that weights (TR (u, v,)) are defined by extra pixels in current picture (I_CUR) and extra reference pixels in reference picture. Application of extra pixels and extra reference pixels allows the determination of spectral weights so that they display the effects of attenuation. Particularly, if reference frame consists of two black-out frames one of which should be forecast with the help of reference frame, then assignment of weights in spectral band allows isolation of significant frame from two frames.

EFFECT: efficient coding in the case of attenuation.

10 cl, 3 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to computer engineering. The method of transmitting a stream of unencrypted images involves encoding the stream of images and sending a compressed stream of images to at least one receiving device. Before encoding, images of the unencrypted stream of images are converted via a secure reversible conversion to obtain a converted stream of images which is encoded and transmitted in place of the unencrypted stream of images. The secure reversible conversion converts each image from a sequence of unencrypted images.

EFFECT: high security of a stream of unencrypted images.

12 cl, 9 dwg

FIELD: information technology.

SUBSTANCE: method of compressing images programmed in a controller of a device, comprising: partitioning an image into one or more blocks; applying gamma conversion to each pixel of the image to generate data with the same number of bits; computing prediction values for each pixel in each block of the one or more blocks using a plurality of prediction modes; applying quantisation to each pixel of each block of the one or more blocks using a plurality of quantisation numbers; computing differential pulse code modulation (DPCM) to generate residuals of the quantised values for each of the plurality of quantisation numbers, wherein the number of bits generated for each block of the one or more blocks is equal to the bit budget; computing pulse code modulation (PCM), which includes shifting each pixel value by a fixed number of bits; selecting for each block of said one or more blocks, DPCM with a quantisation number where the best quantisation accuracy is achieved; selecting an encoding method from the DPCM with said quantisation number and PCM; and generating a bit stream containing data encoded using the selected encoding method.

EFFECT: compression without visual losses.

14 cl, 17 dwg

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to means of digitising a frame image.

EFFECT: frame digitisation with not three converters in each matrix element but with one converter in each matrix element which, during the frame period, concurrently and synchronously performs three successive conversions of colours R, G, B with 15 bits each, and image digitisation ends at the end of the frame period.

4 dwg

FIELD: information technology.

SUBSTANCE: image compression method, based on excluding a certain portion of information, wherein the information is excluded from the space domain through numerical solution of Poisson or Laplace differential equations, and subsequent estimation of the difference between the obtained solution and actual values at discrete points of the image; generating an array of boundary conditions, which includes a considerable number of equal elements which is compressed, and the image is reconstructed by solving Poisson or Laplace partial differential equations using the array of boundary conditions.

EFFECT: eliminating loss of image integrity, high efficiency of compressing images having large areas of the same tone or gradient and maintaining contrast of boundaries between different objects of an image.

2 cl, 16 dwg

FIELD: systems for encoding and decoding video signals.

SUBSTANCE: method and system for statistical encoding are claimed, where parameters which represent the encoded signal are transformed to indexes of code words, so that decoder may restore the encoded signal from aforementioned indexes of code words. When the parameter space is limited in such a way that encoding becomes inefficient and code words are not positioned in ordered or continuous fashion in accordance with parameters, sorting is used to sort parameters into various groups with the goal of transformation of parameters from various groups into indexes of code words in different manner, so that assignment of code word indexes which correspond to parameters is performed in continuous and ordered fashion. Sorting may be based on absolute values of parameters relatively to selected value. In process of decoding, indexes of code words are also sorted into various groups on basis of code word index values relatively to selected value.

EFFECT: increased efficiency of compression, when encoding parameters are within limited range to ensure ordered transformation of code word indexes.

6 cl, 3 dwg

Up!