Method and apparatus for encoding video, and method and apparatus for decoding video

0001The present invention relates to video coding and decoding, and more specifically, a video coding / decoding method for performing intra-prediction by selecting an intra-prediction mode according to the size of an intra-predicted data unit. Regarding the device.

0002In video compression methods such as MPEG (moving picture experts group) -1, MPEG-2, MPEG-4, H.264 / MPEG-4 AVC (advanced video coding), one picture is used to encode the video. Is divided into macroblocks. Then, after coding each macroblock in all the coding modes available for inter-prediction and intra-prediction, the bit ratio required to code the macroblock; and the original macroblock and the decoded macroblock. Select one coding mode according to the degree of distortion with and; and code the macroblock.

0003With the development and widespread use of hardware capable of playing and storing high resolution or high quality video content, there is an increasing need for video codecs that effectively encode and decode high resolution or high quality video content. .. According to existing video codecs, video is encoded in a limited predictive mode based on macroblocks of a predetermined size.

<p num="0004"> According to existing video codecs, video is encoded in a limited predictive mode based on macroblocks of a given size.</p>

<p num="0005"> The present invention provides a video coding / decoding method and an apparatus thereof that apply an intra prediction method having various directions based on hierarchical coding units of various sizes.</p>

<p num="0006"> According to the present invention, video coding efficiency is improved.</p>

0007<figref num="1">It is a block diagram of the image coding apparatus by one Embodiment of this invention.</figref><figref num="2">It is a block diagram of the image decoding apparatus according to one Embodiment of this invention.</figref><figref num="3">It is a figure which illustrates the hierarchical coding unit by one Embodiment of this invention.</figref><figref num="4">It is a block diagram of the video coding part based on the coding unit by one Embodiment of this invention.</figref><figref num="5">It is a block diagram of the image decoding part based on the coding unit by one Embodiment of this invention.</figref><figref num="6">It is a figure which illustrates the coding unit and the prediction unit by depth according to one Embodiment of this invention.</figref><figref num="7">It is a figure which illustrates the relationship of the coding unit and the conversion unit by one Embodiment of this invention.</figref><figref num="8">It is a figure which illustrates the coding information by depth by one Embodiment of this invention.</figref><figref num="9">It is a figure which illustrates the coding unit by depth according to one Embodiment of this invention.</figref><figref num="10">It is a figure which illustrates the relationship of the coding unit, the prediction unit and the conversion unit by one Embodiment of this invention.</figref><figref num="11">It is a figure which illustrates the relationship of the coding unit, the prediction unit and the conversion unit by one Embodiment of this invention.</figref><figref num="12">It is a figure which illustrates the relationship of the coding unit, the prediction unit and the conversion unit by one Embodiment of this invention.</figref><figref num="13">It is a figure which illustrates the relationship of the coding unit, the prediction unit and the conversion unit by the coding mode information of Table 1.</figref><figref num="14">It is a table which illustrates the number of intra prediction modes by the size of the coding unit by one Embodiment of this invention.</figref><figref num="15A">It is a figure for demonstrating an example of the intra prediction mode applied to the coding unit of a predetermined size by one Embodiment of this invention.</figref><figref num="15B">It is a figure for demonstrating an example of the intra prediction mode applied to the coding unit of a predetermined size by one Embodiment of this invention.</figref><figref num="15C">It is a figure for demonstrating an example of the intra prediction mode applied to the coding unit of a predetermined size by one Embodiment of this invention.</figref><figref num="16">It is a figure for demonstrating another example of the intra prediction mode applied to the coding unit of a predetermined size by one Embodiment of this invention.</figref><figref num="17A">It is a reference figure for demonstrating the intra prediction mode which has various directions by one Embodiment of this invention.</figref><figref num="17B">It is a reference figure for demonstrating the intra prediction mode which has various directions by one Embodiment of this invention.</figref><figref num="17C">It is a reference figure for demonstrating the intra prediction mode which has various directions by one Embodiment of this invention.</figref><figref num="18">It is a reference figure for demonstrating the bilinear mode by one Embodiment of this invention.</figref><figref num="19">It is a figure for demonstrating the process of generating the predicted value of the intra prediction mode of the present coding unit by one Embodiment of this invention.</figref><figref num="20A">It is a reference figure for demonstrating the mapping process of the intra prediction mode between coding units having different sizes by one Embodiment of this invention.</figref><figref num="20B">It is a reference figure for demonstrating the mapping process of the intra prediction mode between coding units having different sizes by one Embodiment of this invention.</figref><figref num="21">It is a reference figure for demonstrating the process of mapping the intra prediction mode of a peripheral coding unit to one of the representative intra prediction modes by one Embodiment of this invention.</figref><figref num="22">It is a block diagram which showed the intra prediction apparatus of the image by one Embodiment of this invention.</figref><figref num="23">It is a flowchart which showed the image coding method by one Embodiment of this invention.</figref><figref num="24">It is a flowchart which showed the image decoding method by one Embodiment of this invention.</figref><figref num="25">It is a figure for demonstrating the relationship between the peripheral pixel located on the extension line having the direction of (dx, dy), and the present pixel by this invention.</figref><figref num="26">According to the present invention, it is a figure for demonstrating the change of the peripheral pixel located on the extension line having the direction of (dx, dy) depending on the position of the present pixel.</figref><figref num="27">It is a figure for demonstrating the method of determining the intra prediction mode direction by another embodiment of this invention.</figref><figref num="28">It is a figure for demonstrating the method of determining the intra prediction mode direction by another embodiment of this invention.</figref><figref num="29">It is a figure which illustrated the present pixel and peripheral pixel used for intra prediction by one Embodiment of this invention.</figref>

0008The video coding method according to one embodiment of the present invention is an intra prediction mode applied to the current block depending on the step of dividing the current picture into at least one block of a predetermined size and the size of the coded current block. The intra-prediction mode includes a step of determining an intra-prediction for the current block by the determined intra-prediction mode, and the intra-prediction mode is centered on each pixel inside the current block.^-1It is characterized by including a prediction mode which is located on an extension line having an angle of (dy / dx) (dx, dy is an integer) or makes a prediction using pixels of peripheral blocks close to the extension line.

0009The video decoding method according to the embodiment of the present invention extracts the intra prediction mode information applied to the current block to be decoded from the bit stream and the step of dividing the current picture into at least one block of a predetermined size. The intra-prediction mode includes a step of performing an intra-prediction for the current block by the extracted intra-prediction mode, and the intra-prediction mode is centered on each pixel of the current block.^-1It is characterized by including a prediction mode in which prediction is performed using pixels of peripheral blocks located on an extension line having an angle of (dy / dx) (dx, dy is an integer) or close to the extension line.

0010The video coding apparatus according to the embodiment of the present invention is encoded by an intra prediction mode determining unit that determines the intra prediction mode to be performed according to the size of the current block to be encoded, and the determined intra prediction mode. The intra-prediction mode includes an intra-prediction execution unit that makes an intra-prediction for the block to be performed, and the intra-prediction mode is centered on each pixel inside the current block.^-1It is characterized by including a prediction mode which is located on an extension line having an angle of (dy / dx) (dx, dy is an integer) or makes a prediction using pixels of peripheral blocks close to the extension line.

0011The video decoding apparatus according to the embodiment of the present invention is described by the entropy decoding unit that extracts the intra prediction mode information applied to the current block to be decoded from the bit stream, and the extracted intra prediction mode. The intra-prediction mode includes an intra-prediction execution unit that performs intra-prediction for the current block, and the intra-prediction mode is tan centered on each pixel inside the current block.^-1It is characterized by including a prediction mode which is located on an extension line having an angle of (dy / dx) (dx, dy is an integer) or makes a prediction using pixels of peripheral blocks close to the extension line.

0012The present invention provides a computer-readable recording medium in which a program for embodying a video coding method according to an embodiment is recorded.

0013The present invention provides a computer-readable recording medium in which a program for embodying the video decoding method according to one embodiment is recorded.

0014According to the present invention, it is possible to improve the compression efficiency of video by performing intra-predictive coding in various directions for coding units of various sizes.

0015Hereinafter, a video coding device and a video decoding device, a video coding method, and a video decoding method according to a desirable embodiment of the present invention will be described with reference to the attached drawings.

0016As used herein, the "coding unit" is a coding data unit that is video data encoded on the encoder side and a decoding data unit that is video data decoded on the decoder side. Further, "coded depth" means the depth at which the coding unit is coded. In addition, the video includes a still image and a moving image. In an exemplary embodiment, "unit" may also refer to the size of a unit, depending on the context.

0017First, with reference to FIGS. 1 to 13, a video coding method and its apparatus according to an embodiment of the present invention, and a video decoding method and its apparatus will be described.

0018FIG. 1 is a block diagram of a video coding apparatus according to an embodiment of the present invention. Referring to FIG. 1, the video coding apparatus 100 according to the embodiment of the present invention includes a maximum coding unit dividing unit 110, a coding unit determining unit 120, and an output unit 130.

0019The maximum coding unit division unit 110 divides the current picture or the current slice based on the maximum coding unit, which is the maximum size coding unit. The current picture or current slice is divided into at least one maximum coding unit. The maximum coding unit according to one embodiment is a data unit having a size of 32x32, 64x64, 128x128, 256x256, etc., and may be a square data unit having a size larger than 8 in the vertical and horizontal directions, which is the square of 2. The divided video data is output to the coding unit determination unit 120 for each at least one maximum coding unit.

0020According to one embodiment of the invention, the coding unit is represented using the maximum coding unit and the depth. The maximum coding unit indicates the coding unit having the largest size among the coding units of the current picture, and the depth indicates the number of times the coding unit is spatially divided from the maximum coding unit. As the depth becomes deeper, the depth-specific coding unit is divided from the maximum coding unit to the minimum coding unit, the depth of the maximum coding unit is the highest depth, and the minimum coding unit is the lowest coding unit. Is defined as. Since the size of the depth-specific coding unit decreases as the depth of the maximum coding unit increases, the upper depth coding unit may include a plurality of lower depth coding units.

0021As described above, the video data of the current picture may be divided into the maximum coding units according to the maximum size of the coding units, and each maximum coding unit may include a coding unit divided by depth. Since the maximum coding unit according to the embodiment of the present invention is divided according to the depth, the video data of the spatial domain included in the maximum coding unit is hierarchically classified according to the depth.

0022The maximum depth and the maximum size of the coding unit may be preset to limit the total number of times the height and width of the maximum coding unit can be divided hierarchically.

0023The coding depth determination unit 120 encodes at least one divided region in which the region of the maximum coding unit is divided for each depth, and determines the depth at which the final coding result is output for each at least one divided region. That is, the coding depth determination unit 120 encodes the video data in the coding unit for each depth for each maximum coding unit of the current picture, selects the depth at which the smallest coding error occurs, and determines the coding depth. To do. Therefore, the coded video data of the coding unit corresponding to the determined coding depth is finally output. Also, the coding unit corresponding to the coding depth is regarded as the coded coding unit.

0024The determined coding depth and the video data encoded by the determined coding depth are output to the output unit 130.

0025The video data in the maximum coding unit is encoded based on the depth-specific coding unit by at least one depth below the maximum depth, and the coding results based on the respective depth-specific coding units are compared. As a result of comparing the coding errors of the coding units by depth, the depth with the smallest coding error is selected. At least one coding depth may be determined for each maximum coding unit.

0026As for the size of the maximum coding unit, as the depth becomes deeper, the coding unit is divided hierarchically and the number of coding units increases. Further, even if the coding units have the same depth included in one maximum coding unit, the coding error related to each data is measured, and the division into lower depths is determined. Therefore, even if the data is included in one maximum coding unit, the coding depth for each depth differs depending on the position, so that the coding depth differs depending on the position. Therefore, one or more coding depths are set for one maximum coding unit, and the data of the maximum coding unit is divided according to the coding units of one or more coding depths.

0027Therefore, the coding depth determination unit 120 according to one embodiment can determine the coding unit according to the tree structure currently included in the maximum coding unit. The "tree-structured coding unit" according to one embodiment includes the coding unit of the depth determined by the coding depth among all the coding units by depth currently included in the maximum coding unit. The coding unit of the coding depth is determined hierarchically by the depth in the same region within the maximum coding unit, and independently in the other regions. Similarly, the coding depth for the current region may be determined independently of the coding depth for other regions.

0028The maximum depth according to one embodiment is an index associated with the number of divisions from the maximum coding unit to the minimum coding unit. The first maximum depth according to one embodiment can indicate the total number of divisions from the maximum coding unit to the minimum coding unit. The second maximum depth according to one embodiment can indicate the total number of depth levels from the maximum coding unit to the minimum coding unit. For example, if the depth of the maximum coding unit is 0, the depth of the coding unit where the maximum coding unit is divided once is set to 1, and the depth of the coding unit divided twice is set to 1. It may be set to 2. In this case, if the coding unit divided four times from the maximum coding unit is the minimum coding unit, there are depth levels of 0,1,2,3 and 4, so the first maximum depth is 4. , The second maximum depth may be set to 5.

0029Predictive coding and conversion of the maximum coding unit is performed according to the maximum coding unit. Predictive coding and conversion by the maximum coding unit may also be performed on the basis of a deeper coding unit with a depth below the maximum depth. The conversion may be performed by an orthogonal conversion method or an integer conversion method.

0030The coding process, including predictive coding and transformation, occurs at any lower depth as the depth increases, because each time the maximum coding unit is divided by depth, the number of lower depth coding units increases. Performed for coding units. For convenience of explanation, predictive coding and conversion will be described in maximum coding units, based on the current depth coding units.

0031The coding device 100 can select data units of various sizes or forms for coding video data. In order to encode video data, operations such as predictive coding, conversion, and entropy coding are performed, and at this time, the same data unit is used for all operations, or different data units are used for each operation. May be used.

0032For example, the coding apparatus 100 selects not only a coding unit for coding video data but also a data unit different from the coding unit for predictive coding related to the video data in the coding unit. Can be done.

0033For predictive coding of the maximum coding unit, even if the predictive coding is performed on the coding unit of the coding depth, that is, the coding unit that is not divided into the coding units corresponding to the lower depths. Good. Hereinafter, the data unit that is the basis of the prediction without being further divided will be referred to as a prediction unit. The partition in which the prediction unit is divided may include a data unit in which at least one of the height and width of the prediction unit is divided.

0034For example, if the coding unit of size 2Nx2N (where N is a positive integer) is not further divided, it will be the predicted unit of size 2Nx2N, and the partition size may be 2Nx2N, 2NxN, Nx2N, NxN. The partition type according to one embodiment is not only a symmetric partition whose predicted unit height or width is divided by a symmetric ratio, but also a partition which is divided by an asymmetric ratio such as 1: n or n: 1. Partitions divided into geometric forms, partitions of arbitrary form, and the like may be selectively included.

0035The prediction mode of the prediction unit may be at least one of an intra mode, an inter mode, and a skip mode. For example, intra-mode and inter-mode are performed for partitions of 2Nx2N, 2NxN, Nx2N, NxN size. Also, skip mode is only performed for 2Nx2N size partitions. If there are a plurality of prediction units inside the coding unit, coding is performed independently for each prediction unit, and the prediction mode with the smallest coding error is selected.

0036Further, the video coding device 100 can convert video data based on a data unit having a size different from that of the coding unit.

0037For the conversion of the coding unit, the conversion is performed based on the data unit smaller than or the same size as the coding unit. For example, the data unit for conversion may include an intra-mode data unit and an inter-mode data unit.

0038Hereinafter, the processing unit that is the basis of conversion is referred to as a conversion unit. As for the conversion unit, the conversion depth indicating the number of divisions until the height and width of the coding unit are divided and the conversion unit is reached is set. For example, if the conversion unit of the current coding unit of size 2Nx2N is the conversion unit of size 2Nx2N that is the same size as the current coding unit, the conversion depth is set to 0 and the height and width of the current coding unit are respectively. If it is a conversion unit of size NxN that is halved and divided into 4 ^ 1 in all, the height and width of the current coding unit are each quadranted at the conversion depth of 1, and 4 ^ 2 in total. If it is a conversion unit of size N / 2xN / 2 divided into, the conversion depth may be set to 2. For example, the conversion unit of the upper conversion depth may be divided into four conversion units of the lower conversion depth according to the conversion hierarchical conversion depth, and a conversion unit having a hierarchical tree structure may be set.

0039Similar to the coding unit, the conversion unit within the coding unit is recursively divided into smaller size regions, and the conversion unit is independently determined for each region. Therefore, the residual data of the coding unit is divided by the conversion depth according to the conversion unit by the tree structure.

0040The coding information for each coding depth requires not only the coding depth but also prediction-related information and conversion-related information. Therefore, the coding unit depth determination unit 120 includes not only the coding depth that caused the minimum coding error, but also the partition type in which the coding prediction unit of the coding depth is divided into the prediction unit partitions, the prediction mode for each prediction unit, and the prediction mode. The size of the conversion unit for conversion can be determined.

0041A method for determining a coding unit and a partition according to a tree structure of the maximum coding unit according to an embodiment of the present invention will be described later with reference to FIGS. 3 to 12.

0042The coding depth determination unit 120 measures the coding error of the coding unit by depth by using the rate-distortion optimization based on the Lagrangeian multiplier.

0043The video data coding unit 130 provides video data of the maximum coding unit encoded based on at least one coding depth determined by the coding depth determination unit 120, and coding mode information based on the coding depth. Output to bit stream. ' The encoded video data is acquired by encoding the remaining data of the video.

0044The coding mode information based on the coding depth may include the coding depth, the information related to the partition type of the prediction unit, the prediction mode information, and the conversion unit size information.

0045The coding depth information may be defined by using the depth-specific division information indicating whether or not to code in the lower depth coding unit without coding at the current depth. If the current depth of the current coding unit is the coding depth, then the current coding unit is coded in the coding unit of the current depth, so that the current depth division information is not divided into lower depths. Is defined in. On the other hand, if the current depth of the current coding unit is not the coding depth, coding using the lower depth coding unit must be attempted, so that the current depth division information is the lower depth coding. It is also defined to be divided into units.

0046If the current depth is not the coding depth, coding is done for the coding units divided into lower depth coding units. Since there is at least one lower depth coding unit in one current depth coding unit, the coding is iteratively performed for each lower depth coding unit, and the same depth coding unit. The coding is performed recursively.

0047For one maximum coding unit, since the tree-structured coding units must be determined within one maximum coding unit, and information about at least one coding mode must be determined for each coding depth coding unit. Determines information about at least one coding mode. Further, the data of the maximum coding unit is divided hierarchically according to the depth, and the coding depth may be different for each position. Therefore, information about the coding depth and the coding mode is set for the data.

0048Therefore, the output unit 130 assigns the coding information related to the coding depth and the coding mode to at least one of the coding unit, the prediction unit, and the minimum unit included in the maximum coding unit. Can be done.

0049The minimum unit according to one embodiment is a square data unit having a size in which the minimum coding unit, which is the lowest coding depth, is divided into four. Alternatively, the smallest unit may be any coding unit included in the largest coding unit, a prediction unit, and a square data unit of maximum size contained within the conversion unit.

0050For example, the coding information output via the output unit 130 is classified into coding information for each coding unit and coding information for each prediction unit. The coding information for each coding unit may include prediction mode information and partition size information. The coding information for each prediction unit includes information about the estimation direction of the intermode, information about the reference video index of the intermode, information about the motion vector, information about the chroma component of the intramode, and information about the interpolation method of the intramode. Etc. may be included. In addition, information on the maximum size and maximum depth of coding units defined for each picture, slice, or GOP may be inserted in the header of the SPS (sequence parameter set) or bitstream.

0051In the video coding apparatus 100, the lower depth coding unit is a coding unit having a size obtained by halving the height and width of the upper depth coding unit. That is, if the size of the current depth coding unit is 2Nx2N, then the size of the lower depth coding unit is NxN. Therefore, the current coding unit of 2Nx2N size may include up to four lower depth coding units of NxN size.

0052Therefore, the video coding apparatus 100 encodes the optimum form and size for each maximum coding unit based on the size and maximum depth of the maximum coding unit currently determined in consideration of the characteristics of the picture. The unit can be determined and a tree-structured coding unit can be constructed. In addition, since each maximum coding unit can be coded by various prediction modes and conversion methods, the optimum coding mode is determined in consideration of the video characteristics of the coding units of various video sizes. To.

0053If a video having a very high resolution or a large amount of data is encoded in the conventional macroblock unit, the number of macroblocks per picture becomes excessively large. Therefore, since the amount of compressed information generated for each macroblock also increases, the transmission burden of the compressed information increases, and the data compression efficiency decreases. However, the video coding apparatus 100 according to the embodiment of the present invention can adjust the coding unit in consideration of the video characteristics while increasing the maximum size of the coding unit in consideration of the size of the video. Therefore, the video compression efficiency is increased.

0054FIG. 2 illustrates a block diagram of a video decoding apparatus according to an embodiment of the present invention. Referring to FIG. 2, the video decoding apparatus 200 according to the embodiment of the present invention includes a receiving unit 210, a video data and coded information extraction unit 220, and a video data decoding unit 230. For definitions of various terms such as coding units, depths, prediction units, conversion units, and information about various coding modes for various processing of the video decoding apparatus 200 according to one embodiment, refer to FIG. 1 and the video coding apparatus 100. As described above with reference.

0055The receiver 210 receives the bitstream and parses it. The video data and coding information extraction unit 220 extracts the coded video data related to the coding unit of the tree structure for each maximum coding unit, acquires the extracted video data, and obtains the extracted video data to obtain the video data decoding unit. Output to 230. The video data and coded information extraction unit 220 can extract information related to the maximum size of the coded unit of the received current picture from the header or SPS related to the current picture.

0056Further, the video data and coding information extraction unit 220 parsing the bit stream and extracts information about the coding depth and the coding mode related to the coding unit by the tree structure for each maximum coding unit. Information about the extracted coding depth and coding mode is output to the video data decoding unit 230. In other words, the video data of the bitstream is divided into the maximum coding units, and the video data decoding unit 230 decodes the video data of each maximum coding unit.

0057Information about the maximum coding depth by coding unit and coding mode is set for one or more coding depth information, and information about the coding mode by coding depth is the partition type information of the coding unit. Prediction mode information, size information of conversion units, and the like may be included. Further, as the coding depth information, the division information for each depth is also extracted.

0058The information about the maximum coding depth and coding mode extracted by the coding information extraction unit 220 is coded by the maximum coding unit depth at the coding end together with the video coding apparatus 100 according to one embodiment. Coding is performed repeatedly for each unit to generate a minimum coding error, and it is information about a determined coding depth and coding mode. Therefore, the video decoding apparatus 200 can decode the data by a coding method that causes a minimum coding error and restore the video.

0059Since the coding information about the coding depth and the coding mode is assigned to a predetermined data unit among the coding unit, the prediction unit, and the minimum unit, the coding information extraction unit 220 encodes each predetermined data unit. Information about the depth of conversion and the coding mode can be extracted. If information about the coding depth and coding mode of the maximum coding unit is recorded for each predetermined data unit, the predetermined data unit having the same information about the coding depth and coding mode is It is inferred to be a data unit included in the same maximum coding unit.

0060The video data decoding unit 230 decodes the video data of each maximum coding unit based on the maximum coding depth and coding mode information extracted by the coding information extraction unit to obtain a current picture. Restore. That is, the video data decoding unit 230 is a video encoded based on the partition type, the prediction mode, and the conversion unit for each coding unit among the coding units having a tree structure included in the maximum coding unit. The data can be decrypted. The decoding process may include a motion prediction process including intra prediction and motion compensation, and an inverse transformation process. The inverse transformation process is carried out by the method of inverse orthogonal transformation or inverse integer transformation. The video data decoding unit 230 performs intra-prediction or motion compensation for each coding unit according to each partition and prediction mode based on the partition type information and prediction mode information of the prediction unit of the coding unit according to the coding depth. It can be carried out.

0061Further, the video data decoding unit 230 reverses each coding unit according to each conversion unit based on the size information of the conversion unit of the coding unit according to the coding depth for the reverse conversion by the maximum coding unit. The conversion can be done.

0062The video data decoding unit 230 can determine at least one coding depth of the current maximum coding unit that uses the depth-based division information. If the split information indicates that it will not be split further at the current depth, then the current depth is the coded depth. Therefore, the video data decoding unit 230 can decode the coding unit of the current depth for the video data of the current maximum coding unit by using the division type of the prediction unit, the prediction mode, and the conversion unit size information. ..

0063That is, among the coding unit, the prediction unit, and the minimum unit, the coding information set for the predetermined data unit is observed, and the data units having the coding information including the same division information are collected and the video data is collected. It is regarded by the decoding unit 230 as one data unit to be decoded in the same coding mode.

0064In the coding process, the video decoding apparatus 200 recursively encodes each maximum coding unit, acquires information related to the coding unit that caused the minimum coding error, and currently decodes the picture. It may be used for. That is, for each maximum coding unit, the coded video data of the coding unit according to the tree structure determined by the optimum coding unit is decoded. The maximum size of the coding unit is determined in consideration of the resolution and the amount of video data.

0065Therefore, even for high-resolution video or video with an excessively large amount of data, the coding unit determined adaptively to the characteristics of the video by using the information about the optimum coding mode transmitted from the coding end. Video data can be efficiently decoded and restored depending on the size and coding mode.

0066A method of determining a coding unit, a prediction unit, and a conversion unit of a tree structure according to an embodiment of the present invention will be described with reference to FIGS. 3 to 13.

0067FIG. 3 illustrates a hierarchical coding unit according to an embodiment of the present invention. With reference to FIG. 3, in the example of the coding unit, the size of the coding unit is expressed in width x height and may include 32x32, 16x16, 8x8 from the coding unit having a size of 64x64. The size 64x64 coding unit is divided into size 64x64,64x32, 32x64, 32x32 partitions, the size 32x32 coding unit is divided into size 32x32, 32x16, 16x32, 16x16 partitions, and the size 16x16 coding unit is It is divided into partitions of size 16x16,16x8,8x16,8x8, and the coding unit of size 8x8 is divided into partitions of size 8x8,8x4,4x8,4x4.

0068For video data 310, the resolution is set to 1920x1080, the maximum coding unit size is set to 64, and the maximum depth is set to 2. For video data 320, the resolution is set to 1920x1080, the maximum size of the coding unit is set to 64, and the maximum depth is set to 3. For the video data 330, the resolution is set to 352x288, the maximum size of the coding unit is set to 16, and the maximum depth is set to 1. The maximum depth illustrated in FIG. 3 indicates the total number of divisions from the maximum coding unit to the minimum coding unit.

0069When the resolution is high or the amount of data is large, it is desirable that the maximum size of the coding size is relatively large in order not only to improve the compression rate but also to accurately reflect the video characteristics. Therefore, the maximum code size of the video data 310,320, which has a higher resolution than the video data 330, is selected to be 64.

0070Since the maximum depth of the video data 310 is 2, the coding unit 315 of the video data 310 is divided twice from the maximum coding unit having a major axis size of 64 to increase the depth by two layers, and the major axis size. May include up to coding units where is 32,16. On the other hand, since the maximum depth of the video data 330 is 1, the coding unit 335 of the video data 330 is divided once from the coding unit having a major axis size of 16 to increase the depth by one layer, and the major axis. It may include up to a coding unit of size 8.

0071Since the maximum depth of the video data 320 is 3, the coding unit 325 of the video data 320 is divided three times from the maximum coding unit having a major axis size of 64 to increase the depth by three layers, and the major axis size. May include up to coding units where is 32,16,8. The deeper the depth, the smaller the coding unit is used to encode the image, which is suitable for encoding an image containing a more precise scene.

0072FIG. 4 shows a block diagram of a video coding unit based on a coding unit according to an embodiment of the present invention.

0073The video coding unit 400 according to the embodiment is performed by the coding depth determination unit 120 of the video coding device 100 described above, including the work performed for coding the video data. In other words, the intra prediction unit 410 makes an intra prediction for the prediction unit of the intra mode in the current frame 405, and the motion estimation unit 420 and the motion compensation unit 425 make the current frame 405 and the reference frame 495 for the prediction unit of the inter mode. Use to perform inter-prediction and motion compensation.

0074Residual values were generated based on the prediction units output from the intra prediction unit 410, motion estimation unit 420, and motion compensation unit 425, and the generated residual values were quantized via the conversion unit 430 and the quantization unit 440. It is output as a conversion coefficient. The quantized conversion coefficient is further restored as a residual value via the inverse quantization unit 460 and the inverse conversion unit 470, and the restored residual value is post-processed via the deblocking unit 480 and the loop filtering unit 490. Is output to reference frame 495. The quantized conversion coefficient is output as a bit stream 455 via the entropy coding unit 450.

0075In order to apply to the video coding device 100 according to the embodiment of the present invention, the intra prediction unit 410, the motion estimation unit 420, the motion compensation unit 425, the conversion unit 430, and the quantization unit, which are the components of the video coding unit 400. The 440, the entropy coding unit 450, the inverse quantization unit 460, the inverse conversion unit 470, the deblocking unit 480, and the loop filtering unit 490 are all coded by a tree structure in consideration of the maximum depth for each maximum coding unit. The video coding process is processed based on each coding unit of the coding units.

0076In particular, the intra prediction unit 410, the motion estimation unit 420, and the motion compensation unit 425 consider the maximum size and maximum depth of the current maximum coding unit, and among the coding units by the tree structure, the partition of each coding unit. And the prediction mode is determined, and the conversion unit 430 must determine the size of the conversion unit in consideration of the size of the conversion unit in each coding unit among the coding units in the tree structure.

0077FIG. 5 shows a block diagram of the video decoding unit 500 based on the coding unit according to the embodiment of the present invention.

0078The parsing unit 510 parses the video data encoded from the bitstream 505 and the coding information required for decoding. The encoded video data is output as dequantized data via the entropy decoding unit 520 and the dequantization unit 530, and the dequantized data is output through the inverse conversion unit 540 in the spatial region. It is restored as video data.

0079The intra prediction unit 550 performs intra prediction related to the coding unit in the intra mode for the video data in the spatial region, and the motion compensation unit 560 uses the reference picture 585 to compensate the motion related to the coding unit in the inter mode. I do.

0080The video data in the spatial region that has passed through the intra prediction unit 550 and the motion compensation unit 560 is post-processed and output via the deblocking unit 570 and the loop filtering unit 580. Further, the video data post-processed through the deblocking unit 570 and the loop filtering unit 580 is output as the reference frame 585.

0081In order to apply to the video decoding device 200 according to the embodiment of the present invention, the parsing unit 510, the entropy decoding unit 520, the inverse quantization unit 530, the inverse conversion unit 540, and the intra, which are the components of the video decoding device 500, are used. The prediction unit 550, the motion compensation unit 560, the deblocking unit 570, and the loop filtering unit 580 operate for each maximum coding unit based on the coding unit by the tree structure.

0082In particular, the intra prediction unit 550 and the motion compensation unit 560 operate based on the partition and the prediction mode for each coding unit according to the tree structure, and the inverse conversion unit 540 operates for each coding unit and the size of the conversion unit. Operates based on.

0083FIG. 6 illustrates a depth-coded coding unit and a predicted partition according to an embodiment of the present invention.

0084The video coding device 100 according to one embodiment and the video decoding device 200 according to one embodiment use hierarchical coding units in order to consider video characteristics. The maximum height, width, and maximum depth of the coding unit are variously set according to the user's request, which is adaptively determined by the characteristics of the image. The size of the coding unit for each depth is also determined by the maximum size of the already set coding unit.

0085The hierarchical structure 600 of the coding unit according to one embodiment of the present invention illustrates the case where the maximum height and width of the coding unit is 64 and the maximum depth is 4. Since the depth increases along the vertical axis of the hierarchical structure 600 of the coding unit according to the embodiment, the height and width of the coding unit for each depth are divided. Further, along the horizontal axis of the hierarchical structure 600 of the coding units, the prediction units and partitions that serve as the prediction base of the coding units for each depth are shown.

0086The coding unit 610 has a depth of 0 as the maximum coding unit in the hierarchical structure 600 of the coding unit, and the size of the coding unit, that is, the height and width is 64x64. The depth increases along the vertical axis, with a depth 1 coding unit 620 of size 32x32, a depth 2 coding unit 630 of size 16x16, a size 8x8 depth 3 coding unit 640, and a size 4x4. There is a coding unit 650 with a depth of 4. The coding unit 650 at depth 4 of size 4x4 is the smallest coding unit.

0087Prediction units and partitions of coding units are arranged along the horizontal axis for each depth. That is, if the maximum coding unit 610 of size 64x64 at depth 0 is the prediction unit, the prediction unit is the partition 610 of size 64x64, the partition 612 of size 64x32, and the size 32x64 contained in the coding unit 610 of size 64x64. It is divided into partition 614, partition 616 with a size of 32x32.

0088Similarly, the predictive units of size 32x32 coding unit 620 at depth 1 are size 32x32 partition 620, size 32x16 partition 622, size 16x32 partition 624, size 16x16 partition contained in size 32x32 coding unit 620. Divided into 626.

0089The predictive unit of the size 16x16 coding unit 630 at depth 2 is divided into the size 16x16 partition 630, the size 16x8 partition 632, the size 8x16 partition 634, and the size 8x8 partition 636 contained in the size 16x16 coding unit 630. Will be done.

0090The predictive unit of the size 8x8 coding unit 640 at depth 3 is divided into the size 8x8 partition 640, the size 8x4 partition 642, the size 4x8 partition 644, and the size 4x4 partition 646 contained in the size 8x8 coding unit 640. Will be done.

0091The size 4x4 coding unit 650 at depth 4 is the smallest coding unit and the lowest depth coding unit. The predictive unit of coding unit 650 is set only as a partition of size 4x4.

0092In order to determine the coding depth of the maximum coding unit 610, the coding depth determination unit 120 of the video coding apparatus according to one embodiment codes for each coding unit of each depth included in the maximum coding unit 610. Must be coded.

0093As for the number of depth-specific coding units for containing data of the same range and size, the number of depth-specific coding units increases as the depth increases. For example, for data contained in one coding unit at depth 1, four coding units at depth 2 are required. Therefore, in order to compare the coding results of the same data by depth, one coding unit of depth 1 and four coding units of depth 2 must be used for each coding.

0094In order to perform coding according to the current depth of the depth, coding is performed for each predicted unit of the current depth coding unit along the horizontal axis of the hierarchical structure 600 of the coding unit, and the coding with the smallest current depth is performed. The error is also selected. Further, as the depth becomes deeper along the vertical axis of the hierarchical structure 600 of the coding unit, coding is performed for each depth, and the minimum coding error for each depth is compared to search for the minimum coding error. .. In the maximum coding unit 610, the depth and partition where the minimum coding error occurs are selected as the coding depth and partition type in the maximum coding unit 610.

0095FIG. 7 illustrates the relationship between the coding unit 710 and the conversion unit 720 according to the embodiment of the present invention.

0096The video coding device 100 and the video decoding device 200 divide the video into coding units smaller than or the same size as the maximum coding unit for each maximum coding unit, and encode or decode the video. In the coding process, the size of the conversion units for conversion is selected based on data units that are not greater than each coding unit.

0097For example, when the size of the coding unit 710 is 64x64 size, after converting and coding in 32x32, 16x16, 8x8, 4x4 size conversion units, the conversion unit with the smallest error from the original is selected. ..

0098FIG. 8 illustrates the coded information by depth according to one embodiment of the present invention.

0099The coding information output unit 130 of the video coding apparatus 100 according to the embodiment of the present invention is information about the coding mode, and information 800 about the partition type and prediction for each coding unit of the coding depth. Information 810 about the mode and information 820 related to the conversion unit size can be encoded and transmitted.

0100The information 800 related to the partition type is a data unit for predicting the movement of the current coding unit, and the prediction unit of the current coding unit indicates information related to the form of the divided partition. For example, the current coding unit CU_0 of size 2Nx2N is divided into one of the following types: partition 802 of size 2Nx2N, partition 804 of size 2NxN, partition 806 of size Nx2N, and partition 808 of size NxN. .. In this case, information 800 about the partition type of the current coding unit is set to indicate one of size 2Nx2N partition 802, size 2NxN partition 804, size Nx2N partition 806 and size NxN partition 808. ..

0101Information 810 about the prediction mode indicates the movement prediction mode of each partition. For example, via the information 810 about the prediction mode, the partition pointed to by the information 800 about the division type is one of the intra mode 812, the inter mode 814, and the skip mode 816, and it is set whether or not the motion prediction is performed. Will be done.

0102Further, the information 820 regarding the conversion unit size indicates which conversion unit is currently used for conversion of the coding unit. For example, the conversion unit may be one of the first intra conversion unit size 822, the second intra conversion unit size 824, the first inter conversion unit size 826, and the second intra conversion unit size 828.

0103The coding information extraction unit 210 of the video decoding apparatus 200 according to the embodiment of the present invention relates to information 800 about the partition type, information 810 about the prediction mode, and conversion unit size for each coding unit for each depth. Information 820 can be extracted and used for decryption.

0104FIG. 9 illustrates a depth-based coding unit according to an embodiment of the present invention.

0105Divided information may be used to indicate changes in depth. The division information indicates whether or not the coding unit of the current depth is divided into the coding unit of the lower depth.

0106Predictive unit 910 for predictive coding of depth 0 and 2N_0x2N_0 size coding unit 900 is 2N_0x2N_0 size partition type 912, 2N_0xN_0 size partition type 914, N_0x2N_0 size partition type 916, N_0xN_0 size partition type 918. It may be included. Only partitions 912,914,916,918 whose predictive units are divided by a symmetric ratio are illustrated, but as mentioned above, the partition type is not limited to this, but asymmetric partitions, partitions of arbitrary form, geometry. It may include a partition in a symmetrical form.

0107For each partition type, predictive encoding must be repeated for each 2N_0x2N_0 size partition, two 2N_0xN_0 size partitions, two N_0x2N_0 size partitions, and four N_0xN_0 size partitions. For partitions of size 2N_0xN_0, size N_0xN_0, size N_0x2N_0 and size N_0xN_0, predictive coding is performed in intra-mode and inter-inter-mode. Skip mode is performed on partitions of size 2N_0x2N_0.

0108The coding errors of sizes 2N_0x2N_0, 2N_0xN_0, N_0x2N_0 and N_0xN_0 according to the partition types 912,914,916,918 are compared to determine the smallest coding error of the partition types. If the coding error due to one of the partition types 912,914,916 is the smallest, the prediction unit 910 is not divided into lower depths any further.

0109If the coding error due to partition type 918 of size N_0xN_0 is the smallest, change depth 0 to 1 and divide (920), iteratively coded for partition type coding unit 930 of depth 2 and size N_0xN_0. It is possible to perform the conversion and search for the minimum coding error.

0110Predictive unit 940 for predictive coding of depth 1 and size 2N_1x2N_1 (= N_0xN_0) coding unit 930 is partition type 942 of size 2N_1x2N_1, partition type 944 of size 2N_1xN_1, partition type 946 of size N_1x2N_1, size N_1xN_1 It may include partition type 948.

0111Also, if the coding error due to partition type 948 of size N_1xN_1 is the smallest, iterate over the coding unit 960 of depth 2 and size N_2xN_2 while changing depth 1 to depth 2 and dividing (950). It is possible to perform coding and search for the minimum coding error.

0112When the maximum depth is d, the division operation by depth is performed until the depth d-1, and the division information can be encoded for the depth from 0 to d-2. That is, in operation (970), when the coding unit corresponding to the depth d-2 is divided and the coding is performed up to the depth d-1, the depth d-1 and the size 2N_ (d-1) x2N_ (d-) Prediction unit 990 for coding unit 980 in 1) is partition type 992 of size 2N_ (d-1) x2N_ (d-1), size 2N_ (d-1) xN_ (d-1) Partition type 994, partition type 996 of size N_ (d-1) x2N_ (d-1), partition type 998 of size N_ (d-1) xN_ (d-1) may be included.

0113In the partition type, one partition of size 2N_ (d-1) x2N_ (d-1), two partitions of size 2N_ (d-1) xN_ (d-1), two sizes N_ (d-) 1) The minimum code must be coded repeatedly through predictive coding for each partition of x2N_ (d-1) and four partitions of size N_ (d-1) xN_ (d-1). The partition type that causes the conversion error is searched.

0114Even if the partition type 998 of size N_ (d-1) xN_ (d-1) has the smallest coding error, the maximum depth is d, so the coding unit CU_ (d-1) at depth d-1 is , The coding depth for the maximum coding unit 900 is now determined to be depth d-1 without going through the process of dividing into lower depths, and the partition type is N_ (d-1) xN_ (d-1). It will also be decided. Further, since the minimum coding unit 980 having the maximum depth d and the lowest depth d-1 is not divided into lower depths, no division information is set for the coding unit 980.

0115The data unit 999 may be the smallest unit currently associated with the largest coding unit. The minimum unit according to one embodiment may be a square data unit obtained by dividing the minimum coding unit 980 into four parts. Through such an iterative coding process, the video coding apparatus 100 compares the coding errors by depth of the coding unit 900, selects the depth at which the smallest coding error occurs, and determines the coding depth. It can be determined and the partition type and prediction mode can be set to the coding depth coding mode.

0116In such a way, the minimum coding error by depth of any depth 0,1, ..., d-1, d may be compared, the depth with the smallest error may be selected and determined as the coding depth. .. The coding depth and the partition type and prediction mode of the prediction unit may be encoded and transmitted as information about the coding mode. Also, since the coding unit must be divided from the depth 0 to the coding depth, only the coding depth division information is set to 0, and the depth-specific division information excluding the coding depth is set to 1. To.

0117The coding information extraction unit 220 of the video decoding apparatus 200 can extract information about the coding depth and the prediction unit related to the coding unit 900 and can use it to decode the coding unit 900. The video decoding apparatus 200 can use the depth-specific division information to determine the depth at which the division information is 0 as the coding depth, and use the information about the coding mode related to the depth for decoding. ..

011810 to 12 illustrate the relationship between the coding unit, the prediction unit, and the conversion unit according to the embodiment of the present invention.

0119The coding unit 1010 is a coding unit for each coding depth determined by the video coding apparatus 100 according to one embodiment with respect to the maximum coding unit 1000. The prediction unit 1060 is a partition of the prediction unit of each coding depth-based coding unit in the coding unit 1010, and the conversion unit 1070 is a conversion unit of each coding depth-based coding unit.

0120Assuming that the depth of the maximum coding unit 1000 is 0 in the coding unit 1010 by depth, the coding unit 1012,1054 has a depth of 1, and the coding unit 1014,1016,1018,1028,1050,1052. Has a depth of 2, a coding unit of 1020,1022,1024,1026,1030,1032,1038 has a depth of 3, and a coding unit of 1040,1042,1044,1046 has a depth of 4.

0121In the prediction unit 1060, some partitions 1014,1016,1022,1032,1048,1050,1052,1054 are in the form in which the coding unit is divided. That is, partition 1014,1022,1050,1054 is a 2NxN partition type, partition 1016,1048,1052 is an Nx2N partition type, and partition 1032 is an NxN partition type. The predicted units and partitions of the depth-coded coding unit 1010 are smaller than or the same as the respective coding units.

0122In the conversion unit 1070, the video data of the coding unit 1052 is converted or inversely transformed in a data unit having a size smaller than that of the coding unit 1052. Further, the conversion units 1014,1016,1022,1032,1048,1050,1052,1054 are data units of different sizes or forms as compared with the prediction units and partitions in the prediction unit 1060. That is, the video coding device 100 and the video decoding device 200 can perform prediction and conversion / inverse conversion operations related to the same coding unit based on separate data units.

0123By recursively coding each unit to determine the optimum coding unit, a coding unit having a recursive tree structure is also constructed. The coding information may include division information related to the coding unit, partition type information, prediction mode information, and conversion unit size information. Table 1 below shows an example that can be set by the video coding device 100 and the video decoding device 200.

0124<tables num="1"><img id="000002" he="84" wi="159" file="JP5770185B2_D0001.tif" img-format="tif" img-content="drawing" /></tables> The output unit 130 of the video coding apparatus 100 outputs the coding information related to the coding unit by the tree structure, and the coding information extraction unit 220 of the video decoding apparatus 200 according to one embodiment outputs the coding information from the received bit stream. , It is possible to extract the coding information related to the coding unit by the tree structure.

0125The division information indicates whether or not the current coding unit is divided into lower depth coding units. If the division information of the current depth d is 0, the current coding unit is the coding depth that is not further divided into the lower coding units. Therefore, with respect to the coding depth, the partition type information and the prediction mode , Conversion unit size information may be defined. If the division information requires further division by one step, each of the four divided lower depth coding units must be coded independently.

0126The prediction mode can be indicated by one of an intra mode, an inter mode, and a skip mode. Intra mode and intermode are defined for all partition types, and skip mode is defined only for partition type 2Nx2N.

0127The partition type information includes symmetric partition types 2Nx2N, 2NxN, Nx2N and NxN whose predicted unit height or width is divided by symmetric ratio, and asymmetric partition type 2NxnU, 2NxnD, nLx2N divided by symmetric ratio. , NRx2N and so on. The asymmetric partition types 2NxnU and 2NxnD are divided into heights of 1: 3 and 3: 1, respectively, and the asymmetric partition types nLx2N and nRx2N are divided into widths of 1: 3 and 3: 1, respectively. Shows the form.

0128The conversion unit size may be set to two sizes in the intra mode and two sizes in the inter mode. That is, if the conversion unit division information is 0, the size of the conversion unit is currently set to the size of the coding unit 2Nx2N. If the conversion unit division information is 1, the conversion unit of the size in which the coding unit is currently divided is set. Also, if the partition type related to the current coding unit of size 2Nx2N is a symmetric partition type, the conversion unit size is set to NxN, and if it is an asymmetric partition type, it is set to N / 2xN / 2. May be good.

0129The coding information such as the coding unit according to the tree structure according to the embodiment is assigned to at least one of the coding unit, the prediction unit, and the minimum unit of the coding depth. The coding unit of the coding depth may include one or more prediction units and the minimum unit having the same coding information.

0130Therefore, if the coding information held by each of the adjacent data units is confirmed, it is confirmed whether or not the coding information is included in the coding units having the same coding depth. Further, if the coding information possessed by the data unit is used, the coding unit of the coding depth can be confirmed, so that the distribution of the coding depth and the like within the maximum coding unit can be inferred.

0131Therefore, in this case, when the current coding unit makes a prediction by referring to the peripheral data unit, the coding information of the data unit in the depth-based coding unit adjacent to the current coding unit is directly referred to and used.

0132In another embodiment, when the current coding unit refers to the peripheral coding unit and predictive coding is performed, the coding information of the adjacent depth-specific coding units is used, and within the depth-specific coding unit, Peripheral coding units are also referenced by searching for data that is currently adjacent to the coding unit.

0133FIG. 13 illustrates the relationship between the coding unit, the prediction unit, and the conversion unit based on the coding mode information in Table 1.

0134The maximum coding unit 1300 includes a coding depth coding unit 1302,1304,1306,1312,1314,1316,1318. Since one of the coding units 1318 is a coding unit of the coding depth, the division information is set to 0. The partition type information of the encoding unit 1318 of size 2Nx2N may be set to one of the partition types 2Nx2N 1322, 2NxN 1324, Nx2N 1326, NxN 1328, 2NxnU 1332, 2NxnD 1334, nLx2N 1336 and nRx2N 1338.

0135If the partition type information is set to one of the symmetrical partition types 2Nx2N 1322, 2NxN 1324, Nx2N 1326 and NxN 1328, and the conversion unit division information (TU size flag) is 0, then the size is 2Nx2N. If the conversion unit 1342 is set and the conversion unit division information is 1, the conversion unit 1344 of size NxN may be set.

0136If the partition type information is set to one of the asymmetric partition types 2NxnU 1332, 2NxnD 1334, nLx2N 1336 and nRx2N 1338, the conversion of size 2Nx2N if the conversion unit division information (TU size flag) is 0. If the unit 1352 is set and the conversion unit division information is 1, a conversion unit 1354 of size N / 2xN / 2 may be set.

0137With reference to FIG. 13, the conversion unit division information (TU size flag) has 0 or 1, but the conversion unit division information (TU size flag) is not limited to one bit, and the conversion unit division information (TU size flag) is not limited to one bit. As (TU size flag) increases from 0, the conversion unit can be hierarchically divided to have a tree structure.

0138In this case, the size of the conversion unit actually used may be expressed by using the conversion unit division information (TU size flag) of the conversion unit together with the maximum size and the minimum size of the conversion unit. According to one embodiment, the coding apparatus 100 can encode the maximum conversion unit size information, the minimum conversion unit size information, and the maximum conversion unit division information (TU size flag). The result of encoding the maximum conversion unit size information, the minimum conversion unit size information, and the maximum conversion unit division information (TU size flag) may be inserted into the SPS. According to one embodiment, the decoding device 200 can decode the video by using the maximum conversion unit size information, the minimum conversion unit size information, and the maximum conversion unit division information (TU size flag).

0139Hereinafter, the intra prediction performed by the intra prediction unit 410 of the video coding device 100 and the intra prediction unit 550 of the video decoding device 200 of FIG. 5 according to the embodiment of the present invention of FIG. 4 will be specifically described. In the following description, the coding unit is a term that refers to the block that is currently encoded in the video coding stage, and the decoding unit is the term that refers to the block that is currently decoded in the video decoding stage. Is. The terms coding unit and decoding unit differ only in which stage of the video coding stage or decoding stage, and the coding unit in the coding stage is the decoding stage. It is called the decoding unit in. For the sake of unity of terms, except in special cases, the coding unit and the decoding unit are collectively referred to as the coding unit at the coding stage and the decoding stage. Further, the intra-prediction method according to the embodiment of the present invention and its apparatus can also be applied to intra-prediction with a general video codec, if a person skilled in the art to which the present invention belongs, It will be understood through the specification.

0140FIG. 14 illustrates the number of intra prediction modes according to the size of the coding unit according to one embodiment of the present invention.

0141According to one embodiment of the present invention, the number of intra prediction modes applied to the coding unit can be variously set depending on the size of the coding unit (decoding unit in the decoding stage). As an example, referring to FIG. 14, assuming that the size of the intra-predicted coding unit is NxN, the actual line for each of the 2x2,4x4,8x8,16x16,32x32,64x64,128x128 size coding units. The number of intra prediction modes used is 5,9,9,17,33,5,5, respectively (Example). In case of 2) is set. As described above, the reason why the number of intra prediction modes actually performed is differentiated by the size of the coding unit is that the overhead for coding the prediction mode information differs depending on the size of the coding unit. In other words, in the case of a small-sized coding unit, the overhead for transmitting additional information such as the prediction mode of such a small coding unit can be increased even though the portion occupied by the entire video is small. Therefore, when a small coding unit is coded in too many prediction modes, the amount of bits may increase and the compression efficiency may decrease. Further, a coding unit having a large size, for example, a coding unit having a size of 64x64 or more is generally selected as a coding unit related to a flat region of an image. Coding large size coding units, which are often chosen to encode flat regions, in too many predictive modes is also inefficient in terms of compression efficiency.

0142Therefore, in FIG. 14, at least three coding units are N1xN1 (2N18, N1 is an integer), N2xN2 (16N232, N2 is an integer), and N3xN3 (64N3, N3 is an integer). The number of intra prediction modes performed for each coding unit having N1xN1 size is A1 (A1 is a positive integer), and the number of intra prediction modes performed for each coding unit having N2xN2 size. Is A2 (A2 is a positive integer), and the number of intra prediction modes performed for each coding unit having N3xN3 size is A3 (A3 is a positive integer), the relationship of A3 A1 A2. It is desirable to set the number of intra-prediction modes performed according to the size of each coding unit so as to satisfy. That is, when the current picture is roughly divided into a small size coding unit, an intermediate size coding unit, and a large size coding unit, the medium size coding unit has the largest number of prediction modes. It is desirable that the small size coding unit and the large size coding unit be set to have a relatively smaller number of prediction modes. However, the present invention is not limited to this, and small size and large size coding units can be set to have a larger number of prediction modes. The number of prediction modes according to the size of each coding unit shown in FIG. 14 is only one embodiment, and the number of prediction modes according to the size of each coding unit can be changed.

0143FIG. 15A is a diagram for explaining an example of an intra prediction mode applied to a coding unit of a predetermined size according to an embodiment of the present invention.

0144With reference to FIGS. 14 and 15A, as an example, vertical mode (mode 0), horizontal mode (mode 1), DC (direct) during intra-prediction of a coding unit having a 4 × 4 size. current mode (mode 2), diagonal down-left mode (mode 3), diagonal down-right mode (mode 4), vertical right mode (mode 5), horizontal It can have a horizontal-down mode (mode 6), a vertical-left mode (mode 7) and a horizontal-up mode (mode 8).

0145FIG. 15B is a diagram showing the direction of the intra prediction mode of FIG. 15A. In FIG. 15B, the number at the tip of the arrow indicates the mode value when making a prediction in that direction. Here, mode 2 is a non-directional DC prediction mode and is not shown.

0146FIG. 15C is a diagram illustrating an intra prediction method related to the coding unit shown in FIG. 15A. With reference to FIG. 15C, a predictive coding unit is generated using AM, which is the peripheral pixel of the current coding unit, by the available intra-prediction mode determined by the size of the coding unit. For example, the operation of predictively coding a 4 × 4 size current coding unit by mode 0 of FIG. 15A, that is, a vertical mode will be described. First, the pixel values of pixels A to D adjacent to the upper side of the current coding unit of 4 × 4 size are predicted to be the pixel values of the 4 × 4 current coding unit. That is, the value of pixel A is included in the first column of the 4x4 current coding unit, and the value of pixel B is included in the second column of the 4x4 current coding unit. For each pixel value, the value of pixel C is the value of 4 x 4 currently coded in the 3rd column of the 4 pixel values, and the value of pixel D is the value of 4 x 4 in the 4th column of the current coding unit. Predict each of the four included pixel values. Next, after obtaining the error value between the 4 × 4 current coding unit predicted using the pixels A to D and the actual value of the pixel included in the original 4 × 4 current coding unit, the error value is obtained. The error value is encoded.

0147FIG. 16 is a diagram for explaining another example of the intra prediction mode applied to the coding unit of a predetermined size according to the embodiment of the present invention.

0148With reference to FIGS. 14 and 16, as an example, vertical mode, horizontal mode, direct current mode, plane, when predicting the intra of a coding unit having a size of 2 × 2. There can be 5 modes in all modes and diagonal down-right modes.

0149On the other hand, as shown in FIG. 14, when a coding unit having a 32x32 size has 33 intra prediction modes, it is necessary to set the direction of 33 intra prediction modes. In one embodiment of the present invention, in addition to the intra-prediction modes as shown in FIGS. 15A to 15C and 16, the pixels in the coding unit are centered in order to set the intra-prediction modes in various directions. In, the prediction direction for selecting the peripheral pixel to be used as the reference pixel is set by using the dx and dy parameters. As an example, when each of the 33 prediction modes is defined as mode N (N is an integer from 0 to 32), mode 0 is the vertical mode, mode 1 is the horizontal mode, and mode 2 is the DC mode. mode 3 is set to plain mode, mode 4 ~ mode 31 Each is (1, -1), (1,1), (1,2), (2,1), (1,-2), (2,, as shown in Table 2 below. 1), (1,-2), (2, -1), (2, -11), (5, -7), (10, -7), (11,3), (4,3), (1,11), (1, -1), (12, -3), (1, -11), (1, -7), (3, -10), (5, -6), (7 , -6), (7, -4), (11,1), (6,1), (8,3), (5,3), (5,7), (2,7), (5 Using (dx, dy) expressed by one of the values of, -7) and (4, -3), tan^-1It can be defined as a prediction mode with a direction of (dy / dx).

0150<tables num="2"><img id="000003" he="119" wi="128" file="JP5770185B2_D0001.tif" img-format="tif" img-content="drawing" /></tables> The final mode 32 is set to a bilinear mode that utilizes bilinear interpolation, as will be described later with reference to FIG.

015117A to 17C are reference diagrams for explaining an intra-prediction mode having various directions according to an embodiment of the present invention.

0152As described with reference to Table 2, the intra-prediction mode according to one embodiment of the present invention utilizes a plurality of (dx, dy) parameters and is tan.^-1It can have various directions of (dy / dx).

0153With reference to Figure 17A, tan is determined by the mode-specific (dx, dy) values shown in Table 2, centered on the current pixel P to be predicted inside the current coding unit.^-1Peripheral pixels A and B located on the extension line 150 with an angle of (dy / dx) can now be used as predictors of pixel P. At this time, it is desirable that the peripheral pixels used as the predictor are the pixels of the previous coding unit on the upper side, the left side, the upper right side, and the lower left side of the current coding unit, which have been previously encoded and restored. Also, when the extension line 150 passes between the peripheral pixels at the integer position that are not the peripheral pixels at the integer position, the peripheral pixels closer to the current pixel P among the peripheral pixels close to the extension line 150 may be used as predictors. , Or peripheral pixels close to the extension line 150 can be used for prediction. For example, the average value of peripheral pixels close to the extension line 150 or the weighted average value considering the distance between the intersections of the extension line 150 can be used as the predictor of the current pixel P. Further, as shown in FIG. 17A, as shown in peripheral pixels A and B, either a peripheral pixel on the x-axis or a peripheral pixel on the y-axis that can be used depending on the prediction direction is used. Whether or not to do is signaled for each prediction unit.

015417B and 17C are reference diagrams for explaining the process of generating a predictor when the extension line 150 of FIG. 17A passes between the peripheral pixels of the integer position and not the peripheral pixels of the integer position.

0155Refer to Figure 17B, tan determined by the value by mode (dx, dy)^-1When the extension line 150 having an angle of (dy / dx) passes between the peripheral pixels A151 and B152 at the integer pixel position, as described above, the peripheral pixels A151 and B152 close to the extension line 150 and the extension line 150 A weighted average value that takes into account the distance from the intersection of is currently available as a predictor for pixel P. For example, tan^-1If the distance between the intersection of the extension line 150 with an angle of (dy / dx) and the peripheral pixel A151 is f, and the distance between the intersection and the peripheral pixel B152 is g, then the predictor of the current pixel P is ( It can be obtained as A * g + B * f) / (f + g). Here, f and g are preferably integer-normalized distances. When embodied in actual software or hardware, the current pixel P predictor may be embodied via a shift operation, such as (g * A + f * B + 2) &gt;&gt; 2. As illustrated in FIG. 17B, the extension line 150 passes through a quarter of the points that divide the peripheral pixel A151 at the integer pixel position and the peripheral pixel B152 into four equal parts, which is close to the peripheral pixel A151. If so, the predictor of pixel P is now obtained as (3 * A + B) / 4. Such an operation is also embodied through a shift operation that considers the rounding process, such as (3 * A + B + 2) &gt;&gt; 2.

0156On the other hand, tan determined by the value of each mode (dx, dy)^-1When the extension line 150 having an angle of (dy / dx) passes between the peripheral pixels A151 and B152 at the integer pixel position, the section between the peripheral pixels A151 and the peripheral pixel B152 is divided into a predetermined number, and each division is performed. A weighted average value considering the distance between the peripheral pixel A151 and the peripheral pixel B152 and the intersection can be used as the predicted value for each of the created areas. For example, with reference to FIG. 17C, the section between peripheral pixel A151 and peripheral pixel B152 is shown in five sections P, as shown.₁Or P₅For each section, determine the representative weighted average value considering the distance between the peripheral pixel A151 and the peripheral pixel B152 and the intersection, and use such a representative weighted average value as the predictor of the current pixel P. be able to. Specifically, the extension line 150 is the section P₁Can now determine the value of peripheral pixel A as a predictor of pixel P. Extension line 150 is section P₂When passing through, section P₂(3 * A + 1 * B + 2) &gt;&gt; 2, which is a weighted average value considering the distance between the intermediate point of and the peripheral pixel A and the peripheral pixel B, can be determined as the predictor of the current pixel P. it can. Extension line 150 is section P₃When passing through, section P₃(2 * A + 2 * B + 2) &gt;&gt; 2, which is a weighted average value considering the distance between the intermediate point of and the peripheral pixel A and the peripheral pixel B, can be determined as the predictor of the current pixel P. it can. Extension line 150 is section P₄When passing through, section P₄(1 * A + 3 * B + 2) &gt;&gt; 2, which is a weighted average value considering the distance between the intermediate point of and the peripheral pixel A and the peripheral pixel B, can be determined as the predictor of the current pixel P. it can. Extension line 150 is section P₅If you pass through, you can now determine the value of the peripheral pixel B as the predictor of pixel P.

0157Also, as shown, when there are two peripheral pixels, the upper peripheral pixel A and the left peripheral pixel B that meet the extension line 150, the average value of the upper peripheral pixel A and the left peripheral pixel B. Is currently used as a predictor of pixel P, or if the dx * dy value is positive, the upper peripheral pixel A is used and the left side if the dx * dy value is negative. Peripheral pixel B of can be used.

0158The intra prediction mode having various directions as shown in Table 2 is preset at the coding end and the decoding end, and only the relevant index of the intra prediction mode set for each coding unit is transmitted. It is desirable to let it.

0159FIG. 29 is a diagram illustrating the current coding unit 2900 and the peripheral pixels 2910, 2920 used for intra prediction according to the embodiment of the present invention. With reference to FIG. 29, the upper peripheral pixel 2910 and the left peripheral pixel 2920 of the current coding unit 2900 are utilized for the intra prediction of the current coding unit 2900. In FIG. 29, the lower left pixel contained in the unencoded peripheral block, such as the lower part of the adjacent pixel 2920, is also used for the intra prediction of the current coding unit 2900. The number of peripheral pixels 2910,2920 currently used for intra-prediction of the coding unit 2900 is not limited to this, and can be changed in consideration of the direction of the intra-prediction mode currently applied to the coding unit 2900. Is. Therefore, according to the embodiment of the present invention, the peripheral pixels currently used for the intra prediction of the coding unit may include not only the peripheral pixels on the left side and the upper side but also the peripheral pixels on the lower left side and the upper right side.

0160FIG. 18 is a reference diagram for explaining a bilinear mode according to an embodiment of the present invention. Referring to FIG. 18, the bilinear mode is centered on the current pixel P to be predicted inside the current coding unit, the pixel value of the current pixel P's upper, lower, left and right boundaries, and the current pixel P's upper, lower, left and right. The geometric mean value is calculated considering the distance to the boundary, and the resulting value is used as the predictor of the current pixel P. That is, in the bilinear mode, the predictors of the current pixel P are the pixels A 161, pixel B 162, pixel D 166 and pixel E 167 currently located at the vertical and horizontal boundaries of the pixel P, and up to the vertical and horizontal boundaries of the current pixel P. Use the geometric mean value of the distance of. At this time, since the bilinear mode is also one of the intra prediction modes, the peripheral pixels of the upper side and the left side that have been previously encoded and then restored must be used as the reference pixels at the time of prediction. Therefore, as pixel A 161 and pixel B 162, the virtual pixel value generated by using the peripheral pixels on the upper side and the left side is used instead of using the pixel value currently inside the coding unit as it is.

0161As an example, first, as shown in the following equation (1), the average value of the upper leftmost peripheral pixel (RightUpPixel) 164 and the leftmost lowermost peripheral pixel (LeftDownPixel) 165 adjacent to the current coding unit is used. Therefore, it is possible to calculate the virtual pixel C 163 at the lowermost position on the right side of the current coding unit.

0162C = 0.5 (LeftDownPixel + RightUpPixel) (1) Equation (1) may be calculated using the shift operation as follows equation C = 0.5 (LeftDownPixel + RightUpPixel + 1) &gt;&gt; 1.

0163Next, considering the distance W1 to the left boundary of the current pixel P and the distance W2 to the right boundary, the value of the virtual pixel A 161 located at the bottom boundary when the current pixel P is extended to the lower end. May be set using the average value of the leftmost peripheral pixel (LeftDownPixel) 165 and pixel C in consideration of the distances W1 and W2. As an example, the value of pixel A 161 is calculated as one of the following equations (2).

0164A = (C * W1 + LeftDownPixel * W2) / (W1 + W2); A = (C * W1 + LeftDownPixel * W2 + ((W1 + W2) / 2)) / (W1 + W2) (2) In equation (2), if W1 + W2 has an exponent value of 2, such as 2 ^ n, then A = (C * W1 + LeftDownPixel * W2 + ((W1 + W2) / 2)) / (W1 + W2) can be calculated via shift operation without division operation like A = (C * W1 + LeftDownPixel * W2 + 2 ^ (n-1)) &gt;&gt; n.

0165Similarly, considering the distance h1 to the upper boundary of the current pixel P and the distance h2 to the lower boundary, the value of the virtual pixel B 162 located at the rightmost boundary when the current pixel P is extended to the right. Can be set by considering the distances h1 and h2 and using the average value of the upper rightmost peripheral pixel (RightUpPixel) 164 and pixel C. As an example, the value of pixel B 162 may be calculated as one of the following equations (3).

0166B = (C * h1 + RightUpPixel * h2) / (h1 + h2); B = (C * h1 + RightUpPixel * h2 + ((h1 + h2) / 2)) / (h1 + h2) (3) In equation (3), if h1 + h2 has an exponential value of 2, such as 2 ^ m, then B = (C * h1 + RightUpPixel * h2 + ((h1 + h2) / 2)) / (h1 + h2) can be calculated via shift operation without division operation, such as B = (C * h1 + RightUpPixel * h2 + 2 ^ (m-1)) &gt;&gt; m.

0167If equations (1) to (3) are used to determine the values of the virtual pixel A 161 currently on the lower border of pixel P 160 and the virtual pixel B 162 on the right border, then A + B The average value of + D + E can be used to determine the predictor of the current pixel P 160. Specifically, as the predictor of the current pixel P 160, the above-mentioned average value of A + B + D + E can be used, or the current pixel P 160 and the virtual pixel A 161, the virtual pixel B 162, A weighted average value that takes into account the distances from pixel D 166 and pixel E 167 can be used. For example, in Figure 18, where the block size is 16x16 and the weighted mean is used, the predictor of the current pixel P is (h1 * A + h2 * D + W1 * B + W2 * E + 16). ) &gt;&gt; 5 and so on. In this way, the bilinear prediction process is applied to every pixel inside the current coding unit, and the predicted coding unit of the current coding unit is generated by the bilinear prediction mode.

0168According to one embodiment of the present invention, predictive coding is performed by an intra-prediction mode variously set according to the size of the coding unit, thereby enabling more efficient compression depending on the characteristics of the image.

0169According to one embodiment of the present invention, the size of the coding unit is compatible with the conventional codec in order to utilize a large number of intra-prediction modes as compared with the number of intra-prediction modes used in the conventional codec. Can be a problem. According to prior art, up to nine intra-prediction modes are available, as illustrated in FIGS. 13A and 13B. Therefore, it is necessary to map the intra-prediction modes in various directions selected by one embodiment of the present invention with one of a smaller number of intra-prediction modes. That is, when the number of available intra prediction modes of the current coding unit is N1 (N1 is an integer), N2 (N2 is an integer) intra prediction modes different from the currently available intra prediction modes of the coding unit. For compatibility of coding units of a given size with, the intra prediction mode of the current coding unit can be mapped to the intra prediction mode in the most similar direction of the N2 intra prediction modes. For example, for the current coding unit, a total of 33 intra-prediction modes are available, as shown in Table 2 above, and the finally applied intra-prediction mode for the current coding unit is mode 14, ie. , (Dx, dy) = (4,3), tan^-1It is assumed that the directionality is (3/4) 36.87 (°). In this case, the direction of 36.87 (°) and most to match the intra prediction mode currently applied to the block with one of the nine intra prediction modes as illustrated in Figures 15A and 15B. Mode 4 (down_right) mode with a similar direction is selected. That is, mode 4 in Table 2 may be mapped to mode 4 illustrated in FIG. 15A. Similarly, the intra-prediction mode currently applied to the coding unit is mode 15, i.e. (dx, dy) = (1,11) in all 33 available intra-prediction modes in Table 2. If selected, the direction of the intra-prediction mode currently applied to the coding unit is tan.^-1(11) Since it has 84.80 (°), it can be mapped to the mode 0 (vertical) mode in Fig. 13B, which is most similar to such a direction.

0170On the other hand, in order to decode the coding unit encoded via intra-prediction, prediction mode information relating to what intra-prediction mode the coding unit is currently encoded is required. Therefore, when the video is coded, the information about the intra prediction mode of the current coding unit is added to the bit stream, but if the intra prediction mode information is added to the bit stream as it is for each coding unit, the overhead increases. However, the compression efficiency may be low. Therefore, the information about the intra prediction mode of the current coding unit determined as a result of the coding of the current coding unit is not transmitted as it is, but the predicted value of the intra prediction mode predicted from the peripheral coding unit. And, only the difference value from the actual intra prediction mode is transmitted.

0171When using the intra-prediction modes in various directions selected by one embodiment of the present invention, the number of available intra-prediction modes may differ depending on the size of the coding unit. Therefore, in order to predict the intra prediction mode of the current coding unit, it is first necessary to map the intra prediction mode of the peripheral coding unit to the representative intra prediction mode. Here, the representative intra prediction mode is a smaller number of intra prediction modes among the available peripheral coding unit intra prediction modes, or nine intra prediction modes as shown in FIG. Is desirable.

0172FIG. 19 is a diagram for explaining a process of generating a predicted value of the intra prediction mode of the current coding unit by one embodiment of the present invention. With reference to FIG. 19, when the current coding unit is A 170, the intra prediction mode of the current coding unit A 170 can be predicted from the intra prediction mode determined by the peripheral coding unit. For example, if the determined intra prediction mode of the left coding unit B 171 of the current coding unit A 170 is mode 3 and the intra prediction mode of the upper coding unit C 172 is mode 4, then the current code The intra prediction mode of the conversion unit A 170 is predicted as mode 3 having the smaller value of the prediction modes of the upper coding unit C 172 and the left coding unit B 171. If the intra prediction mode determined as a result of the actual intra prediction coding related to the current coding unit A 170 is mode 4, the value of the intra prediction mode predicted from the peripheral coding unit is used as the intra prediction mode information. Mode Only 1 which is the difference from 3 is transmitted, and at the time of decoding, the predicted value of the intra prediction mode of the current decoding unit is generated by the same method as described above, and the mode difference transmitted via the bit stream. The value can be added to the predicted value of the intra prediction mode to obtain the intra prediction mode information actually applied to the currently decoded unit. In the above description, only the peripheral coding units currently located above and to the left of the coding unit are used, but other peripheral codes such as E and D in FIG. 19 are also used. The conversion unit can be used to predict the intra-prediction mode of the current coding unit A 170.

0173On the other hand, since the intra prediction mode actually executed differs depending on the size of the coding unit, the intra prediction mode predicted from the peripheral coding unit may not match the intra prediction mode of the current coding unit. Therefore, in order to predict the intra prediction mode of the current coding unit from the peripheral coding units having different sizes, a mapping process for unifying the intra prediction modes between the coding units having different intra prediction modes is required. is there.

017420A and 20B are reference diagrams for explaining the mapping process of the intra prediction mode between coding units having different sizes according to the embodiment of the present invention.

0175With reference to FIG. 20A, it is assumed that the current coding unit A 180 has a size of 16x16, the left coding unit B 181 has a size of 8x8, and the upper coding unit C 182 has a size of 4x4. In addition, as in the first example (Example 1) of FIG. 14 described above, the number of intra prediction modes that can be used in the coding units of 4x4, 8x8, 16x16 size differs from 9, 9, 33, respectively. Suppose. In this case, the left coding unit B 181 is different from the intra prediction mode available in the left coding unit B 181 and the upper coding unit C 182, and the intra prediction mode currently available in the coding unit A 180. And the intra prediction mode predicted from the upper coding unit C 182 is not suitable for use as the prediction value of the intra prediction mode of the current coding unit A 180. Therefore, according to one embodiment of the present invention, the peripheral coding unit B 181 and the peripheral coding unit C The 182 intra prediction modes are changed to the 1st representative intra prediction mode and the 2nd representative intra prediction mode in the most similar directions among the predetermined number of representative intra prediction modes, respectively, and the 1st representative intra prediction mode and the 2nd representative intra prediction mode are changed. Select as the final representative intra prediction mode with a smaller mode value of the prediction modes. Then, the intra prediction mode having the direction most similar to the representative intra prediction mode selected from the currently available intra prediction modes according to the size of the currently coded unit A 180 is predicted as the intra prediction mode of the current coding unit.

0176As another example, referring to FIG. 20B, it is assumed that the current coding unit A has a size of 16x16, the left coding unit B has a size of 32x32, and the upper coding unit C has a size of 8x8. In addition, as in the first example (Example 1) of FIG. 14 described above, the number of intra prediction modes that can be used in the coding units of 8x8, 16x16, 32x32 size is different from 9,9,33, respectively. Suppose. Further, it is assumed that the intra prediction mode of the left coding unit B is mode 4 and the intra prediction mode of the upper coding unit C is mode 31. In this case, since the intra prediction modes of the left coding unit B and the upper coding unit C are not compatible, the intra prediction modes of the left coding unit B and the upper coding unit C are shown in FIG. 21, respectively. Map to one of the representative intra prediction modes. As shown in Table 2, mode 31, which is the intra-prediction mode of the left-hand coding unit B, has the direction of (dx, dy) = (4, -3), so that the representative intra-prediction in FIG. 21 Of the modes, tan^-1Mode 4, which is mapped to mode 5 having the direction most similar to (-3 / 4) and is the intra prediction mode of the upper coding unit C, is the same direction as mode 4 among the representative intra prediction modes in FIG. Since it has sex, it is mapped to mode 4 as it is.

0177Next, of the mapped intra prediction mode mode 5 of the left coding unit B and the mapped intra prediction mode mode 4 of the upper coding unit C, mode 4 having a smaller mode value is currently encoded. Determined as the predicted value of the unit's intra-prediction mode, only the mode difference between the actual intra-prediction mode of the current coding unit and the predicted intra-prediction mode is encoded as the prediction mode information of the current coding unit. ..

0178FIG. 21 is a reference diagram for explaining a process of mapping an intra prediction mode of a peripheral coding unit to one of the representative intra prediction modes according to an embodiment of the present invention. In FIG. 21, vertical mode, horizontal mode, DC (direct current) mode, diagonal left side mode, diagonal right side mode, vertical right side mode, vertical left side mode, horizontal upper mode, and horizontal lower side mode are set as representative intra prediction modes. The case is shown in the figure. However, the representative intra prediction mode is not limited to the one shown in the figure, and may be set to have various numbers of directions.

0179With reference to FIG. 21, a predetermined number of representative intra prediction modes are set in advance, and the intra prediction mode of the peripheral coding unit is mapped to the representative intra prediction mode in the most similar direction. For example, if the determined intra prediction mode of the upper coding unit A has the direction shown as MODE_A 190, the intra prediction mode MODE_A 190 of the upper coding unit A has nine preset representative intras. It maps to MODE 1 which has the most similar direction of prediction modes 1-9. Similarly, if the determined intra prediction mode of the left coding unit B has the direction shown as MODE_B 191 then the intra prediction mode MODE_B 191 of the left coding unit B has nine preconfigured representative intras. It is mapped to MODE 5 which has the most similar direction among prediction modes 1 to 9.

0180Next, of the first representative intra prediction mode and the second representative intra prediction mode, the intra prediction mode having a smaller mode value is selected as the representative intra prediction mode of the final peripheral coding unit. As described above, the reason for selecting the representative intra prediction mode having a smaller mode value is that the smaller mode value is generally set by the more frequently occurring intra prediction mode. That is, when different intra-prediction modes are predicted from peripheral coding units, and because the intra-prediction modes having smaller mode values are the prediction modes having a higher probability of occurrence, the different prediction modes compete with each other. It is desirable to select a prediction mode with a smaller mode value as the predictor of the prediction mode of the current coding unit.

0181Even if the representative intra prediction mode is selected from the peripheral coding unit, the representative intra prediction mode may not be used as it is as a predictor of the intra prediction mode of the current coding unit. As shown in FIG. 20, the representative intra prediction mode corresponds to the representative intra prediction mode if the current coding unit A 180 has 33 intra prediction modes and only 9 intra prediction modes. Currently, there is no intra-prediction mode for coding unit A 180. In such a case, the intra-prediction mode of the peripheral coding unit described above is most similar to the representative intra-prediction mode selected among the intra-prediction modes by the size of the current coding unit so as to map to the representative intra-prediction mode. The directional intra-prediction mode can finally be selected as the intra-prediction mode predictor for the current coding unit. For example, in FIG. 21, if the representative intra-prediction mode finally selected from the peripheral coding units is mode 1, then mode of the currently available intra-prediction modes depending on the size of the coding unit. The intra prediction mode having the most similar direction to 1 is finally selected as the predictor of the intra prediction mode of the current coding unit.

0182On the other hand, as explained with reference to FIGS. 15A to 15C above, when the predictor of the current pixel P is generated by using the peripheral pixels located at the extension line 150 or close to the extension line 150, the extension is extended. Line 150 is actually tan^-1It has a direction of (dy / dx), but in order to calculate such a direction, a division operation of (dy / dx) is required. It may be a factor to increase the amount of calculation including. Therefore, according to another embodiment of the present invention, as described in Table 2 above, the prediction direction for selecting peripheral pixels to be used as reference pixels is determined centering on the pixels in the coding unit. The process of setting dx, dy is disclosed so that the amount of calculation can be reduced when setting using the, dx, dy parameters.

0183FIG. 25 is a diagram for explaining the relationship between the peripheral pixels located on the extension line having the direction of (dx, dy) and the current pixels according to the present invention. With reference to FIG. 25, the position of the current pixel P is P (j, i), and the tan currently passes through the pixel P.^-1The directionality of (dy / dx), that is, the upper peripheral pixel located on the extension line 2510 having the inclination is defined as A, and the left peripheral pixel is defined as B. Assuming that the position of the upper peripheral pixel corresponds to the x-axis on the coordinate plane and the position of the left peripheral pixel corresponds to the y-axis on the coordinate plane, the extension line 2510 indicates that the upper peripheral pixel A corresponds to (j +. On i * dx / dy, 0), the left peripheral pixel B is located on (0, i + j * dy / dx). Therefore, a division operation such as dx / dy or dy / dx is required to determine either the upper peripheral pixel A or the left peripheral pixel B for the prediction of the current pixel P. As described above, such a division operation may cause a decrease in operation speed when it is realized by software or hardware due to its high operation complexity.

0184Therefore, in another embodiment of the present invention, at least one value of dx and dy indicating the direction of the prediction mode for determining the peripheral pixels can be determined as a power of 2. That is, when n and m are integers, respectively, dx and dy may be 2 ^ n and 2 ^ m, respectively.

0185Referring to FIG. 25, if the left peripheral pixel B is currently used as the predictor of pixel P and dx has a value of 2 ^ n, it is the position of the left peripheral pixel B (0, i + j * dy /). The j * dy / dx operation required to determine dx) is (i * dy) / (2 ^ n), and the operation that divides by an exponent of 2 is (i * dy). ) &gt;&gt; n, which can be realized via shift calculation, so the amount of calculation is reduced.

0186Similarly, if the upper peripheral pixel A is currently used as the predictor of the pixel P and dy has a value of 2 ^ m, then the position of the upper peripheral pixel A is determined (j + i * dx / dy, 0). The i * dx / dy operation required to do this is (i * dx) / (2 ^ m), and the operation that divides by the exponential power of 2 is (i * dx) &gt;&gt; m. It can be realized through a shift operation.

0187FIG. 26 is a diagram for explaining a change in peripheral pixels located on an extension line having a direction of (dx, dy) depending on the current pixel position according to the present invention.

0188Depending on the current pixel position, one of the upper peripheral pixels or the left peripheral pixel is selected as the peripheral pixel required for prediction.

0189Referring to FIG. 26, suppose that the current pixel is P (j, i) illustrated by drawing reference numeral 2610 and is predicted using peripheral pixels located in the prediction direction as shown. If the upper pixel A is used when predicting the current pixel P 2610 and the current pixel is Q (b, a) illustrated by drawing code 2610, then the left pixel B is when predicting the current pixel Q 2620. It will be used.

0190If only the dy component in the y-axis direction has the exponentiation value of 2 in the 2 ^ m form of the (dx, dy) pointing in the prediction direction, the upper pixel A in FIG. 24 is (j + (i * dx). ) &gt;&gt; m, 0) can be determined by a shift operation without a division operation, but the left pixel B requires a division operation like (0, a + b * 2 ^ m / dx). become. Therefore, both dx and dy can have an exponential form of 2 in order to exclude the division operation when the predictor is generated for every pixel of the current block.

019127 and 28 are diagrams for explaining a method of determining the intra prediction mode direction according to another embodiment of the present invention.

0192In general, the linear pattern that appears in a video or video signal is often in the vertical or horizontal direction. Therefore, when defining an intra-prediction mode with various directions using the parameters of (dx, dy), the coding efficiency of the video is improved by defining the values of dx, dy as follows. be able to.

0193Specifically, when dy has a fixed value of 2 ^ m, the absolute value of dx is the prediction direction close to the horizontal direction, with the spacing between the prediction directions close to the vertical direction set narrow. The more the interval between the prediction modes can be set, the wider the interval can be set. For example, referring to Figure 27, if dy has a value of 2 ^ 4, i.e. 16, then the value of dx is 1, 2, 3, 4, 6, 9, 12, 16, 0, -1,-. By setting 2, -3, -4, -6, -9, -12, -16, the interval between the prediction directions close to the vertical direction is set narrow and the prediction direction is close to the horizontal direction. The more the interval between the prediction modes can be set, the wider the interval can be set.

0194Similarly, when dx has a fixed value of 2 ^ n, the absolute value of dy is such that the spacing between prediction directions closer to the horizontal direction is set narrower and the prediction direction closer to the vertical direction. The interval between prediction modes is set to be wide. For example, referring to FIG. 28, if dx has a value of 2 ^ 4, i.e. 16, then the value of dy is 1, 2, 3, 4, 6, 9, 12, 16, 0, -1, By setting such as -2, -3, -4, -6, -9, -12, -16, the interval between the prediction directions close to the horizontal direction is set narrow and the prediction direction close to the vertical direction is set. The more it is, the wider the interval between the prediction modes is set.

0195Further, when any one of dx and dy is fixed, the remaining non-fixed values may be set to increase for each prediction mode. For example, when dy is fixed, the interval between dx may be set to increase by a predetermined value. Further, such an increase width may be set by dividing the angle between the horizontal direction and the vertical direction by a predetermined unit and by the divided angle. For example, when dy is fixed, the value of dx has an increase of a in the interval where the angle with the vertical axis is within 15 °, and an increase of b between 15 ° and 30 °. And above 30 °, it may be set to have an increase width of c. In this case, in order to have the form as shown in FIG. 25 described above, a &lt;b &lt;c may be set.

0196As an example, the prediction mode according to another embodiment of the present invention described with reference to FIGS. 23 to 28 described above utilizes (dx, dy) as shown in Tables 3 to 5 below. And tan^-1It can be defined as a prediction mode with a direction of (dy / dx).

0197<tables num="3"><img id="000004" he="211" wi="126" file="JP5770185B2_D0001.tif" img-format="tif" img-content="drawing" /></tables><img id="000005" he="104" wi="132" file="JP5770185B2_D0001.tif" img-format="tif" img-content="drawing" /> FIG. 22 is a block diagram showing an intra-prediction device for video according to an embodiment of the present invention. The video intra-prediction device 2000 according to an embodiment of the present invention may operate as the intra-prediction unit 410 of the video coding device 400 of FIG. 4 and the intra-prediction unit 550 of the video decoding device 500 of FIG.

0198With reference to FIG. 22, the intra-prediction mode determination unit 2010 applies the intra-prediction mode currently applied to the coding unit according to the size of each coding unit divided based on the maximum coding unit and the depth, as described above. To determine. That is, the intra-prediction mode determination unit 2010 selects an intra-prediction mode candidate to be applied according to the size of the current coding unit from the intra-prediction modes in various directions.

0199The intra-prediction mode execution unit 2020 applies the determined intra-prediction mode and performs intra-prediction related to each coding unit. In the intra-prediction mode execution unit 2020, the optimum intra-prediction mode having the minimum error value is the coding unit based on the error value between the prediction coding unit generated as a result of the intra-prediction and the original coding unit. Is determined as the final intra-prediction mode of.

0200On the other hand, when the intra-prediction device 2000 illustrated in FIG. 22 is used as the decoding device, the intra-prediction mode determination unit 2010 is the maximum extracted from the bitstream encoded by the entropy decoding unit 520 of FIG. The size of the current decoding unit is determined using the depth information, which is the hierarchical division information of the coding unit and the maximum coding unit, and the execution is performed using the intra prediction mode information currently applied to the decoding unit. Determine the intra prediction mode. In addition, the intra prediction mode execution unit 2020 executes the intra prediction related to the decoding unit by the extracted intra prediction mode to generate the prediction decoding unit. The predictive decoding unit is added to the residual data restored from the bitstream, and the decoding related to the decoding unit is performed.

0201FIG. 23 is a flowchart showing a video coding method according to an embodiment of the present invention. Referring to FIG. 23, at step 2110, the current picture is now divided into at least one block. As described above, the current picture is divided based on the maximum coding unit, which is the coding unit having the maximum size, and the depth, which is the hierarchical division information of the maximum coding unit.

0202At stage 2120, the size of the divided current block determines the intra-prediction mode performed for the current block. As mentioned above, according to one embodiment of the invention, the intra-prediction mode is currently centered on each pixel inside the block, tan.^-1Includes a prediction mode that makes predictions using pixels in peripheral blocks that are located on or near the extension line with an angle of (dy / dx).

0203In stage 2130, the determined intra-prediction mode is used to make intra-prediction for the current block. Among the intra prediction modes, the intra prediction mode having the smallest prediction error value is selected as the final intra prediction mode of the current block.

0204FIG. 24 is a flowchart showing a video decoding method according to an embodiment of the present invention. Referring to FIG. 24, at step 2210, the current picture is divided into at least one block of a predetermined size. Here, the current picture is divided based on the maximum decoding unit which is the decoding unit having the maximum size extracted from the bit stream and the depth information which is the hierarchical division information of the maximum decoding unit.

0205At step 2220, the intra-prediction mode information currently applied to the block is extracted from the bitstream. Intra prediction mode is currently centered around each pixel of the block, tan^-1It includes a prediction mode in which prediction is performed using pixels of peripheral blocks located on an extension line having an angle of (dy / dx) (dx, dy is an integer) or close to the extension line. As described with reference to FIGS. 19 to 21, when using the intra prediction mode predictor predicted from the intra prediction mode of the peripheral decoding unit, the intra prediction mode currently possessed by the peripheral decoding unit of the decoding unit is used. After mapping to the representative intra prediction mode, the representative intra prediction mode having a smaller mode value among the representative intra prediction modes is selected as the final representative intra prediction mode. Then, among the available intra-prediction modes determined by the size of the currently decoding unit, the intra-prediction mode having the direction most similar to the final representative intra-prediction mode is used as the predictor of the intra-prediction mode of the current decoding unit. By selecting, extracting the difference between the actual intra-prediction mode provided in the bit stream and the predictor of the intra-prediction mode, and adding this to the predictor of the intra-prediction mode, the intra-prediction mode of the current decoding unit Can be determined.

0206At step 2230, the extracted intra-prediction mode performs intra-prediction related to the decoding unit to decode the decoding unit.

0207On the other hand, the above-described embodiment of the present invention can be created by a computer-executable program, and is embodied in a general-purpose digital computer that operates the program by using a computer-readable recording medium. The computer-readable recording medium is a magnetic recording medium (for example, ROM (read-only memory), floppy (registered trademark) disk, hard disk, etc.), and an optical recording medium (for example, CD-ROM or DVD). ..

0208The present invention has been described above, focusing on its preferred embodiments. Those skilled in the art to which the present invention belongs will appreciate that the present invention can be embodied in a modified form to the extent that it does not deviate from the essential properties of the present invention. Therefore, the disclosed embodiments must be considered from a descriptive point of view, not from a limited point of view. The scope of the present invention is shown in the claims, not in the above description, and any differences within the equivalent scope must be construed as included in the present invention. is there.