Apparatus for generating residual blocks

Abstract

To provide a favorable video data decoding device.SOLUTION: An apparatus is configured to: derive a luma intra prediction mode and a chroma intra prediction mode; determine a size of a luma transform unit and a size of a chroma transform unit using luma transform size information; adaptively filter reference pixels of a current luma block on the basis of the luma intra prediction mode and the size of the luma transform unit; generate prediction blocks of a luma block and a chroma block; and generate a luma residual block and a chroma residual block. Therefore, a distance of intra prediction becomes short, and an amount of coding bits required to encode intra prediction modes and residual signals of luma and chroma components is reduced and coding complexity is reduced by adaptively encoding the intra prediction modes and adaptively filtering the reference pixels.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a video decoding device, and more particularly to a method for decoding a luminance and chrominance component block. To recover the original image, an intra prediction mode is derived, and a prediction block and a residual block are generated. This relates to an apparatus for [Background technology]

[0002] In H.264 / MPEG-4 AVC, one picture is divided into multiple macroblocks. and generating a predicted block using intra prediction or inter prediction. Each macroblock is coded. The difference between the original block and the predicted block is converted. A transform block is generated by the quantization parameter and a plurality of predetermined quantization matrices. The transform block is quantized using one of the quantization matrices. The quantization coefficients of the quantization block are scanned according to a predetermined scan type. The quantization parameter is adjusted for each macroblock, The quantization parameter is coded using the previous quantization parameter as a predictor.

[0003] On the other hand, to improve coding efficiency, various sizes of coding units are used. In addition, a technique is introduced to generate a predicted block that is more similar to the original block. To address this issue, techniques for increasing the number of intra prediction modes have also been introduced.

[0004] However, as the number of intra-prediction modes increases, the coding requirements for the intra-prediction modes become The amount of coding bits to be used increases. Also, the size of the coding unit increases. As the value of the difference between the original block and the predicted block increases, the difference between the original block and the predicted block increases. A novel and efficient method for encoding and decoding luminance and chrominance component video data. is required. Summary of the Invention [Problem to be solved by the invention]

[0005] The object of the present invention is to induce an intra prediction mode and adaptively select reference pixels. The present invention provides an apparatus for filtering and generating a prediction block. [Means for solving the problem]

[0006] The video decoding apparatus according to the present invention is configured to select a luma intra prediction mode and a chroma intra prediction mode. A prediction mode decoder determines the size of the luminance transform unit using the luminance transform size information. and a prediction size determination unit for determining the size of a color difference transform unit; a reference pixel generator that generates reference pixels when the reference pixel is not available; Adaptively filter the reference pixels of the current luminance block based on the size of the luminance transformation unit. Reference pixel filtering: Filters the reference pixels of the current chrominance block without filtering. a prediction block generation unit for generating a prediction block of the current luminance block and the current chrominance block; and a residual block generator for generating a luminance residual block and a chrominance residual block. It is characterized by: The present invention further provides, for example: (Item 1) 1. An apparatus for decoding video data, comprising: a prediction mode decoding unit for deriving a luma intra prediction mode and a chroma intra prediction mode; The size of the luminance conversion unit and the size of the chrominance conversion unit are calculated using the luminance conversion size information. a prediction size determination unit that determines a reference pixel generator for generating reference pixels if at least one reference pixel is not available; Reference of the current luma block based on the luma intra prediction mode and the size of the luma transform unit. The pixels are adaptively filtered, and the reference pixels of the current chrominance block are not filtered. A reference pixel filtering unit; a prediction block generator for generating a prediction block of the current luminance block and the current chrominance block; and Beauty, a residual block generator for generating a luminance residual block and a chrominance residual block; A device for doing so. (Item 2) When the color difference intra prediction mode is a DM mode, the color difference intra prediction mode is 2. The device according to item 1, characterized in that the mode is the same as the luma intra prediction mode. (Item 3) The luma intra prediction mode is a set of three MPM candidates, a mode group indicator, and a prediction mode. The device described in item 1, characterized in that it is guided by utilizing a dot index. (Item 4) Only one of the left intra prediction mode and the above intra prediction mode is available. In this case, the available intra-prediction modes and the available intra-prediction modes are used to determine the available intra-prediction modes. The two additional intra prediction modes are set to the three MPM candidates. Item 3. The device according to claim 3. (Item 5) If the available intra-prediction modes are non-directional modes, one remaining non-directional mode and a vertical mode is set as the additional intra prediction mode. 4. The apparatus according to claim 4. (Item 6) If the size of the prediction unit is larger than the size of the transform unit, the chrominance intra prediction is performed. The prediction mode is used to generate a chrominance prediction block for another block in the chrominance prediction unit. 2. The device according to item 1, characterized in that it is used. Effect of the Invention

[0007] Induces luma intra prediction mode and chroma intra prediction mode, and uses luma transform size information. The size of the luminance conversion unit and the size of the chrominance conversion unit are determined using the Adaptively adjusts the reference pixels of the current luma block based on the prediction mode and the size of the luma transform unit. Then, the luminance residual block is filtered to generate the predicted luminance and chrominance blocks. Therefore, the distance of intra prediction is reduced, The amount of bits required for coding the intra prediction modes of the luminance and chrominance components and the residual signal is reduced. A little bit, adaptive coding of intra prediction modes and adaptive filtering of reference pixels This can reduce the coding complexity. [Brief description of the drawings]

[0008] [Figure 1] 1 is a block diagram showing a video encoding device according to the present invention; [Diagram 2] 1 is a block diagram showing a video decoding device according to the present invention; [Diagram 3] FIG. 2 is a block diagram showing an apparatus for generating a prediction block according to the present invention; [Figure 4] FIG. 2 is a conceptual diagram illustrating an intra prediction mode according to the present invention. [Diagram 5] FIG. 2 is a block diagram showing an apparatus for generating a residual block according to the present invention; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Various embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It will be appreciated that modifications and variations of the present invention may be made without departing from the spirit and scope of the present invention. The present invention is not limited to the above, and all modifications and equivalents included in the spirit and technical scope of the present invention are intended to be included. In the description of the various figures, similar elements are designated by the same reference numerals. Similar reference numbers have been used for the same.

[0010] FIG. 1 is a block diagram showing a video encoding device 100 according to the present invention.

[0011] Referring to FIG. 1, a video encoding device 100 according to the present invention includes a picture division unit 110, an image division unit 120, an image division unit 130, an image division unit 140, an image division unit 150, an image division unit 160, an image division unit 170, an image division unit 180, an image division unit 190, an image division unit 20 ... Intra prediction unit 120, inter prediction unit 130, transform unit 140, quantization unit 150, scanning a dequantization unit 160, an entropy coding unit 170, an inverse quantization unit 155, an inverse transform unit 145, and a post-processing unit 166. The image processor 100 includes a processing unit 180, a picture store unit 190, a subtraction unit 192 and an addition unit 194.

[0012] The picture division unit 110 divides a picture into slices, and divides the slices into a plurality of LCUs ( Each LCU is divided into one or more Largest Coding Units (LCUs). The picture division unit 110 divides the picture into coding units. Determines the prediction mode and size of the prediction unit. A unit consists of an array of luminance components (luminance array) and two arrays of chrominance components (chrominance arrays). The chrominance block is half the height and half the width of the luminance block. A block is an LCU, a coding unit, or a prediction unit. The coding unit, the luma prediction unit and the luma transformation unit are each a coding unit , prediction unit and transform unit.

[0013] One LCU consists of one or more coding units. The LCU includes a recursive quad to indicate a division structure of a coding unit. It has a recursive quadtree structure. The parameters indicating the maximum and minimum size of the coding unit are included in the sequence parameters. The split structure is included in the sequence parameter set. The structure is one or more split coding unit flags (split_cu_flag ) The coding unit has a size of 2N×2N.

[0014] A coding unit is made up of one or more prediction units. In intra prediction, the size of the prediction unit is 2N×2N or In inter prediction, the size of the prediction unit is 2N×2N, 2N ×N, N×2N or N×N.

[0015] A coding unit consists of one or more transform units. The transformation unit includes a recursive quadtree structure (r The partition structure is 1 The split_tu_flag is a number of split transform unit flags. The parameters indicating the maximum and minimum size of the transformation unit are the sequence parameters. If the transform unit is not 4×4, the chroma transform unit is included in the transform set. The minimum size of the chrominance conversion unit is 1 / 2 the width and 1 / 2 the height of the chrominance conversion unit. , 4×4.

[0016] The intra prediction unit 120 determines an intra prediction mode of the current prediction unit, A prediction block is generated using the transform prediction mode. The prediction block is the same as the transform unit. The 100 mm diameter 110 mm thick 120 mm thick 130 mm thick 140 mm thick

[0017] The inter prediction unit 130 predicts one or more reference pictures stored in the picture storage unit 190. The motion information of the current prediction unit is determined using the motion vector, and the prediction block of the current prediction unit is determined using the motion vector. The motion information includes one or more reference picture indexes and one or more motion vectors. Includes Kutlu.

[0018] The transform unit 140 transforms a residual signal generated by using an original block and a predicted block. The residual signal is transformed into a transform unit. The transform type is The type of transformation is determined by the prediction mode and the size of the transform unit. In inter prediction, DCT-based integer transform is used. In intra prediction, the size of the transform unit is set to a predetermined size. If the size is smaller than the limit, use DST-based integer transform, otherwise use DCT-based The chrominance conversion unit uses integer transform. The predetermined size is 8×8. The transform type of each transform block is the same as that of the corresponding transform block. The transform type of the transform unit is a DCT-based integer transform.

[0019] The quantization unit 150 determines a quantization parameter for quantizing the transform block. The quantization parameter is the quantization step size. The luminance quantization parameter is the quantization The quantization parameter is determined for each quantization unit. The size of the quantization unit is the minimum size of the coding unit. The size of the coding unit is one of the allowable sizes of the quantization unit. If the coding unit is larger than or equal to the minimum size of the unit, the coding unit is Multiple coding units can be included in the minimum quantization unit. The minimum size of the quantization unit is determined for each picture, and the maximum size of the quantization unit is The parameter specifying the small size is included in the picture parameter set. The quantization parameter is determined by the quantization parameter. The mapping relationship between the encoding parameters can be determined for each picture. A parameter indicating the mapping relationship is transmitted in the picture parameter set. The mapping relationship can be changed for each slice. Other parameters can be transmitted in the slice header.

[0020] The quantizer 150 generates a quantization parameter predictor and quantizes the quantization parameter from the quantization parameter. A differential quantization parameter is generated by subtracting a parameter predictor from the differential quantization parameter. is encoded.

[0021] The quantization parameter predictor includes the quantization parameters of the neighboring coding units and It is generated using the quantization parameter of the previous coding unit as follows:

[0022] The left quantization parameter, the upper quantization parameter and the previous quantization parameter are in the order If more than one quantization parameter is available, they are searched in the order given above. The average of the first two available quantization parameters is set as the quantization parameter predictor. If only one quantization parameter is available, the available quantization parameter is set as the quantization parameter predictor. That is, the left and upper quantization parameters are both If both the left and top quantization parameters are available, the average of the left and top quantization parameters is used as the quantization parameter. Only one of the left and top quantization parameters is available. If so, the average value of the available quantization parameter and the previous quantization parameter is The quantization parameter predictor is set to the left and top quantization parameters. If not available, the previous quantization parameter is set to the quantization parameter predictor. The average value is rounded off.

[0023] The quantization unit 150 converts the transform block into a quantization matrix and a quantization parameter. The quantization block is then passed to the inverse quantization unit 155. The signal is provided to the scanning unit 160.

[0024] The scanning unit 160 determines a scan pattern and calculates the scan pattern based on the amount of light. For entropy coding, if CABAC is used, The scan pattern is determined as follows.

[0025] In intra prediction, the scan pattern is determined based on the intra prediction mode and the transform unit. The size of the transform unit, the size and quantity of the transform block are determined by the size of the unit. The size of the child blocks is the same. Diagonal scan , vertical scan and horizontal scan The scan pattern is determined from the quantized data of the quantization block. The transform coefficients are marked with significant flags, coefficient signs, and ficient signs and coefficient levels The scan pattern is divided into the importance flag, the coefficient sign and the coefficient level. The significance flag indicates whether the corresponding quantized transform coefficient is zero or not. The coefficient sign indicates the sign of a non-zero quantized transform coefficient. Indicates the absolute value of the conversion coefficient.

[0026] When the size of the transformation unit is smaller than or equal to the first size, the vertical mode and the vertical A horizontal scan is selected for a predetermined number of intra prediction modes adjacent to the vertical mode. a horizontal mode and a predetermined number of intra-prediction modes adjacent to the horizontal mode; selects vertical scan, while the remaining intra prediction modes select diagonal scan. If the size of the transform unit is larger than the first size, a diagonal scan is used. The first size is 8×8. When the size of the transform unit is 8×8, The predetermined number is eight.

[0027] In inter prediction, a predefined scan pattern is used regardless of the size of the transform unit. The predetermined scan pattern is a diagonal scan.

[0028] The scan pattern of the chrominance conversion unit is the same as the scan pattern of the corresponding luma conversion unit. Therefore, if the size of the chrominance transform unit is 4 × 4, the diagonal scan The scan pattern is determined from among a vertical scan, a horizontal scan, and a chrominance transformation. If the unit size is greater than 4x4, a diagonal scan is used.

[0029] If the size of the transform unit is greater than the second size, the quantization block a plurality of remaining subsets, and the determined scan pattern is The significance flag, coefficient sign and coefficient level of each subset are The main subset is scanned by the determined scan pattern. The residual subset includes coefficients that cover an area other than that covered by the main subset. The second size is 4×4. The subset includes 16 transform coefficients. The chrominance subset is also a 4×4 block containing 16 transform coefficients. be.

[0030] A scan pattern for scanning the subsets is a quantized version of each of the subsets. The scan pattern for scanning the transform coefficients obtained by the quantization of each subset is the same as the scan pattern for scanning the transform coefficients obtained by the quantization of each subset. The transformed coefficients are then scanned backwards. The subset is also scanned backwards. .

[0031] last non-zero coefficient p osition) is encoded and sent to the decoder. The last non-zero coefficient position is the decoded The last non-zero coefficient position is used to determine the number of subsets transmitted by the Indicates the position of the last non-zero quantized transform coefficient within a transform unit. The non-zero subset flag indicates the main subset and the last Set for each subset other than the last subset, which is not 0 Covers the last coefficient. The nonzero subset flag indicates that the subset contains nonzero coefficients. Indicates whether to include

[0032] The inverse quantization unit 155 inverse quantizes the quantized transform coefficients of the quantization block.

[0033] The inverse transform unit 145 inverse transforms the dequantized block to generate a spatial domain residual signal.

[0034] The adder 194 adds the residual block and the prediction block to generate a reconstructed block.

[0035] The post-processing unit 180 removes blocking artifacts that occur in the reconstructed picture. A deglossing filtering process is then performed to obtain the

[0036] The picture storage unit 190 receives the post-processed video from the post-processing unit 180 and stores the video in a picture unit. The image is stored in a pixel position. A picture can be a frame or a field.

[0037] The entropy coding unit 170 converts the one-dimensional coefficient information received from the scanning unit 160 into , intra prediction information received from the intra prediction unit 120, and inter prediction information received from the inter prediction unit 130. The transmitted motion information is then entropy coded.

[0038] FIG. 2 is a block diagram illustrating a video decoding device 200 according to the present invention.

[0039] The video decoding apparatus 200 according to the present invention includes an entropy decoding unit 210, a reverse scanning 220, an inverse quantization unit 230, an inverse transform unit 240, an intra prediction unit 250, and an inter prediction unit 260, a post-processing unit 270, a picture storage unit 280 and an adder unit 290.

[0040] The entropy decoding unit 210 extracts intra prediction information, The entropy decoding unit 210 extracts inter prediction information and one-dimensional coefficient information. The inter prediction information is sent to the inter prediction unit 260, and the intra prediction information is sent to the intra prediction unit 25 0, and transmits the coefficient information to the inverse scanning unit 220.

[0041] The inverse scanning unit 220 generates the quantization blocks using the inverse scan pattern. When CABAC is used as the entropy coding method, the inverse scan pattern is It is determined as follows:

[0042] In intra prediction, the inverse scan path is determined according to the intra prediction mode and the size of the transform unit. The reverse scan pattern consists of a diagonal scan, a vertical scan and a horizontal scan. The selected inverse scan pattern is selected from among the significance flag, coefficient code, The inverse sequence of the chrominance transform unit is applied to the quantization signal and the coefficient level, respectively, to generate the quantization block. The scan pattern is the same as the scan pattern of the corresponding luma conversion unit. The minimum size of a transformation unit is 4x4.

[0043] If the size of the transformation unit is smaller than or equal to the first size, the vertical mode and the front A horizontal scan is selected for a predetermined number of intra prediction modes adjacent to the vertical mode. A horizontal mode and a predetermined number of intra prediction modes adjacent to the horizontal mode are selected. For the most part, vertical scanning is selected, while the remaining intra prediction modes select diagonal scanning. If the size of the transform unit is greater than the first size, a diagonal scan The first size is 8×8. When the transform unit is 8×8, The predetermined number is eight.

[0044] In inter prediction, a diagonal scan is used.

[0045] If the size of the transformation unit is greater than the second size, the determined scan pattern The significance flags, coefficient signs and coefficient levels are reverse scanned by the subset. A subset is generated, and the subset is inverse scanned to generate a quantization block. The second size is the same as the size of the subset. The subset is a set of 16 transformations. The chrominance subset is also a 4x4 block. Therefore, If the chrominance transform unit is larger than the second size, a subset is generated, The bits are reverse scanned.

[0046] The inverse scan pattern used to generate each subset produces a quantization block. The inverse scan pattern is the same as that used to calculate the significance flag, coefficient sign, and coefficient register. The signal is scanned backwards. The subset is also scanned backwards.

[0047] The last non-zero position and the non A zero subset flag is received from the encoder. The number of subsets coded is determined by the scan pattern. The flags are used to select which subsets are generated: the main subset and the final The subset is generated by the inverse scan pattern.

[0048] The inverse quantization unit 230 receives the differential quantization parameter from the entropy decoding unit 210. quantization parameter predictor 15 in FIG. It is generated through a process similar to the operation by 0. Then, the difference quantization parameter and the previous The quantization parameter predictor is added to generate the quantization parameter of the current coding unit. If no differential quantization parameter for the current coding unit is received from the encoder, In this case, the differential quantization parameter is set to 0.

[0049] A parameter indicating a mapping relationship between the quantization parameter and the chrominance quantization parameter. The mapping relationship is changed for each slice. If it is allowed to do so, additional parameters can be included in the slice header. Therefore, the chrominance quantization parameter is calculated by the quantization parameter and the picture parameter set or using the quantization parameter and the two parameters It is generated by

[0050] The inverse quantization unit 230 inverse quantizes the quantized block.

[0051] The inverse transform unit 240 inverse transforms the dequantized block to restore a residual block. The inverse transform type is determined by the prediction mode and the size of the transform unit. The types are DCT-based integer transforms or DST-based integer transforms. For prediction, a DCT-based integer transform is used. For intra prediction, the transform unit If the size is smaller than the predefined size, DST-based integer conversion is used, so Otherwise, a DCT-based integer transform is used. The inverse transform type of the chroma transform unit is The inverse transform type is the same as the corresponding transform unit. Therefore, the inverse of the chrominance transform unit The transform type is a DCT-based integer transform.

[0052] The intra prediction unit 250 predicts the current prediction unit using the received intra prediction information. and restoring an intra-prediction mode and performing a prediction block according to the restored intra-prediction mode. Generate.

[0053] The inter prediction unit 260 predicts the current prediction unit using the received inter prediction information. The motion information is restored and a prediction block is generated using the motion information.

[0054] Post-processing unit 270 operates similarly to post-processing unit 180 in FIG.

[0055] The picture storage unit 280 receives the post-processed video from the post-processing unit 270 and stores the video in a picture unit. The image is stored in a pixel position. A picture can be a frame or a field.

[0056] The adder 290 adds the reconstructed residual block and the prediction block to generate a reconstructed block. do.

[0057] FIG. 3 is a block diagram illustrating an apparatus 300 for generating a prediction block according to the present invention.

[0058] The device 300 according to the present invention includes a parsing unit 310, a prediction mode decoding unit 320, a prediction subunit 330, and a prediction subunit 340. A size determination unit 330, a reference pixel generation unit 340, a reference pixel filtering unit 350, and a prediction block The lock generator 360 is included.

[0059] The parsing unit 310 extracts intra-prediction data of a current prediction unit from a received bitstream. Parse the prediction parameters.

[0060] The luma intra prediction parameters include a mode group indicator and a prediction mode index. The mode group indicator indicates whether the intra prediction mode of the current prediction unit is an MPM group. Indicates whether the mode belongs to the most probable mode group. If the flag is 1, the intra prediction mode of the current prediction unit is MP If the flag is 0, the intra prediction mode of the current prediction unit is , belong to the residual mode group. The mode group is a group of all intra prediction modes other than those belonging to the MPM group. The prediction mode index includes a prediction mode group indicator. Identifies the intra prediction mode of the current prediction unit in the identified group. The tiger prediction parameters are specified by the chrominance prediction mode index.

[0061] The luma intra prediction modes are derived as follows.

[0062] The MPM group is formed using the intra prediction modes of the adjacent prediction units. The intra prediction modes of the MPM group are left intra prediction mode and top intra prediction mode. The left intra-prediction mode is adaptively determined according to the left adjacent prediction mode. The intra prediction mode of the unit, and the upper intra prediction mode is the intra prediction mode of the adjacent upper The MPM group includes three intra prediction modes: It consists of modes.

[0063] If the adjacent prediction unit on the left or upper side does not exist, The intra prediction mode of the prediction unit is set to unavailable. For example, If the unit is located on the left or top boundary of the picture, the adjacent prediction units on the left or top If the adjacent prediction unit on the left or top side belongs to another slice, The intra prediction mode of the adjacent prediction unit to the left or above is set as unavailable. can be.

[0064] FIG. 4 is a conceptual diagram illustrating an intra prediction mode according to the present invention. As shown in FIG. The number of intra prediction modes is 35. DC mode and planar mode are non-directional modes. the remaining are directional modes.

[0065] Both the left intra-prediction mode and the top intra-prediction mode are available, and If different, the left intra prediction mode and the upper intra prediction mode are and one additional intra prediction mode is added to the MPM group. will be done.

[0066] Among the left and top intra prediction modes, one is a non-directional mode. tional mode, and the other is directional mode. mode), the remaining non-directional mode is set as an additional intra prediction mode. If the left and top intra prediction modes are both non-directional modes, the vertical mode is It is set as an additional intra prediction mode.

[0067] When only one of the left intra prediction mode and the top intra prediction mode is available The available intra prediction modes are included in the MPM group, and two additional intra prediction modes are included in the MPM group. A tra prediction mode is added to the MPM group.

[0068] If the available intra-prediction modes are non-directional modes, one remaining non-directional mode The available intra prediction modes are set as additional intra prediction modes. If the prediction mode is a directional mode, two non-directional modes are added as additional intra prediction modes. It is set as follows.

[0069] If neither the left intra prediction mode nor the top intra prediction mode is available, , DC mode, planar mode and perpendicular mode are added to the MPM group.

[0070] If the mode group indicator indicates the MPM group, the prediction mode index The intra prediction mode in the MPM group specified by the The error prediction mode is set.

[0071] If the mode group indicator does not indicate the MPM group, The three intra prediction modes are rearranged in the order of mode numbers. Among the intra prediction modes, the intra prediction mode having the smallest mode number is selected as the first candidate. the intra prediction mode having the intermediate mode number is set to the second candidate, and the intra prediction mode having the intermediate mode number is set to the most The intra prediction mode with the larger mode number is set as the third candidate.

[0072] The prediction mode index is greater than the mode number of the first candidate in the MPM group. If the prediction mode index is greater than or equal to the prediction mode index, the prediction mode index is increased by 1. If the index is greater than or equal to the mode number of the second candidate in the MPM group When the prediction mode index is changed from the previous prediction mode index, the value of the prediction mode index is increased by 1. If the mode number is greater than or equal to the mode number of the third candidate in the MPM group, the prediction mode The index value is incremented by 1. The last prediction mode index is the current prediction unit. The mode number is set to the mode number of the intra prediction mode of the frame.

[0073] The color difference intra prediction mode is an intra prediction mode specified by the color difference prediction mode index. When the chrominance prediction mode index indicates a DM mode, the chrominance prediction mode is set as the DM mode. In this case, the chrominance intra prediction mode is set to be the same as the luma intra prediction mode.

[0074] The prediction size determination unit 330 determines the size of the transform unit based on the transform size information. The transformation size information is determined by one or more split_t The size of the color difference prediction block is also determined based on the transform size information. The minimum size of a chrominance prediction block is 4x4.

[0075] If the size of the transform unit is the same as the size of the current prediction unit, the size of the prediction block is equal to the size of the current prediction unit.

[0076] If the size of the transform unit is smaller than the size of the current prediction unit, the size of the prediction block is The size of the transform unit is the same as the size of the transform unit. In this case, the process of generating the reconstruction block is , is performed for each subblock of the current prediction unit. A prediction block and a residual block are generated, and the prediction block and the residual block are added to the current subblock. A block recovery block is generated. Then, the prediction of the next subblock in the decoding order is A block, a residual block and a reconstructed block are generated. It is used to generate prediction blocks for all subblocks. Some pixels of the original sub-block are used as reference pixels for the next sub-block. It is possible to generate a predicted block that is more similar to the subblock.

[0077] If one or more reference pixels of the current block are unavailable, the reference pixel generator 340: Generate a reference pixel. The reference pixel of the current block is located at (x=0, 2N-1, y=-1). The upper reference pixel is located at (x=-1, y=0, 2M-1), the left reference pixel is located at ( x=-1, y=-1) and the corner reference pixel located at the current block. M is the horizontal length of the current block, and M is the vertical length of the current block. The current chrominance block is a sub-block having a size of a unit or a transform unit. If reference pixels are also not available, they are generated.

[0078] If all reference pixels are not available, all reference pixels are L-1 Replaced by L. The value is the number of bits used to represent the value of a luminance pixel.

[0079] If available reference pixels are present only on one side of unavailable reference pixels, The value of an unavailable reference pixel is the value of the reference pixel closest to the unavailable reference pixel. is replaced by the value.

[0080] If there are available reference pixels on both sides of the unavailable reference pixel, the average value of the closest reference pixels or the closest to the unavailable reference pixel in a predetermined direction The pixel value is replaced by the value of the reference pixel at the new position.

[0081] The reference pixel filtering unit 350 is configured to filter the intra prediction mode and the size of the transform unit. The reference pixels of the current luminance block are adaptively filtered based on the pixel size.

[0082] In DC mode, the reference pixels are not filtered. In directional modes other than the vertical and horizontal modes, the illuminated pixels are not filtered. The reference pixels are adaptively filtered according to the size of the current block.

[0083] If the size of the current block is 4×4, the reference pixels are full-width in all intra prediction modes. Not filtered. For sizes 8x8, 16x16 and 32x32, the reference pixels are filtered. The number of intra prediction modes that must be filtered increases with the current block size. For example, the vertical mode and a predetermined distance adjacent to the vertical mode are In the horizontal and intra prediction modes, the reference pixels are not filtered. A predetermined number of intra prediction modes closest to the horizontal mode also have full pixel references. The predetermined number is between 0 and 7, and is not filtered. The larger the noise, the smaller the

[0084] The reference pixel filtering unit 350 performs the filtering process according to the intra prediction mode and the size of the transform unit. Regardless, the reference pixels of the current chrominance block are not filtered.

[0085] The prediction block generation unit 360 selects reference pixels according to the restored intra prediction mode. It is used to generate a prediction block of the current block.

[0086] In DC mode, N reference pixels are located at (x=0,...,N-1, y=-1) and , M reference pixels located at (x=-1, y=0,...,M-1) are averaged to obtain a predicted pixel. The predicted pixels of the predicted block adjacent to the reference pixels are generated by dividing one or two of the predicted pixels. The chrominance prediction pixels are filtered using the reference pixels adjacent to the pixel. It won't be done.

[0087] In vertical mode, the predicted pixel is generated by copying the value of the vertical reference pixel. The adjacent predicted pixels are filtered using the left adjacent reference pixels and the corner reference pixels. The chrominance prediction pixels are not filtered.

[0088] In horizontal mode, predicted pixels are generated by duplicating the values ​​of horizontal reference pixels. The adjacent predicted pixels are filtered using the upper adjacent reference pixels and the corner reference pixels. The chrominance prediction pixels are not filtered.

[0089] FIG. 5 is a block diagram illustrating an apparatus 400 for generating residual blocks according to the present invention.

[0090] The device 400 according to the present invention includes an entropy decoding unit 410, a reverse scanning unit 420, It includes an inverse quantization unit 430 and an inverse transform unit 440 .

[0091] The entropy decoding unit 410 entropy-decodes the encoded residual signal to obtain a quantity When CABAC is used for entropy coding, The quantized coefficient components include a significance flag, a coefficient sign, and a coefficient level. The significance flag is , indicates whether the corresponding quantized transform coefficient is zero. The coefficient level indicates the absolute value of a non-zero quantized transform coefficient.

[0092] The inverse scanning unit 420 determines an inverse scan pattern and performs a scan on the inverse scan pattern. The inverse scanning unit 420 generates a quantized block. It works the same as 220.

[0093] The inverse quantization unit 430 derives a quantization parameter, selects an inverse quantization matrix, and The quantized block is inverse quantized to generate a transformed block.

[0094] The luma quantization parameter is derived as follows:

[0095] A minimum size of the quantization unit is determined. The minimum size of the quantization unit is determined based on the picture size. It is determined on a picture-by-picture basis using the QU size indicator contained in the parameter set. The size indicator specifies the minimum size of the quantization unit.

[0096] The differential quantization parameter (dQP) of the current coding unit is restored. , generated for each quantization unit through entropy decoding. If the quantization unit does not contain a coded dQP, then dQP is set to 0. If multiple coding units are included, the first coding unit that contains a coefficient with a non-zero dQP is used. It is included in the bitstream of the encoding unit.

[0097] The quantization parameter predictor for the current coding unit is generated. The parameter predictors are generated in a manner similar to the operation of the inverse quantization unit 230 in FIG. If a contains multiple coding units, the first coding unit is A quantization parameter predictor for a set of bits is generated, and the generated quantization parameter predictor is used as a This is used for all quantization units within the quantization unit.

[0098] A quantization parameter is generated using the dQP and the quantization parameter predictor.

[0099] The chrominance quantization parameter is a luminance quantization parameter and the luminance quantization parameter and chrominance The offset parameter (offset parameter) indicates the mapping relationship with the quantization parameter. The offset parameter is generated using the picture parameter. If the offset is allowed to vary from slice to slice, The offset parameter is an offset adjustment parameter (o It can be changed by ffset adjusting parameter.

[0100] The inverse transform unit 440 inversely transforms the transform block to generate a residual block. The type of transformation is determined by the prediction mode and the size of the transform unit. In intra prediction, the luma transform unit is a T-based integer transform or a DST-based integer transform. If the unit size is smaller than 8x8, use DST-based integer conversion, otherwise For chrominance transformation blocks, DCT-based integer transformation is used. Use integer conversion.

[0101] The predicted block and the residual block are added to generate a reconstructed block. The size of the prediction unit is the same as the size of the transform unit. If is larger than the transform unit, the last reconstruction block is generated after the first reconstruction block is generated. Until the block is generated, the next block's prediction block and residual block are The intra prediction mode of the current prediction unit is It is used to generate the lock and residual blocks.

[0102] Although the present invention has been described with reference to the embodiments, those skilled in the art will understand the present invention as described in the claims. Various modifications and variations of the present invention are possible without departing from the spirit and scope of the present invention as described in the scope. It can be seen that the above-mentioned configuration can be changed. [Explanation of symbols]

[0103] 120 Intra prediction unit

Claims

1. 1. An apparatus for generating a residual block, comprising: an entropy decoding module for decoding the residual signal and generating quantized coefficient components; an inverse scanning module for determining an inverse scan pattern and generating a quantization block using the inverse scan pattern; an inverse quantization module for deriving a quantization parameter and for inverse quantizing the quantized block using the quantization parameter to generate a transform block; and an inverse transform module for inverse transforming the transform block to generate a residual block; When the quantization parameter is a chrominance quantization parameter, the chrominance quantization parameter is generated using a luma quantization parameter and an offset parameter indicating a relationship between the luma quantization parameter and the chrominance quantization parameter; The luma quantization parameter is derived using a differential quantization parameter and a quantization parameter predictor, and the differential quantization parameter is generated by entropy decoding; When two or more quantization parameters are available among a left quantization parameter, an upper quantization parameter, and a previous quantization parameter of a current coding unit, the quantization parameter predictor is generated using two available quantization parameters determined according to a predetermined order; An apparatus, wherein if only one quantization parameter is available, the available quantization parameter is set as the quantization parameter predictor.

2. 2. The apparatus of claim 1, wherein when the transform block is a luminance block and the prediction mode is intra prediction, an inverse transform type used to inverse transform the transform block is a DST-based integer transform when the size of the transform block is smaller than 8x8, and is a DCT-based integer transform when the size of the transform block is not smaller than 8x8.

3. The apparatus of claim 1 , wherein if the transform block is a chrominance block, then an inverse transform type utilized to inverse transform the transform block is a DCT-based integer transform.

4. The apparatus of claim 1 , wherein if the left quantization parameter and the top quantization parameter are available, an average value of the left quantization parameter and the top quantization parameter is set as the quantization parameter predictor.

5. The apparatus of claim 1 , wherein the offset parameter indicating a relationship between the luma quantization parameter and the chroma quantization parameter is included in a picture parameter set.

Images