JP4855418B2 - Color video encoding and decoding method and apparatus using correlation of color difference components - Google Patents

Description

  The present invention relates to encoding and decoding of color video data. More specifically, the present invention finds a correlation between color difference components Cb and Cr of color video data configured in the YCbCr format, and reduces color data with a smaller amount of data. The present invention relates to a method and apparatus for encoding video data and decoding the video data.

  FIG. 1 is a diagram showing data constituting an RGB format video and a YCbCr format video.

  There is an RGB format as one of the formats for expressing a color video. The RGB format expresses video color components by dividing them into three color components (Red, Green, Blue, hereinafter referred to as R, G, B). . At this time, the three color components R, G, and B all have the same data amount. For example, if there is a 16 × 16 macroblock, the R component, G component, and B component also have a 16 × 16 size. However, the human eye reacts more sensitively to the luminance component representing brightness than the color difference component representing hue. Therefore, the amount of data is reduced by using a format in which a color video is divided into a luminance component and a color difference component, and this format is called a YCbCr format.

  In the YCbCr format, more data is allocated to the luminance component than the color difference component. Referring to FIG. 1, if an RGB format video for a 16 × 16 macroblock is expressed as a YCbCr format video, it can be seen that it consists of a 16 × 16 size luminance block and 8 × 8 size color difference blocks Cb and Cr. At this time, the Y, Cb, and Cr values that are the luminance component and the color difference component are calculated by a weighted combination of the R, G, and B values. For example, Y, Cb, Cr value can be calculated by mathematical formulas such as Y = 0.259900R + 0.58700G + 0.11400B, Cb = −0.16874R−0.33126G + 0.50000B, Cr = 0.0000R−0.41869G−0.08131. Is calculated. As described above, the color moving image data in the YCbCr format is composed of the luminance component and the two color difference components, and each component is encoded separately at the time of encoding. That is, encoding is performed separately without considering the relationship between color difference components.

  FIG. 2 is a diagram showing the configuration of video data in 4: 4: 4, 4: 2: 2, and 4: 2: 0 formats.

  When a moving image is encoded, the color format of the moving image is indicated by displaying the ratio between the luminance component and the color difference component of the pixels included in the horizontal pixel lines of the picture constituting the moving image. Hereinafter, the luminance component is displayed as Y, and the color difference components are displayed as Cb and Cr. The luminance is a level indicating the brightness of the video, and the luminance of one pixel is represented by 8 bits in the ITU-R recommendation. The color difference is information representing the color of the video, and the pixel color is represented by an 8-bit value of the two color difference components Cb and Cr. A coordinate system representing colors is called a color space. In the Motion Picture Experts Group (MPEG), which is a moving picture coding standard, the color format of a moving picture is expressed by three pieces of 8-bit information of Y, Cb, and Cr.

  In expressing a color moving image using Y, Cb, and Cr, it can be expressed in various color formats depending on the ratio. However, in the case of different color formats, the Y component that is a luminance component is the same, and Cb, Only the CR component changes. Referring to FIG. 2, the 4: 2: 2 format video is obtained by down-sampling the chrominance component of the 4: 4: 4 format video by 1/2 in the horizontal direction. It can be seen that the video in 4: 2: 0 format can be obtained by downsampling 2 times.

  As described above, in the conventional general codec (MPEG, H.26x, VC1), the RGB color video is converted into the YCbCr color video, and the luminance component and the color difference component are separated and encoded. At this time, various formats such as 4: 4: 4, 4: 2: 2, and 4: 2: 0 can be used. In general, in (MPEG, H.26x, VC1) and the like, YCbCr is 4: Video data in 2: 0 format is input and encoded. Hereinafter, video data in 4: 2: 0 format will be described as an example.

  In a general moving image encoding method, input Y, Cb, and CR components are encoded so as not to have temporal and spatial redundancy. Spatial redundancy is removed through intra-prediction, which is prediction with neighboring blocks, and temporal redundancy is removed by inter-prediction between the previous picture and the current picture. That is, only the difference component between the adjacent block and the current block is encoded through intra prediction, and only the difference component between the previous picture and the current picture is encoded through inter prediction, thereby improving the compression efficiency.

  That is, conventionally, only the prediction for removing temporal and spatial redundancy in the Y, Cb, and CR components is performed, and the redundancy is removed using the correlation between the Y, Cb, and CR components. Not. However, H. When compressing a high-quality video such as H.264 high profile, the data amount of Y, Cb, and CR components increases, so a method for efficiently compressing this is necessary.

  The technical problem to be solved by the present invention is to reduce the amount of data to be encoded by using the correlation between Cb and CR components constituting the color difference component of a color video, An object is to provide a color video encoding apparatus and method with an increased encoding speed, and an apparatus and method for decoding the same.

  The technical problem is that, according to the present invention, Cb and Cr values, which are color difference components of a color image, are combined by giving a weight value with a predetermined coefficient to generate a plurality of converted values, and among the generated converted values, A chrominance component conversion unit that selects and outputs two conversion values having the lowest cost calculated by a predetermined cost function, and an entropy encoding unit that performs entropy encoding on the selected conversion value This is achieved by an encoding device characterized in that.

  If the color difference components Cb and Cr are already converted and quantized, or if the color difference components Cb and Cr are values before the conversion and quantization, the color difference component conversion unit It is desirable to further include a conversion and quantization unit that performs conversion and quantization on the conversion value output from the.

  The color difference component conversion unit calculates a conversion value of the color difference component by axCb + bxCr + c, the a, b, and c values are constants, and a plurality of (a, b, c) pairs are predetermined by the user. Preferably, the predetermined cost function includes RDCost, SAD, SATD, SSA, and MAD.

  Further, the color difference component conversion unit, when the color difference components Cb and Cr are inputted, converts the conversion values in all cases that can be generated by giving weight values with a plurality of (a, b, c) coefficients and combining them. A conversion value calculation unit for calculating, a cost calculation unit for calculating a cost for the converted conversion value by a predetermined cost function, and two conversion values having the lowest cost among the calculated cost values are selected. It is desirable to include a determination unit that outputs the data.

  In addition, it is preferable that the color difference component conversion unit encodes information on (a, b, c) coefficients corresponding to the two conversion values with the lowest cost by a run length encoding method.

  Further, the technical problem is that, when the color difference components Cb and Cr are input, the conversion values in all cases that can be generated by combining the weight values with a plurality of (a, b, c) coefficients are calculated. A conversion value calculation unit, a cost calculation unit for calculating a cost for the converted conversion value by a predetermined cost function, and selecting two conversion values having the lowest cost among the calculated cost values It is also achieved by an encoding apparatus comprising: a determination unit for outputting; and an entropy encoding unit for entropy encoding the output transformation value.

  Further, the technical problem is that (a) Cb and Cr values that are color difference components of a color image are combined by giving weight values with predetermined coefficients to generate a plurality of conversion values, and A step of selecting and outputting two transform values having the lowest cost calculated by a predetermined cost function; and (b) an entropy coding step of performing entropy coding on the selected transform value. It is also achieved by an encoding method characterized by including.

  In the step (a), (a1) if the color difference components Cb and Cr are input, conversion values in all cases that can be generated by giving weight values with a plurality of (a, b, c) coefficients and combining them. (A2) a cost calculation step for calculating a cost for the converted conversion value by a predetermined cost function, and (a3) two of the calculated cost values having the lowest cost And selecting and outputting the conversion value.

  In the step (b), information on (a, b, c) coefficients corresponding to the two transformation values having the lowest cost determined in the step (a) may be encoded by a run length encoding method. desirable.

  On the other hand, according to another field of the present invention, the technical problem is that an entropy decoding unit for entropy decoding an encoded bitstream and the decoded data are luminance components, When the color difference component is bypassed, the information on the coefficients used for combining the Cb and CR components with weights is extracted, and the original color difference components Cb and CR components are generated and output. It is also achieved by a decoding device comprising a component inverse transform unit.

  The chrominance component inverse transform unit is information indicating which (a, b, c) coefficient pair is used to encode the chrominance component, and the information encoded and transmitted by the run-length encoding method. It is desirable to extract and calculate Cb and CR components.

  In addition, the technical problem is that (a) an entropy decoding step for entropy decoding the encoded bitstream; and (b) if the decoded data is a luminance component, bypassing the color difference In the case of a component, the method includes extracting information on coefficients used for combining the Cb and CR components by giving weights, and generating and outputting the color difference components Cb and CR components. This is also achieved by the decoding method.

  The technical problem is that, in the color video encoding apparatus according to the present invention, the color difference component of the color video is converted for each of two or more interpretation modes, and the value converted for each mode is converted into a predetermined cost function. A color difference component conversion unit that calculates a cost for each interprediction mode, selects one of the two or more interprediction modes according to the calculation result, and outputs a conversion value of the selected mode; It is also achieved by an encoding apparatus comprising: an entropy encoding unit that performs entropy encoding on the output transform value.

  The selection of the interprediction mode is performed in a predetermined block unit, and the interprediction mode information selected for the predetermined block is preferably encoded in a group unit including a plurality of blocks. .

  It is desirable to generate a plurality of mode planes obtained by classifying interpredation mode information related to a plurality of blocks in a predetermined group for each mode, and to encode the generated plurality of mode planes.

  The generated mode plane may include information indicating whether or not an interpredation mode corresponding to the current mode plane is applied for each predetermined block.

  As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. The corresponding mode information is preferably obtained by setting to '0'.

  Generating a plurality of mode planes obtained by classifying interprediction mode information for a plurality of blocks in a predetermined group according to each mode, and arranging the generated plurality of mode planes in a selected order; It is desirable to modify the information of the next mode plane according to the mode information of the mode plane and encode the modified mode plane.

  The generated mode plane includes information indicating whether or not an interpredation mode corresponding to the current mode plane is applied for each predetermined block, and the modification of the information of the mode plane includes the previous mode plane. It is desirable to delete the information on the block to which the prediction mode of the previous mode plane is applied from the information of the next mode plane.

  As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. The corresponding mode information is preferably obtained by setting to '0'.

  It is preferable that the information regarding the block to which the previous mode is applied is deleted by setting the information about the block to which the previous mode is applied to '0'. .

  It is preferable that the predetermined block is a macro block and the group is a picture.

  The color difference component conversion unit includes an interprediction mode table storage unit that stores an inter prediction mode table including two or more inter prediction modes, and Cb and Cr values that are color difference components of a color image. Based on a mode table, a conversion value calculation unit that calculates a conversion value for each mode, and a mode that selects an interpretation mode with the lowest cost calculated by a predetermined cost function among the calculated conversion values It is desirable to include a selection unit.

  It is preferable to further include a run-length encoding unit that performs run-length encoding on the selected interprediction mode information.

  The technical problem is that, in the color video encoding method according to the present invention, the color difference component of the color video is converted for each of two or more interpretation modes, and the value converted for each mode is converted into a predetermined cost function. And calculating a cost for each mode using the calculation result, selecting one of the two or more interpredation modes according to the calculation result, and outputting a conversion value of the selected mode; and the output conversion Performing an entropy encoding on the value.

  Further, the technical problem is that in an encoded color video decoding device, an entropy decoding unit for entropy decoding an input bitstream, and a current block of a predetermined size extracted from the bitstream. A chrominance component inverse transform unit that restores the original chrominance component based on the applied inter-prediction mode information is provided, and the inter-prediction mode information is applied to the current block among two or more inter-prediction modes. The original chrominance component is obtained from a transform value corresponding to the interprediction mode applied to the current block.

  The technical problem is applied to an encoded bitstream decoding method for entropy decoding an input bitstream and a current block of a predetermined size extracted from the bitstream. Reconstructing an original color difference component based on the interprediction mode information, wherein the interprediction mode information includes an interprediction mode applied to the current block among two or more interpredation modes. The original chrominance component is obtained from a transform value corresponding to the interpretation mode applied to the current block.

  Further, the technical problem is that the color difference component of the color image is converted for each of two or more interpretation modes, the value converted for each mode is calculated for each mode using a predetermined cost function, Selecting one of the two or more interpredation modes according to a calculation result, outputting a conversion value of the selected mode, and performing entropy encoding on the output conversion value The present invention is also achieved by a computer-readable recording medium on which a program for executing a color video encoding method is recorded.

  Further, the technical problem is that the input color stream is subjected to entropy decoding, and the original color difference is extracted based on the prediction mode information extracted from the bit stream and applied to the current block of a predetermined size. Restoring the component, wherein the interprediction mode information indicates an interpredation mode applied to the current block among two or more interpredation modes, and the original color difference component is the current block And a computer-readable recording medium having recorded thereon a program for executing a method for decoding a coded color image, which is obtained from a conversion value corresponding to an interpretation mode applied to Achieved.

  According to the present invention, it is possible to increase the compression rate of moving images and greatly reduce the number of bits required for encoding. In addition, more efficient run-length encoding is possible.

  Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 3 is a block diagram of a moving picture coding apparatus according to an embodiment of the present invention.

  The encoding apparatus includes a motion estimation unit 302, a motion compensation unit 304, an intra-prediction execution unit 306, a conversion unit 308, a quantization unit 310, a rearrangement unit 312, an entropy encoding unit 314, an inverse quantization unit 316, and an inverse transform. A unit 318, a filter 320, and a frame memory 322.

  The encoding apparatus performs encoding on a macroblock of the current picture under one encoding mode selected from various encoding modes. For this purpose, encoding is performed under all possible modes of interpredation and intraprediction to calculate a rate-distortion cost (RDCost), and the mode having the lowest value is optimized. A mode is determined, and encoding is performed under that mode. Rate (R) means a bit rate and indicates the number of bits used to encode one macroblock. That is, R is the number of bits obtained by encoding a residual signal (residual) after interpredation or intraprediction, and the number of bits obtained by encoding motion vector information. It is the value which added all. Distortion (D) is the difference between the original macroblock and the decoded macroblock when the encoded video is decoded. Therefore, D is a value that can be understood if decoding is performed and the original macroblock is restored.

  However, according to the coding mode determination method of the present invention, the optimum mode can be determined not only by calculating the rate-distortion cost but also by various methods. That is, not only RDCost but also cost calculation is performed by various methods. For example, the cost functions that can be used are SAD (Sum of Absolute Difference value), SATD (Sum of Absolute Transformed Difference), SSD (Sum of Sequential Difference), MD (Sum of Aquired Difference), MAD There is.

  The motion estimation unit 302 searches the reference picture for the prediction value of the macroblock of the current picture for the purpose of interpredation. Then, when the reference block is found in units of 1/2 pixel or 1/4 pixel, the motion compensation unit 304 calculates these intermediate pixel values to determine the reference block data value. As described above, the interprediction is performed by the motion estimation unit 302 and the motion compensation unit 304.

  Also, intra prediction execution unit 306 performs intra prediction for searching for a predicted value of a macroblock of the current picture in the current picture. Whether the current macroblock is to be interpreted or intrapredicted is calculated by calculating the rate-distortion cost under all coding modes and selecting the mode with the lowest value as the block. The coding mode is determined and the coding is performed on the macroblock.

  As described above, if the prediction data referred to by the macroblock of the current frame is found after the interpredation or intraprediction is performed, the prediction data is subtracted from the macroblock of the current picture, and is then sent to the color difference component conversion unit 330. input. When the chrominance component is input, the chrominance component conversion unit 330 generates the chrominance component into various conversion values by a chrominance component conversion method described later, and selects two conversion values among them. If a luminance component that is not a color difference component is input, it passes through as it is. The luminance component or the selected color difference component is input to the conversion unit 308 and converted, and then quantized by the quantization unit 310. A difference obtained by subtracting a motion-estimated reference block from a macroblock of the current frame is referred to as a residual, but data input to the color difference component conversion unit 330 in order to reduce the amount of data at the time of encoding is a residual value. The quantized residual value passes through the realignment unit 312 for encoding by the entropy encoding unit 314.

  On the other hand, to obtain a reference picture used for interpredation, the quantized picture is restored through the inverse quantization unit 316 and the inverse transform unit 318 to restore the current picture. The current picture restored in this way is stored in the frame memory and is used for interpredation on the next picture. If the restored picture passes through the filter 320, the original picture becomes a picture including some coding errors.

  The detailed operation of the color difference component conversion unit 330 will be described as follows. If Cb and Cr data as luminance components are input, a converted value is calculated by the following equation (1).

Conversion value = axCb + bxCr + c (1)
Here, the a, b, and c values are determined by experiments. For example, in the case of interpredation, (a, b, c) is changed to (1, 0, 0), (0, 1, 0), (-1, 1, 0), (1, 1, 0). For example, the conversion values are Cb, Cr, -Cb + Cr, and Cb + Cr. Costs are calculated for these four. The types of cost calculation and cost function used are as described above. Among the calculated costs, two (a, b, c) sets having the lowest value are selected and input to the conversion unit 308. For example, if Cb and -Cb + Cr are selected, the conversion unit 308 converts Cb and -Cb + Cr components. In this case, since the cost is the lowest, the values of the Cb and -Cb + Cr components are the lowest, thereby reducing the amount of bits required for encoding. On the other hand, in the case of a macroblock in intra prediction, (a, b, c) is (-1, 1, 14), (1, 1, -250), (1, 0, 14), (0, 1,14). In the case of a macroblock in intra prediction, the cost is calculated for the (a, b, c) coefficient pair as in the case of the macroblock in interprediction. The color difference component determined by a, b, c) is searched and encoded.

  On the other hand, the color difference component conversion unit 330 may be positioned after the conversion unit 308 and the quantization unit 310. That is, realignment and entropy coding can be performed by calculating the cost using the frequency-converted Cb and Cr components on the frequency domain that is not the spatial domain.

  FIG. 4 is a reference diagram for explaining calculation of a color difference component conversion value according to an embodiment of the present invention. Referring to FIG. 4, one pixel value is read for each of the Cb block and the Cr block, and the converted value is calculated by multiplying or adding the (a, b, c) value according to the above-described equation (1). I understand.

  FIG. 5 is a detailed block diagram of the color difference component conversion unit 330 of FIG.

  The color difference component conversion unit 330 includes a conversion value calculation unit 510, a cost calculation unit 520, and a determination unit 530. When the color difference components Cb and Cr are input, the conversion value calculation unit 510 calculates the conversion values in all cases that can be generated using the given (a, b, c) coefficient set according to Equation (1). To do. The cost calculation unit 520 calculates a cost for each calculated case. The determination unit 530 selects the two cases where the calculated cost value is the lowest and outputs the converted value in that case.

  FIG. 6 is a flowchart of an encoding method according to an embodiment of the present invention.

  If video data is input, motion estimation (S610) and motion prediction (S620) are performed in the case of interpredation. In the case of intra prediction, the motion estimation step (S610) and the motion prediction step (S620) are omitted. The execution of motion estimation and motion prediction is as described with reference to FIG. After motion prediction is performed, the cost is calculated as described with reference to FIGS. 4 and 5 for all cases obtained using the predetermined (a, b, c) coefficients (S630). ). After calculating the cost, two conversion values having the lowest value are selected (S640). A transformation step (S650), a quantization step (S660), and entropy coding are performed on the selected case (S670). As a result, the chrominance components Cb and Cr were conventionally encoded, but the bits necessary for encoding by encoding the converted chrominance component and removing the redundancy between the chrominance components Cb and Cr in steps S630 and S640. Reduce the number.

  On the other hand, the selected (a, b, c) coefficient information is also encoded and transmitted. In order to indicate which color difference component is encoded and transmitted for each macroblock, the selected coefficient information is recorded in the picture header for each macroblock. In the example of the above-mentioned interpredation, the first and third of the (1, 0, 0), (0, 1, 0), (-1, 1, 0), (1, 1, 0) coefficients If a coefficient is selected, this information is run-length encoded.

  More specifically, the selected coefficient information is run-length encoded only when the chrominance component block is encoded. In this case, the chrominance encoding block pattern of the conventional syntax or CBPC (Coded Block) is used. (Pattern for Chroma). When encoding in (run, length) format, the number of bits allocated to run depends on how many sets are used, and the number of bits allocated to length depends on how many consecutive run are coded together. change. For example, if the number of sets is S1, S2, S3, S4 and a total of four types, when A is 2 bits and B is 5 bits, one (run, length) is encoded into 7 bits. In other words, if S1 is continuously output as No. 11, (S1, 10) can be encoded to “0001010”. Since the (a, b, c) coefficient information selected for each macroblock is likely to have a value similar to the coefficient information of the adjacent macroblock, it is necessary for encoding if the run-length encoding method is used. The number of bits can be reduced. Further, information indicating whether or not the color difference block is encoded in units of blocks is transmitted using the color difference encoded block pattern or CBPC.

  Since the (a, b, c) coefficient information selected for each macroblock is likely to have a value similar to the coefficient information of the adjacent macroblock, it is necessary for encoding if the run-length encoding method is used. The number of bits can be reduced.

  FIG. 7 is a block diagram of a decoding apparatus according to an embodiment of the present invention.

  The decoding apparatus includes an entropy decoding unit 702, a rearrangement unit 704, an inverse quantization unit 706, an inverse transform unit 708, a color difference component inverse transform unit 710, a motion compensation unit 712, an intra prediction execution unit 714, a filter 716, and A frame memory 718 is provided. When the encoded bit stream is input to the decoding apparatus, the encoded bit stream is input to the color difference component inverse transform unit 708 through entropy decoding, rearrangement, and inverse transform. If the input data is a luminance component, it is bypassed. If it is a chrominance component, the chrominance component inverse transform unit 708 encodes the chrominance component with any (a, b, c) coefficient. The color difference components Cb and Cr are generated by finding out whether the color difference has been done. Since the information indicating which (a, b, c) coefficient is used to encode and transmit the color difference component is also encoded and transmitted by the run-length encoding method, the color difference component inverse transform unit 710 transmits this information. To generate Cb and Cr components. On the other hand, the chrominance component inverse transform unit 710 may be positioned before the inverse quantization unit 706 and the inverse transform unit 708, as in the case of the encoding apparatus in FIG.

  FIG. 8 is a flowchart of a decoding method according to an embodiment of the present invention.

  After the entropy decoding step (S810), the inverse quantization step (S820), and the inverse transformation step (S830), the received (a, b, c) coefficient information is decoded, and any combination of Cb and Cr components is determined. It is searched whether it has been encoded, and this is inversely converted to obtain Cb and Cr components (S840). Then, decoding is performed through motion compensation (S850). In the case of intra prediction, the motion compensation step (S850) is omitted.

  Hereinafter, an encoding method according to another embodiment of the present invention will be described with reference to FIGS. 9 to 15.

  FIG. 9 is a table showing an interpretation mode applied to an embodiment of the present invention. Referring to FIG. 9, the inter-prediction mode applied to each macro block is set to five types of 0 to 4, and as shown in FIG. Interpredation mode is selected. According to the selected mode and the table shown in FIG. 9, the Cb and Cr blocks are replaced with the conversion value 1 and the conversion value 2. The relationship between the conversion value and the Cb, Cr value in each of the interpredation modes shown in FIG. 9 is an example, and a mode having another relationship is selectively added or removed. It is possible.

  For example, if the selected interpredation mode for a given macroblock is 0, the conversion value 1 is the Cb value of the Cb block itself, and the conversion value 2 is the Cr value of the Cr block itself. . When the selected interpretation mode is 1, the conversion value 1 is the Cb value itself of the Cb block, and the conversion value 2 is the converted Cb value Cb ′ to the Cr value of the Cr block. Subtracted value. The reason for using the Cb and Cr values restored in the Interpredation mode table of FIG. 9 is to restore more accurate Cb and Cr values at the decoding stage. Optionally, the original Cb and Cr values can be used instead of the recovered Cb and Cr values. Here, the reconstructed Cb and Cr values mean values obtained by performing inverse quantization and inverse transformation after performing the transformation and quantization process, and the original Cb and Cr values are transformed and transformed. The value before the quantization process is performed.

  FIG. 10 is a table for explaining the deinterpretation method for each mode. Referring to FIG. 10, when the interpredation mode of one macroblock is 0, the Cb and Cr values of the corresponding macroblock are immediately obtained from the conversion value 1 and the conversion value 2. On the other hand, when the interpretation mode is 1, the Cb value of the corresponding macroblock is obtained immediately from the converted value 1, but the Cr value is converted from the value (Cb ′) obtained by restoring the converted value 1 to the converted value 2 (Cb It is obtained from the value obtained by subtracting '-Cr). At this time, the reason why the conversion value 2 is subtracted from the value obtained by restoring the conversion value 1 is that the conversion value 2 includes Cb ′.

  FIG. 11 is a detailed block diagram of the color difference component conversion unit 330 of FIG. 3 according to another embodiment.

  The color difference component conversion unit 330 includes a conversion value calculation unit 1110, an interpretation mode table storage unit 1112, a cost calculation unit 1120, a mode selection / conversion value output unit 1130, and a selection mode storage unit 1132.

  If the color difference components Cb and Cr are input, the conversion value calculation unit 1110 calculates the conversion value 1 and the conversion value 2 for each of the interpretation modes stored in the interpretation mode table storage unit 1112. For example, the conversion value 1 and the conversion value 2 are generated for each mode shown in FIG. In the present embodiment, the interprediction mode table is stored in the interprediction mode table 1112 that exists separately from the conversion value calculation unit 1110. However, the interprediction mode table is selectively used as the conversion value calculation unit 1110. It may be stored in a predetermined place.

  The cost calculation unit 1120 calculates a cost for the conversion value calculated for each mode.

  The mode selection / conversion value output unit 1130 selects the mode having the lowest calculated cost value, and outputs the conversion value at that time. For example, when mode 1 is selected according to the interpretation mode table shown in FIG. 9, Cb and Cb′-Cr values are output as conversion value 1 and conversion value 2.

  The selection mode storage unit 1132 stores mode information for each macroblock selected by the mode selection and conversion value output unit 1130. The mode information for each macroblock stored in the selection mode storage unit 1132 is used to generate a picture unit interprediction mode table shown in FIG. In this embodiment, the mode information selected in units of macroblocks is stored in the selection mode storage unit 1132, but information on the selectively selected mode is stored in the mode selection and conversion value output unit 1132. It may be stored in a predetermined place.

  The conversion value output from the mode selection / conversion value output unit 1130 is output to the conversion unit 308 and the quantization unit 310 as illustrated in FIG. 3, and the conversion and quantization process is performed.

  As described above, since the transform value is determined according to the selected mode and encoding is performed based on the determined transform value, the decoding unit determines which interpredation mode is selected for each macroblock. It is necessary to inform. Hereinafter, a method for transmitting the interprediction mode information selected for each macroblock will be described with reference to FIGS. 12 to 14.

  FIG. 12 shows the interpredation mode selected for each macroblock in one picture. The values at the respective positions have values of 0, 1, 2, 3, and 4, which indicate the interpredation mode applied to the macroblock corresponding to the corresponding position. For example, the value “0” at the uppermost leftmost position substitutes Cb and Cr for the corresponding macroblock by the interpretation mode 0 in Table 9, that is, conversion value 1 = Cb and conversion value 2 = Cr. Indicates that there is. Further, the value '22' corresponding to the next macroblock at the uppermost end indicates that the macroblock corresponding to these positions substitutes Cb and Cr by the interprediction mode 2 in FIG. 9, that is, the conversion value 1 = Cb, The conversion value 2 = Cb ′ + Cr.

  FIGS. 13A to 13E are diagrams illustrating the inter-prediction mode values for each macroblock shown in FIG. 12 for each inter-prediction mode plane.

  FIG. 13A is an inter-prediction mode table shown in FIG. 12, and is reconfigured to indicate 1 for a macroblock indicating mode 0 and 0 for the remaining macroblocks not indicating mode 0. Mode 0 plane. For example, the first, fourth, fifth, seventh, eighth, tenth, and fourteenth values having 0 value among the interpredation mode values shown in FIG. The value is set to 0.

  FIG. 13B is an inter-prediction mode table shown in FIG. 12, and is reconfigured to indicate 1 for a macroblock indicating mode 1 and 0 for the remaining macroblocks not indicating mode 0. Mode 1 plane. For example, the 6th and 9th values having 1 value are set to 1 and the remaining values are set to 0 among the top end values of the interprediction mode table shown in FIG.

  FIG. 13C is an inter-prediction mode table shown in FIG. 12, and is reconfigured to indicate 1 for a macroblock indicating mode 2 and 0 for the remaining macroblocks not indicating mode 0. Mode 2 plane. For example, the 2nd, 3, 13, 15, 17, 18, 19, 20, 21, and 22th values having two values among the top end values of the interprediction mode table shown in FIG. And the remaining values are set to zero.

  FIG. 13D is an inter-prediction mode table shown in FIG. 12, and is reconfigured to indicate 1 for a macroblock indicating mode 3 and 0 for the remaining macroblocks not indicating mode 0. Mode 3 plane. For example, the interprediction mode values shown in FIG. 12 are all set to 0 because there is no ternary value at the uppermost value.

  FIG. 13E is an inter-prediction mode table shown in FIG. 12, and is reconfigured to indicate 1 for a macroblock indicating mode 4 and 0 for the remaining macroblocks not indicating mode 0. Mode 4 plane. For example, the 11th, 12th, and 16th values having 4 values are set to 1 and the remaining values are set to 0 among the top end values of the interprediction mode table shown in FIG.

  As described above, when the interprediction mode table shown in FIG. 12 is divided into the respective mode planes as shown in FIGS. 13A to 13E, the length of 0 run (length of 0 run) becomes even longer.

  14A to 14D are diagrams for explaining a method of encoding interprediction mode information according to the present invention. 14A to 14D show modified mode planes in which the length of one run in each mode plane shown in FIGS. 13A to 13E is further increased by employing the mode plane removing method according to the present invention.

  FIG. 14A shows a modified mode 1 plane in which the “0” value corresponding to the macroblock whose value is 1 in the mode 0 plane of FIG. 13A is deleted from the mode 1 plane of FIG. 13B. As shown in FIG. 14A, the uppermost 22 bits '0000010100000000000000' of FIG. 13B indicating the original mode 1 plane are 1, 4, 5, 7, 8, 10, 14th '0' value is removed and transformed to 15 bits '001100000000000000'. The modified mode 1 plane has a reduced amount of bits and a longer run than the original mode 1 plane.

  FIG. 14B shows a modified mode 2 plane in which the “0” value corresponding to the macroblock whose value is 1 in the mode 0 plane and the mode 1 plane is deleted from the mode 2 plane of FIG. 13C. As shown in FIG. 14B, the uppermost 22 bits '0110000000001010111111' of FIG. 13C showing the original mode 2 plane are 1, 4, 5, 7, 8, 10, The 14th '0' value and the value in the mode 1 plane are 1, and the 6th and 9th '0' values are removed and transformed to 13 bits '1100110111111'. The modified mode 2 plane has a reduced amount of bits and a longer run than the original mode 2 plane.

  FIG. 14C is a modified mode 3 plane in which the “0” value corresponding to the macroblock whose value is 1 in the mode 0 plane, the mode 1 plane, and the mode 2 plane is deleted from the mode 3 plane of FIG. 13D. Indicates. As shown in FIG. 14C, the uppermost 22 bits '0000000000000000000000' of FIG. 13D showing the original mode 3 plane are 1, 4, 5, 7, 8, 10, The 14th '0' value, the value in the mode 1 plane is 1, the 6th and 9th '0' values, and the value in the mode 2 plane is 1, 2, 3, 13, 15, 17, It is transformed into 3-bit '000' with the '0' value in 18, 19, 20, 21, 22 removed. The modified mode 3 plane has a reduced bit amount compared to the original mode 3 plane.

  FIG. 14D is modified by deleting the “0” value corresponding to the macroblock whose value is 1 in the mode 0 plane, the mode 1 plane, the mode 2 plane, and the mode 3 plane from the mode 4 plane of FIG. 13E. Mode 4 plane is shown. As shown in FIG. 14D, the uppermost 22 bits '0000000000000110001000' of FIG. 13E showing the original mode 4 plane are 1, 4, 5, 7, 8, 10, The 14th '0' value, the value in the mode 1 plane is 1, the 6th and 9th '0' values, and the value in the mode 2 plane is 1, 2, 3, 13, 15, 17, The '0' values at 18, 19, 20, 21, and 22 are transformed into the removed 3-bit '111'. The modified mode 4 plane has a reduced bit amount and a longer run than the original mode 3 plane. Moreover, since all the modified mode 4 planes have a single value, encoding is not necessary.

  In the present embodiment, it is possible to reduce the amount of data transmitted by performing run-length encoding on the original mode 0 plane and the modified mode 1-4 plane having a longer '1' run. is there. Alternatively, transmission may be performed by selectively performing run-length encoding on the original mode plane.

  The decoding unit sequentially restores the mode 4 plane from the mode 1 plane of FIGS. 14A to 14D in order to generate the original mode plane of FIGS. 13B to 13E. Further, based on the restored mode plane of FIGS. 13A to 13E, the interpretation mode table of FIG. 12 is restored, and based on this, the original Cb and Cr values are restored.

  FIG. 15 is a flowchart for explaining a modified mode plane generation method and an interpretation mode information encoding method shown in FIGS. 14A to 14D according to an embodiment of the present invention.

  In step 1510, run-length encoding is performed on the mode 0 plane.

  In step 1520, after run-length encoding is performed on the mode 0 plane, the mode 1, 2, 3, and 4 planes shown in FIGS. 13B to 13E correspond to macroblocks having a value of 1 in the mode 0 plane. The '0' value is removed to generate a first-order modified mode plane. Thereafter, run-length encoding is performed on the mode 1 plane that has been linearly modified, that is, the modified mode 1 plane shown in FIG. 14A.

  In step 1530, after run-length coding is performed on the modified mode 1 plane, the first modified mode 2, 3, and 4 planes correspond to macroblocks having a value of 1 in the mode 1 plane. The '0' value is removed to generate a second-order modified mode plane. Then, run-length encoding is performed on the second-order modified mode 2 plane, that is, the modified mode 2 plane shown in FIG. 14B.

  In step 1540, after run-length encoding is performed on the modified mode 2 plane, '0' corresponding to a macroblock having a value of 1 in the mode 2 plane is obtained from the second modified mode 3 and 4 planes. 'Remove the value and generate a cubically transformed mode plane. Then, run-length encoding is performed on the third-order modified mode 3 plane, that is, the modified mode 3 plane shown in FIG. 14C.

  In step 1550, after run-length encoding is performed on the modified mode 3 plane, a value having a value of 1 in the mode 3 plane is removed from the third modified mode 4 plane, and the fourth modified. Mode 4 plane is generated. Then, run-length encoding is performed on the fourth-order modified mode 4 plane, that is, the modified mode 3 plane shown in FIG. 14D. Optionally, the values in the last mode plane all have a value of '1', so even if there is no information in the modified mode 4 plane, only the information in the remaining modified mode is used for the original mode plane. Therefore, it is not necessary to separately compress and encode the modified mode 4 plane. Alternatively, skipping may be skipped without performing step 1550 selectively.

  Optionally, modified mode planes may be generated in a different order than this embodiment.

  After performing steps 1510 to 1550, the run-length encoded interpredation information is inserted into the picture header of the bitstream and transmitted.

  In the present embodiment, description has been given by taking, as an example, the color difference component of a color image, that is, the interpredation using the correlation between Cb and Cr. However, optionally, the present invention can be applied between any two regions of any color space to increase compression efficiency. For example, the present invention is not applicable to Cb and Cr, but also to interpredation between regions on other color spaces, that is, using the correlation between Y and Cb or Y and Cr in the YCbCr color space. It is.

  Hereinafter, a decoding apparatus according to still another embodiment of the present invention will be described with reference to FIG.

  When the encoded bit stream is input to the decoding apparatus, the encoded bit stream is input to the color difference component inverse transform unit 708 through entropy decoding, rearrangement, and inverse transform. If the input data is a luminance component, it is bypassed, and if it is a chrominance component, it is input to the chrominance component inverse transform unit 710.

  The inter-prediction mode determination unit (not shown) restores the run-length encoded mode plane extracted from the picture header of the input bitstream and restores sequentially from the mode plane 0, then FIG. The inter-prediction mode table for each picture shown in FIG. 6 is generated, and based on this, the inter-prediction mode applied for each macro-block unit is determined and input to the color difference component inverse transform unit 710.

  The color difference component inverse conversion unit 710 generates Cb and Cr components from the decoded converted values using the determined interpredation mode information.

  FIG. 16 is a flowchart of a decoding method according to an embodiment of the present invention.

  After the entropy decoding step (S1610), the inverse quantization step (S1620), and the inverse transformation step (S1630), the original mode plane is restored from the decoded transformation mode plane shown in FIG. For example, an inter-prediction mode table indicating the inter-prediction mode applied to each macro block is generated in units of pictures.

  From the generated inter-prediction mode table, an inter-prediction mode applied to the corresponding macroblock is determined, and a conversion value decoded by the determined inter-prediction mode is inversely converted to obtain Cb and Cr components. Calculate (S1640). Then, decoding is performed through motion compensation (S1650). In the case of intra prediction, the motion compensation step (S1650) is omitted.

  As described above, according to the present invention, there is an effect of increasing the compression ratio of a moving image by searching for a correlation between color difference components of the moving image and removing unnecessary components. In addition, in order to search for a correlation and remove unnecessary components, the coefficient information constituting the combination of Cb and Cr components is encoded by the run-length encoding method, thereby greatly reducing the number of bits required for encoding. Further, by converting the Cb and Cr components for each interpredation mode, dividing the information indicating the mode applied for the conversion into the respective mode planes, and performing the run length coding, it is further efficient. Run-length encoding is possible.

  On the other hand, the encoding and decoding methods described above can be created by a computer program. The code and code segments that make up the program can be easily inferred by computer programmers in the field. In addition, the program is stored in a computer-readable information recording medium, and is read and executed by the computer, thereby realizing an entropy encoding and decoding method. The information recording medium includes a magnetic recording medium, an optical recording medium, and a carrier wave medium.

  In the above, this invention was demonstrated centering on the desirable embodiment. Those skilled in the art will appreciate that the present invention may be embodied in variations that do not depart from the essential characteristics of the invention. Accordingly, the disclosed embodiments should be considered in an illustrative, not a limiting sense. The scope of the present invention is shown not in the above description but in the claims, and all differences within the scope equivalent thereto should be construed as being included in the present invention.

  The present invention is applicable to a method of searching for a correlation between color difference components Cb and Cr of color video data configured in the YCbCr format, encoding color video data with a smaller amount of data, and decoding this. .

It is a figure which shows the data which comprise the image | video of a RGB format, and the image | video of a YCbCr format. It is a figure which shows the structure of the video data of 4: 4: 4, 4: 2: 2, and 4: 2: 0 format. 1 is a block diagram of a moving image encoding apparatus according to an embodiment of the present invention. It is a reference figure for demonstrating calculation of the color difference component conversion value by one Embodiment of this invention. FIG. 4 is a detailed block diagram of a color difference component conversion unit 330 in FIG. 3. 4 is a flowchart of an encoding method according to an embodiment of the present invention. FIG. 4 is a block diagram of a decoding device according to an embodiment of the present invention. 4 is a flowchart of a decoding method according to an embodiment of the present invention. It is a table which shows the interpretation mode applied to one Embodiment of this invention. It is a table for demonstrating the deinterpretation method with respect to each interpredation mode. It is a detailed block diagram according to still another embodiment of the color difference component conversion unit 330 of FIG. It is a block diagram which shows the interpredation mode selected with respect to each macroblock in one picture. It is a figure which shows the interpretation mode plane by this invention. It is a figure which shows the interpretation mode plane by this invention. It is a figure which shows the interpretation mode plane by this invention. It is a figure which shows the interpretation mode plane by this invention. It is a figure which shows the interpretation mode plane by this invention. It is a figure for demonstrating the encoding method of the interference prediction mode information by this invention. It is a figure for demonstrating the encoding method of the interference prediction mode information by this invention. It is a figure for demonstrating the encoding method of the interference prediction mode information by this invention. It is a figure for demonstrating the encoding method of the interference prediction mode information by this invention. 5 is a flowchart illustrating a method for encoding interprediction mode information according to an exemplary embodiment of the present invention. 5 is a flowchart illustrating a decoding method according to an embodiment of the present invention.

Claims (101)

Cb and Cr values, which are color difference components of a color image, are generated by combining weighted values with predetermined coefficients to generate a plurality of converted values, and among the generated converted values, a cost calculated by a predetermined cost function A color difference component converter that selects two conversion values having the lowest value and outputs the two selected conversion values;
An entropy encoding unit that performs entropy encoding on the two transform values.   The encoding apparatus according to claim 1, wherein the color difference components Cb and Cr have already been converted and quantized. A conversion unit that converts the conversion value output from the color difference component conversion unit;
The encoding apparatus according to claim 1, further comprising: a quantization unit that performs quantization on the conversion value output from the conversion unit.   The encoding apparatus according to claim 1, further comprising a motion estimation and motion compensation unit for performing an interpredation.   The color difference component conversion unit calculates a conversion value of the color difference component by axCb + bxCr + c, the a, b, c values are predetermined coefficients, and a plurality of (a, b, c) pairs are predetermined by the user. The encoding apparatus according to claim 1, wherein:   The encoding apparatus according to claim 1, wherein the predetermined cost function includes RDCost, SAD, SATD, SSA, and MAD. The color difference component converter is
If the color difference components Cb and Cr are input, a conversion value calculation unit that calculates conversion values in all cases that can be generated by giving weights with a plurality of (a, b, c) coefficients and combining them;
A cost calculation unit for calculating a cost with a predetermined cost function for the converted conversion value;
The encoding apparatus according to claim 1, further comprising: a determination unit that selects and outputs two conversion values having the lowest cost among the calculated cost values. The color difference component converter is
The encoding apparatus according to claim 1, wherein information relating to (a, b, c) coefficients corresponding to two transformed values having the lowest cost is encoded by a run-length encoding method. A step of generating Cb and Cr values, which are color difference components of a color image, by giving weight values with predetermined coefficients and combining them into a plurality of converted values;
Selecting two transformation values with the lowest cost calculated by a given cost function;
Performing an entropy encoding on the two selected transform values.   10. The encoding method according to claim 9, wherein the color difference components Cb and Cr have already been converted and quantized.   The code according to claim 9, further comprising a step of performing conversion and quantization on the converted value if the color difference components Cb and Cr are values before conversion and quantization. Method.   The encoding method according to claim 9, further comprising performing motion estimation and motion compensation for performing interpredation.   As the conversion value, a conversion value of a color difference component is calculated by axCb + bxCr + c, the a, b, and c values are predetermined coefficients, and a plurality of (a, b, c) pairs are predetermined by the user. The encoding method according to claim 9.   The encoding method according to claim 9, wherein the predetermined cost function includes RDCost, SAD, SATD, SSD, and MAD. The step of generating the plurality of conversion values includes:
Multiplying the color difference components Cb, Cr by a plurality of (a, b, c) coefficients, and combining the multiplied results with the coefficient c;
The encoding method according to claim 9, further comprising: combining a coefficient c and a result obtained by multiplying coefficients a and b among the plurality of (a, b, c) coefficients.   10. The encoding method according to claim 9, wherein information on (a, b, c) coefficients corresponding to the two transformation values having the lowest cost is encoded by a run-length encoding method. An entropy decoding unit for entropy decoding the encoded bitstream and outputting the decoded data;
If the decoded data is a luminance component, it is bypassed, and if it is a chrominance component, it extracts information on the coefficients used to combine the Cb and Cr components with weights, A decoding device comprising: a color difference component inverse transform unit that generates and outputs original color difference components Cb and Cr components. An inverse quantization unit that performs inverse quantization on the entropy decoded data;
The decoding apparatus according to claim 17, further comprising: an inverse transform unit that performs inverse transform on the inversely quantized data.   The decoding apparatus according to claim 17, further comprising a motion compensation unit for performing interpredation. The color difference component inverse transform unit includes:
Information indicating which (a, b, c) coefficient pair is used to encode the color difference component, and extracting the information encoded and transmitted by the run-length encoding method to obtain the Cb and Cr components The decoding device according to claim 17, wherein: An entropy decoding stage for entropy decoding the encoded bitstream;
If the decoded data is a luminance component, it is bypassed, and if it is a chrominance component, it extracts information on the coefficients used to combine the Cb and Cr components with weights, Generating and outputting color difference components Cb and Cr components, and a decoding method comprising:   The decoding method according to claim 21, further comprising: performing inverse quantization and inverse transform on the entropy-decoded data.   The decoding method according to claim 21, further comprising a motion compensation step for performing an interpredation.   Information indicating which (a, b, c) coefficient pair was used to encode the color difference component, and using the information encoded and transmitted by the run-length encoding method, Cb and Cr components The decoding method according to claim 21, wherein: In a color video encoding device,
The color difference component of the color image is converted for each of the two or more interpredation modes, and the value converted for each mode is calculated for each interpredation mode using a predetermined cost function. A color difference component converter that selects one of two or more interpredation modes and outputs a conversion value of the selected mode;
An entropy encoding unit that performs entropy encoding on the output transformed value.   The selection of the inter-prediction mode is performed in units of predetermined blocks, and the inter-prediction mode information selected for the predetermined block is encoded in units of groups composed of a plurality of blocks. The encoding device according to claim 25.   A plurality of mode planes are generated by classifying interprediction mode information regarding a plurality of blocks in a predetermined group for each mode, and the generated plurality of mode planes are encoded. Item 27. The encoding device according to Item 26.   The encoding apparatus according to claim 27, wherein the generated mode plane includes information indicating whether or not an interpretation mode corresponding to a current mode plane is applied to each predetermined block.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. 28. The encoding apparatus according to claim 27, wherein the corresponding mode information is obtained by setting to '0'.   Generating a plurality of mode planes obtained by classifying interprediction mode information for a plurality of blocks in a predetermined group according to each mode, and arranging the generated plurality of mode planes in a selected order; 27. The encoding apparatus according to claim 26, wherein the information of the next mode plane is modified according to the mode information of the mode plane, and the modified mode plane is encoded.   The generated mode plane includes information indicating whether or not an interpredation mode corresponding to the current mode plane is applied for each predetermined block, and the modification of the information of the mode plane includes the previous mode plane. 31. The encoding apparatus according to claim 30, wherein the encoding is performed by deleting information on a block to which an interprediction mode of a previous mode plane is applied from information of a next mode plane based on the information of .   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. 32. The encoding apparatus according to claim 31, wherein the corresponding mode information is obtained by setting to '0'.   The deletion of the information regarding the block to which the previous mode is applied is performed by setting the information about the block to which the previous mode is applied to '0'. The encoding device according to claim 32.   27. The encoding apparatus according to claim 26, wherein the predetermined block is a macro block, and the group is a picture.   The encoding apparatus according to claim 25, wherein the color difference components are Cb and Cr, and the color difference components Cb and Cr have already been converted and quantized. A conversion unit that converts the conversion value output from the color difference component conversion unit;
The encoding apparatus according to claim 25, further comprising: a quantization unit that performs quantization on the conversion value output from the conversion unit.   The encoding apparatus according to claim 25, further comprising a motion estimation and motion compensation unit for performing interpredation.   The encoding apparatus according to claim 25, wherein the predetermined cost function includes RDCost, SAD, SATD, SSA, and MAD. The color difference component converter is
An inter-prediction mode table storage unit for storing an inter-prediction mode table including two or more inter-prediction modes;
A conversion value calculation unit that calculates Cb and Cr values, which are color difference components of a color image, for each mode based on the Interpredation mode table;
The encoding apparatus according to claim 25, further comprising: a mode selection unit that selects an interpretation mode having a lowest cost calculated by a predetermined cost function among the calculated conversion values.   The encoding apparatus according to claim 25, further comprising a run-length encoding unit that performs run-length encoding on the selected interprediction mode information. In the color video encoding method,
The color difference component of the color image is converted for each of two or more interpretation modes, the value converted for each mode is calculated for each mode using a predetermined cost function, and the two or more are calculated according to the calculation result. Selecting one of the Interpredation modes and outputting a conversion value of the selected mode;
Performing an entropy encoding on the output transformed value.   The selection of the inter-prediction mode is performed in units of predetermined blocks, and the inter-prediction mode information selected for the predetermined block is encoded in units of groups composed of a plurality of blocks. The encoding method according to claim 41.   A plurality of mode planes are generated by classifying interprediction mode information regarding a plurality of blocks in a predetermined group for each mode, and the generated plurality of mode planes are encoded. Item 43. The encoding method according to Item 42.   44. The encoding method of claim 43, wherein the generated mode plane includes information indicating whether or not an interpretation mode corresponding to a current mode plane is applied to each predetermined block.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. 44. The encoding method according to claim 43, wherein the corresponding mode information is obtained by setting to '0'.   Generating a plurality of mode planes obtained by classifying interprediction mode information for a plurality of blocks in a predetermined group according to each mode, and arranging the generated plurality of mode planes in a selected order; 43. The encoding method according to claim 42, wherein information on a next mode plane is modified according to mode mode information, and the modified mode plane is encoded.   The generated mode plane includes information indicating whether or not an interpredation mode corresponding to the current mode plane is applied for each predetermined block, and the modification of the information of the mode plane includes the previous mode plane. 47. The encoding method according to claim 46, wherein the encoding method is performed by deleting information on a block to which an interprediction mode of a previous mode plane is applied from information of a next mode plane based on the information of the next mode plane. .   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. 48. The encoding method according to claim 47, wherein the corresponding mode information is obtained by setting to '0'.   The deletion of the information regarding the block to which the previous mode is applied is performed by setting the information about the block to which the previous mode is applied to '0'. The encoding method according to claim 48.   43. The encoding method according to claim 42, wherein the predetermined block is a macroblock, and the group is a picture.   The encoding method according to claim 41, wherein the color difference components are Cb and Cr, and the color difference components Cb and Cr have already been transformed and quantized.   If the color difference components Cb and Cr are values before conversion and quantization, the method further includes a step of performing conversion and quantization on the conversion value output from the color difference component conversion unit. The encoding method according to claim 41.   42. The encoding method according to claim 41, further comprising performing interpredation.   The encoding method according to claim 41, wherein the predetermined cost function includes RDCost, SAD, SATD, SSA, and MAD. The encoding method is:
Calculating conversion values for Cb and Cr values, which are color difference components of a color image, for each mode;
The step of selecting one of the two or more inter prediction modes includes the step of selecting an interpredation mode having a lowest cost calculated by a predetermined cost function among the calculated conversion values. The encoding method according to claim 41, further comprising:   The encoding method of claim 41, further comprising performing run length encoding on the selected interprediction mode information. In a decoding apparatus for encoded color video,
An entropy decoding unit for entropy decoding the input bitstream;
A color difference component inverse transform unit that restores the original color difference component based on the interpredation mode information extracted from the bitstream and applied to a current block of a predetermined size,
The inter-prediction mode information indicates an inter-prediction mode applied to the current block among two or more inter-prediction modes, and the original color difference component is changed to an inter-prediction mode applied to the current block. A decoding device characterized in that it is obtained from a corresponding conversion value.   58. The inter prediction mode information extracted from the bitstream is a plurality of mode planes generated by classifying the inter prediction mode information for a plurality of blocks according to each mode. The decoding device according to 1.   59. The decoding apparatus according to claim 58, wherein the generated mode plane includes information indicating whether or not an interpretation mode corresponding to a current mode plane is applied to each predetermined block.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. 59. The decoding apparatus according to claim 58, wherein the corresponding mode information is obtained by setting to '0'.   The interprediction mode information extracted from the bitstream includes a plurality of mode planes generated by classifying interpredation mode information related to a plurality of blocks in a predetermined group for each mode in the selected order. 59. The decoding apparatus according to claim 58, wherein the decoding apparatus is a modified mode plane that is generated by modifying the information of the next mode plane according to the mode information of the previous mode plane. The generated mode plane includes information indicating whether or not an interpredation mode corresponding to the current mode plane is applied for each predetermined block, and the modification of the information of the mode plane includes the previous mode plane. Based on the information of the following mode plane, by deleting the information about the block to which the previous mode plane interprediction mode was applied from the information of the next mode plane,
The decoding apparatus according to claim 61, wherein the decoding apparatus restores the original mode plane by restoring the modified mode planes in the order of selection.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. 62. The decoding apparatus according to claim 61, wherein the corresponding mode information is obtained by setting to '0'.   The deletion of the information regarding the block to which the previous mode is applied is performed by setting the information about the block to which the previous mode is applied to '0'. 64. The decoding device according to claim 63.   59. The decoding apparatus according to claim 58, wherein the block of the predetermined size is a macroblock, and the mode plane includes interpredation mode information regarding a macroblock in units of pictures.   58. The decoding apparatus according to claim 57, wherein the color difference components are Cb and Cr. An inverse quantization unit that performs inverse quantization on the entropy decoded data;
58. The decoding apparatus according to claim 57, further comprising: an inverse transform unit that performs inverse transform on the data subjected to the inverse quantization.   The decoding apparatus according to claim 57, further comprising a motion compensation unit for performing interpredation.   59. The decoding according to claim 58, wherein the decoding apparatus extracts interpredation mode information encoded and transmitted by a run-length encoding method, and calculates a color difference component based on the extracted information. Device. In a method for decoding an encoded color video,
Entropy decoding the input bitstream;
Restoring the original color difference component based on the interpretation mode information extracted from the bitstream and applied to a current block of a predetermined size,
The inter-prediction mode information indicates an inter-prediction mode applied to the current block among two or more inter-prediction modes, and the original color difference component is changed to an inter-prediction mode applied to the current block. A decoding method characterized in that it is obtained from the corresponding conversion value.   The interprediction mode information extracted from the bitstream is a plurality of mode planes generated by classifying interpredation mode information related to a plurality of blocks according to each mode. Decoding method described in 1.   The decoding method according to claim 71, wherein the generated mode plane includes information indicating whether or not an interpretation mode corresponding to a current mode plane is applied to each predetermined block.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. 72. The decoding method according to claim 71, wherein the corresponding mode information is obtained by setting to '0'.   The interprediction mode information extracted from the bitstream includes a plurality of mode planes generated by classifying interpredation mode information related to a plurality of blocks in a predetermined group for each mode in the selected order. 72. The decoding method according to claim 71, wherein the decoding method is a modified mode plane that is generated by modifying the information of the next mode plane according to the mode information of the previous mode plane. The generated mode plane includes information indicating whether or not an interpredation mode corresponding to the current mode plane is applied for each predetermined block, and the modification of the information of the mode plane includes the previous mode plane. Based on the information of the following mode plane, by deleting the information about the block to which the previous mode plane interprediction mode was applied from the information of the next mode plane,
The decoding method according to claim 74, wherein the decoding apparatus restores the original mode plane by restoring the modified mode plane in the selected order.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. The decoding method according to claim 75, wherein the corresponding mode information is obtained by setting to '0'.   The deletion of the information regarding the block to which the previous mode is applied is performed by setting the information about the block to which the previous mode is applied to '0'. The decoding method according to claim 76.   The decoding method according to claim 71, wherein the block of the predetermined size is a macroblock, and the mode plane includes interpredation mode information regarding a macroblock in units of pictures.   The decoding method according to claim 70, wherein the color difference components are Cb and Cr.   The decoding method according to claim 70, further comprising: performing inverse quantization and inverse transform on the entropy decoded data.   The decoding method according to claim 70, further comprising a motion compensation step for performing interpredation.   The decoding according to claim 71, wherein the decoding method extracts the interpredation mode information encoded and transmitted by the run-length encoding method and calculates a color difference component based on the extracted information. Method. A step of generating Cb and Cr values, which are color difference components of a color image, by giving weight values with predetermined coefficients and combining them into a plurality of converted values;
Selecting two transformation values with the lowest cost calculated by a given cost function;
A computer-readable recording medium having recorded thereon a program for performing an encoding method, comprising: entropy encoding with respect to the two selected transformation values. Converting the color difference component of the color image into two or more interpretation modes;
Calculating a cost for each mode using a predetermined cost function;
Selecting one of the two or more interpredation modes according to the calculation result;
A computer-readable recording medium having recorded thereon a program for performing a color image encoding method, comprising: entropy encoding the output conversion value of the selected interprediction mode value.   The selection of the inter-prediction mode is performed in units of predetermined blocks, and the inter-prediction mode information selected for the predetermined block is encoded in units of groups composed of a plurality of blocks. The recording medium according to claim 84.   A plurality of mode planes are generated by classifying interprediction mode information regarding a plurality of blocks in a predetermined group for each mode, and the generated plurality of mode planes are encoded. Item 86. The recording medium according to Item 85.   The recording medium according to claim 86, wherein the generated mode plane includes information indicating whether or not an interpretation mode corresponding to a current mode plane is applied to each predetermined block.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. The recording mode according to claim 86, wherein the corresponding mode information is obtained by setting to '0'.   Generating a plurality of mode planes obtained by classifying interprediction mode information for a plurality of blocks in a predetermined group according to each mode, and arranging the generated plurality of mode planes in a selected order; 86. The recording medium according to claim 85, wherein the information of the next mode plane is modified according to the mode information of the mode plane, and the modified mode plane is encoded.   The generated mode plane includes information indicating whether or not an interpredation mode corresponding to the current mode plane is applied for each predetermined block, and the modification of the information of the mode plane includes the previous mode plane. 90. The recording medium according to claim 89, wherein the recording medium is deleted by deleting information on a block to which the previous mode plane interprediction mode is applied from the information of the next mode plane based on the information of the next mode plane.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. The recording medium according to claim 90, wherein the corresponding mode information is obtained by setting to '0'.   The deletion of the information regarding the block to which the previous mode is applied is performed by setting the information about the block to which the previous mode is applied to '0'. The recording medium according to claim 91. An entropy decoding stage for entropy decoding the encoded bitstream to generate decoded data;
If the decoded data is a luminance component, bypassing;
If the decoded data is a chrominance component, information relating to coefficients used for combining the Cb and Cr components by applying weights is extracted to generate and output the chrominance components Cb and Cr components. And a computer-readable recording medium storing a program for performing a decoding method. Entropy decoding the input bitstream;
Restoring the original color difference component based on the interpretation mode information extracted from the bitstream and applied to a current block of a predetermined size,
The inter-prediction mode information indicates an inter-prediction mode applied to the current block among two or more inter-prediction modes, and the original color difference component is changed to an inter-prediction mode applied to the current block. A computer-readable recording medium storing a program for performing a method of decoding an encoded color video obtained from a corresponding conversion value.   95. The inter-prediction mode information extracted from the bitstream is a plurality of mode planes generated by classifying the inter-prediction mode information related to a plurality of blocks for each mode. The recording medium described in 1.   The recording medium according to claim 95, wherein the generated mode plane includes information indicating whether or not an interpretation mode corresponding to a current mode plane is applied to each predetermined block.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. 96. The recording medium according to claim 95, wherein the corresponding mode information is obtained by setting to '0'.   The interprediction mode information extracted from the bitstream includes a plurality of mode planes generated by classifying interpredation mode information related to a plurality of blocks in a predetermined group for each mode in the selected order. 96. The recording medium according to claim 95, wherein the recording medium is a modified mode plane that is generated by modifying the information of the next mode plane according to the mode information of the previous mode plane. The generated mode plane includes information indicating whether or not an interpredation mode corresponding to the current mode plane is applied for each predetermined block, and the modification of the information of the mode plane includes the previous mode plane. Based on the information of the following mode plane, by deleting the information about the block to which the previous mode plane interprediction mode was applied from the information of the next mode plane,
99. The recording medium according to claim 98, wherein the decoding device restores the original mode plane by restoring the modified mode plane in the selected order.   As for the predetermined mode plane, the mode information corresponding to the block to which the prediction mode corresponding to the current mode plane is applied is set to '1', and the block to which the interpolation prediction mode corresponding to the current mode plane is not applied. The recording medium according to claim 99, wherein the corresponding mode information is obtained by setting to '0'.   The deletion of the information regarding the block to which the previous mode is applied is performed by setting the information about the block to which the previous mode is applied to '0'. The recording medium according to claim 100.

Images