JP6262380B2 - Electronic device, decoding method and program - Google Patents

Description

  Embodiments described herein relate generally to an electronic device, a decoding method, and a program.

  In recent years, an image coding method with greatly improved coding efficiency has been developed in collaboration with ITU-T (International Telecommunication Union Telecommunication Standardization Sector) and ISO (International Organization for Standardization) / IEC (International Electrotechnical Commission). T REC. H. H.264 and ISO / IEC 14496-10 (hereinafter referred to as “H.264”).

  H. H.264 discloses an inter-prediction coding scheme that eliminates temporal redundancy and realizes high coding efficiency by performing fractional accuracy motion compensated prediction using a coded image as a reference image. ing.

  Also, a scheme for encoding moving images including fade and dissolve effects with higher efficiency than the inter prediction encoding schemes in ISO / IEC MPEG (Moving Picture Experts Group) -1, 2, 4 has been proposed. In this method, as a framework for predicting a change in brightness in the time direction, fractional accuracy motion compensation prediction is performed on an input moving image having luminance and two color differences. Then, the prediction image is multiplied by the weighting factor using the reference image, the luminance and the weighting factor for each of the two color differences, and the offset for each of the luminance and the two color differences, and the offset is added. The weighting coefficient can represent fractional values with a predetermined accuracy by a parameter indicating fixed-point accuracy, and it is possible to perform weighted motion compensation prediction with finer accuracy for pixel value changes between images. Yes.

JP 2004-7377 A

  In the conventional technology as described above, the reference image, the weighting factor, the offset, and the like are encoded as an index. However, since the index is specified to be expressed with a predetermined bit precision, the weighting factor cannot be expressed. There is. The problem to be solved by the present invention is to provide an electronic device, a decoding method, and a program capable of improving the coding efficiency while allowing the weighting coefficient to be expressed with a predetermined bit precision.

The electronic device according to the embodiment is an electronic device to which encoded data is input, and includes a processing unit. The processing unit decodes the first fixed-point precision of the luminance weight coefficient from the encoded data, and is the same as the value obtained by left-shifting “1” by the number of bits determined by the first fixed-point precision. A value is determined as a first reference value of the luminance weighting factor, and is the same value as the difference between the luminance weighting factor and the first reference value from the encoded data, and the first fixed-point precision Regardless of whether the first difference value of the luminance weighting coefficient, which is a value that can take a fixed range from −128 to +127, is decoded, and the first reference value and the first difference value are added, The luminance weighting factor, which is a value that can take a range from (the first reference value−128) to (the first reference value + 127), is determined, and the median value is determined from the median value of the maximum pixel values. Is multiplied by a color difference weighting factor and said A second reference value of color difference offset is calculated by subtracting a value obtained by right shifting by the number of bits determined by the second fixed point precision of the difference weighting coefficient, and the color difference is calculated from the encoded data. of the same value as the difference between the second reference value and the offset, - a value can range from (2 ⁿ⁾ to (+2 n ^-1), the second difference value of the offset of the color difference Decoding and adding the second reference value and the second difference value, a value that can take a range from (the second reference value -2 ⁿ ) to (the second reference value +2 ⁿ -1) The color difference offset is determined.

The block diagram which shows the example of the encoding apparatus of 1st Embodiment. Explanatory drawing which shows the example of the prediction encoding order of the pixel block of 1st Embodiment. Explanatory drawing which shows the other example of the prediction encoding order of the pixel block of 1st Embodiment. The figure which shows the block size example of the coding tree block in 1st Embodiment. The figure which shows the specific example of the coding tree block of 1st Embodiment. The figure which shows the specific example of the coding tree block of 1st Embodiment. The figure which shows the specific example of the coding tree block of 1st Embodiment. The block diagram which shows the example of the estimated image generation part of 1st Embodiment. The figure which shows the example of the relationship of the motion vector of the motion compensation prediction in the bidirectional | two-way prediction of 1st Embodiment. The block diagram which shows the example of the multi-frame motion compensation part of 1st Embodiment. The reference figure for description of a weighting coefficient. H. FIG. 6 is a reference diagram illustrating a selection range of H.264 weight coefficients. Explanatory drawing which shows the example of the selection range of the weighting coefficient of 1st Embodiment. Explanatory drawing which shows the specific example of the selection range of the weighting coefficient of 1st Embodiment. Explanatory drawing which shows the specific example of the selection range of the weighting coefficient of 1st Embodiment. H. The reference figure which shows the minimum value and maximum value of a H.264 weighting coefficient. Explanatory drawing which shows the example of the minimum value of the weighting coefficient of 1st Embodiment, and a maximum value. The figure which shows the WP parameter information example of 1st Embodiment. The figure which shows the WP parameter information example of 1st Embodiment. 6 is a flowchart illustrating an example of a process for deriving a selection range of weighting factors according to the first embodiment. The figure which shows the example of the syntax of 1st Embodiment. The figure which shows the example of the picture parameter set syntax of 1st Embodiment. The figure which shows the example of the slice header syntax of 1st Embodiment. The figure which shows the example of the tread weight table syntax of 1st Embodiment. Explanatory drawing which shows the example of the relationship of the value of the syntax element of 1st Embodiment. The block diagram which shows the structural example of the decoding apparatus of 2nd Embodiment. Explanatory drawing which shows the example of the selection range of the offset of the modification 1. FIG. 10 is a flowchart illustrating an example of a process for deriving an offset selection range according to Modification 1; Explanatory drawing which shows the example of the selection range of the weighting coefficient of the modification 2. 10 is a flowchart showing an example of a derivation process of a weight coefficient selection range according to Modification 2; Explanatory drawing which shows the example of the range of the difference value of the weighting coefficient of the encoding object of the modification 3. Explanatory drawing which shows the example of the relationship of the value of the syntax element of the modification 3. Explanatory drawing which shows the example of the range of the difference value of the weighting coefficient of the modification 4. Explanatory drawing which shows the example of the selection range of the weighting coefficient after the decoding of the modification 4. 10 is a flowchart illustrating an example of wrap processing of a difference value of weighting factors according to a fifth modification. 10 is a flowchart illustrating an example of restoration processing of a difference value of weighting factors according to a fifth modification.

  Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. The encoding device and decoding device of each of the following embodiments can be realized by hardware such as an LSI (Large-Scale Integration) chip, a DSP (Digital Signal Processor), or an FPGA (Field Programmable Gate Array). In addition, the encoding device and the decoding device of each of the following embodiments can be realized by software by causing a computer to execute a program. In the following description, the term “image” can be appropriately replaced with a term such as “video”, “pixel”, “image signal”, “picture”, or “image data”.

(First embodiment)
In the first embodiment, an encoding apparatus that encodes a moving image will be described.

  FIG. 1 is a block diagram illustrating an example of a configuration of the encoding device 100 according to the first embodiment.

  The encoding apparatus 100 divides each frame or each field constituting the input image into a plurality of pixel blocks, and uses the encoding parameters input from the encoding control unit 111 to predict the encoded code for the divided pixel blocks. To generate a predicted image. The encoding apparatus 100 subtracts the input image divided into a plurality of pixel blocks and the prediction image to generate a prediction error, orthogonally transforms and quantizes the generated prediction error, and further performs entropy encoding to perform encoding. Generate and output data.

  The encoding apparatus 100 performs predictive encoding by selectively applying a plurality of prediction modes in which at least one of the block size of the pixel block and the prediction image generation method is different. The generation method of the prediction image is roughly classified into two types: intra prediction that performs prediction within the encoding target frame and inter prediction that performs motion compensation prediction using one or more reference frames that are temporally different. Note that intra prediction is also referred to as intra prediction or intra frame prediction, and inter prediction is also referred to as inter prediction, inter frame prediction, motion compensation prediction, or the like.

  FIG. 2A is an explanatory diagram illustrating an example of a predictive coding order of pixel blocks in the first embodiment. In the example illustrated in FIG. 2A, the encoding device 100 performs predictive encoding from the upper left to the lower right of the pixel block, and the left side of the encoding target pixel block c in the encoding processing target frame f and The encoded pixel block p is located on the upper side.

  FIG. 2B is an explanatory diagram illustrating another example of the predictive coding order of pixel blocks in the first embodiment. In the example illustrated in FIG. 2B, the encoding device 100 divides the screen into a plurality of tiles or slices in advance, and predictively encodes the pixel blocks in each tile or each slice from the upper left to the lower right. In the encoding target frame f, the encoded pixel block p is located on the left side and the upper side of the encoding target pixel block c. Note that a tile means an area cut out from an arbitrary rectangular area, and a slice means an area cut out by an arbitrary number of large coding tree blocks to be described later according to the predictive coding order.

  As shown in FIG. 2B, after the screen is divided into a plurality of tiles or a plurality of slices, an encoding process is performed for each tile or each slice, thereby enabling a decoding process for each tile or each slice. Become. For this reason, it is possible to distribute the amount of calculation required for decoding by performing decoding processing of high-resolution video in parallel. That is, in the example shown in FIG. 2B, the encoding process and the decoding process can be speeded up.

  In the following, for simplification of description, the encoding apparatus 100 performs predictive encoding in the order shown in FIG. 2A, but the order of predictive encoding is not limited to this.

  The pixel block represents a unit for processing an image, and corresponds to, for example, an M × N size block (M and N are natural numbers), a coding tree block, a macro block, a sub block, or one pixel. In the following description, the pixel block is basically used in the meaning of the coding tree block, but may be used in other meanings. For example, in the description of the prediction unit, a pixel block is used to mean a pixel block of the prediction unit. A block may be called by a name such as a unit. For example, a coding block is called a coding unit.

  FIG. 3A is a diagram illustrating an example of a block size of a coding tree block according to the first embodiment. The coding tree block is typically a 64 × 64 pixel block as shown in FIG. 3A. However, the present invention is not limited to this, and it may be a 32 × 32 pixel block, a 16 × 16 pixel block, an 8 × 8 pixel block, a 4 × 4 pixel block, or the like. Further, the coding tree block may not be a square, and may be, for example, an M × N size (M ≠ N) pixel block.

  3B to 3D are diagrams illustrating specific examples of the coding tree block according to the first embodiment. FIG. 3B shows a coding tree block having a block size of 64 × 64 (N = 32). N represents the size of the reference coding tree block. The size when divided is defined as N, and the size when not divided is defined as 2N. FIG. 3C shows a coding tree block obtained by dividing the coding tree block of FIG. 3B into quadtrees. The coding tree block has a quadtree structure as shown in FIG. 3C. When the coding tree block is divided, the four pixel blocks after the division are numbered in the Z-scan order as shown in FIG. 3C.

  The coding tree block can be further divided into quadtrees within one quadtree number. Thereby, a coding tree block can be divided | segmented hierarchically. In this case, the depth of division is defined by Depth. FIG. 3D shows one of the coding tree blocks obtained by dividing the coding tree block of FIG. 3B into quadtrees, and the block size is 32 × 32 (N = 16). The Depth of the coding tree block shown in FIG. 3B is 0, and the Depth of the coding tree block shown in FIG. The coding tree block having the largest unit is called a large coding tree block, and the input image signal is encoded in this unit in the raster scan order.

  In the following description, the encoding target block or coding tree block of the input image may be referred to as a prediction target block or a prediction pixel block. The encoding unit is not limited to a pixel block, and at least one of a frame, a field, a slice, a line, and a pixel can be used.

  As illustrated in FIG. 1, the encoding device 100 includes a subtraction unit 101, an orthogonal transformation unit 102, a quantization unit 103, an inverse quantization unit 104, an inverse orthogonal transformation unit 105, an addition unit 106, and a prediction The image generation unit 107, the index setting unit 108, the motion evaluation unit 109, and the encoding unit 110 are provided. Note that the encoding control unit 111 illustrated in FIG. 1 controls the encoding apparatus 100 and can be realized by, for example, a CPU (Central Processing Unit).

  The subtraction unit 101 subtracts the corresponding prediction image from the input image divided into pixel blocks to obtain a prediction error. The subtraction unit 101 outputs a prediction error and inputs it to the orthogonal transform unit 102.

  The orthogonal transform unit 102 performs orthogonal transform such as discrete cosine transform (DCT) or discrete sine transform (DST) on the prediction error input from the subtraction unit 101 to obtain transform coefficients. The orthogonal transform unit 102 outputs transform coefficients and inputs them to the quantization unit 103.

  The quantization unit 103 performs a quantization process on the transform coefficient input from the orthogonal transform unit 102 to obtain a quantized transform coefficient. Specifically, the quantization unit 103 performs quantization according to quantization information such as a quantization parameter or a quantization matrix specified by the encoding control unit 111. More specifically, the quantization unit 103 divides the transform coefficient by the quantization step size derived from the quantization information to obtain a quantized transform coefficient. The quantization parameter indicates the fineness of quantization. The quantization matrix is used for weighting the fineness of quantization for each component of the transform coefficient. The quantization unit 103 outputs the quantized transform coefficient and inputs it to the inverse quantization unit 104 and the encoding unit 110.

  The inverse quantization unit 104 performs inverse quantization processing on the quantized transform coefficient input from the quantization unit 103 to obtain a restored transform coefficient. Specifically, the inverse quantization unit 104 performs inverse quantization according to the quantization information used in the quantization unit 103. More specifically, the inverse quantization unit 104 multiplies the quantization transform coefficient by the quantization step size derived from the quantization information to obtain a restored transform coefficient. Note that the quantization information used in the quantization unit 103 is loaded from an internal memory (not shown) of the encoding control unit 111 and used. The inverse quantization unit 104 outputs the restored transform coefficient and inputs it to the inverse orthogonal transform unit 105.

  The inverse orthogonal transform unit 105 performs inverse orthogonal transform such as inverse discrete cosine transform (IDCT) or inverse discrete sine transform (IDST) on the restored transform coefficient input from the inverse quantization unit 104, Get the restoration prediction error. Note that the inverse orthogonal transform performed by the inverse orthogonal transform unit 105 corresponds to the orthogonal transform performed by the orthogonal transform unit 102. The inverse orthogonal transform unit 105 outputs the reconstruction prediction error and inputs it to the addition unit 106.

  The adding unit 106 adds the restored prediction error input from the inverse orthogonal transform unit 105 and the corresponding predicted image to generate a locally decoded image. The adding unit 106 outputs the local decoded image and inputs it to the predicted image generation unit 107.

  The predicted image generation unit 107 stores the locally decoded image input from the addition unit 106 as a reference image in a memory (not shown in FIG. 1), outputs the reference image stored in the memory, and inputs the reference image to the motion evaluation unit 109. . Further, the predicted image generation unit 107 performs weighted motion compensation prediction based on the motion information and WP parameter information input from the motion evaluation unit 109, and generates a predicted image. The predicted image generation unit 107 outputs a predicted image and inputs it to the subtracting unit 101 and the adding unit 106.

  FIG. 4 is a block diagram illustrating an example of the configuration of the predicted image generation unit 107 of the first embodiment. As shown in FIG. 4, the predicted image generation unit 107 includes a multi-frame motion compensation unit 201, a memory 202, a unidirectional motion compensation unit 203, a prediction parameter control unit 204, a reference image selector 205, and a frame memory 206. And a reference image control unit 207.

  The frame memory 206 stores the locally decoded image input from the addition unit 106 as a reference image under the control of the reference image control unit 207. The frame memory 206 has a plurality of memory sets FM1 to FMN (N ≧ 1) for temporarily storing reference images.

  The prediction parameter control unit 204 prepares a plurality of combinations of reference image numbers and prediction parameters as a table based on the motion information input from the motion evaluation unit 109. Here, the motion information indicates a motion vector indicating a shift amount of motion used in motion compensation prediction, a reference image number, information on a prediction mode such as unidirectional / bidirectional prediction, and the like. The prediction parameter refers to information regarding a motion vector and a prediction mode. Then, the prediction parameter control unit 204 selects and outputs a combination of the reference image number and the prediction parameter used for generating the prediction image based on the input image, inputs the reference image number to the reference image selector 205, and outputs the prediction parameter. Is input to the unidirectional motion compensation unit 203.

  The reference image selector 205 is a switch for switching which output terminal of the frame memories FM1 to FMN included in the frame memory 206 is connected according to the reference image number input from the prediction parameter control unit 204. For example, if the reference image number is 0, the reference image selector 205 connects the output end of FM1 to the output end of the reference image selector 205. If the reference image number is N-1, the reference image selector 205 refers to the output end of the FMN. Connected to the output terminal of the image selector 205. The reference image selector 205 outputs the reference image stored in the frame memory to which the output terminal is connected among the frame memories FM1 to FMN included in the frame memory 206, and the unidirectional motion compensation unit 203 and the motion evaluation unit 109. To enter.

  The unidirectional motion compensation unit 203 performs a motion compensation prediction process according to the prediction parameter input from the prediction parameter control unit 204 and the reference image input from the reference image selector 205, and generates a unidirectional prediction image.

  FIG. 5 is a diagram illustrating an example of a motion vector relationship of motion compensation prediction in bidirectional prediction according to the first embodiment. In motion compensated prediction, interpolation processing is performed using a reference image, and a unidirectional prediction image is generated based on the amount of motion deviation from the pixel block at the encoding target position between the generated interpolation image and the input image. . Here, the amount of deviation is a motion vector. As shown in FIG. 5, in a bi-directional prediction slice (B-slice), a prediction image is generated using a set of two types of reference images and motion vectors. As the interpolation process, an interpolation process with 1/2 pixel accuracy, an interpolation process with 1/4 pixel accuracy, or the like is used, and the value of the interpolation image is generated by performing the filtering process on the reference image. For example, H.P. capable of performing interpolation processing up to 1/4 pixel accuracy on a luminance signal. In H.264, the shift amount is expressed by four times the integer pixel accuracy.

  The unidirectional motion compensation unit 203 outputs a unidirectional prediction image and temporarily stores it in the memory 202. Here, when the motion information (prediction parameter) indicates bidirectional prediction, the multi-frame motion compensation unit 201 performs weighted prediction using two types of unidirectional prediction images. The first unidirectional prediction image corresponding to the first is stored in the memory 202, and the second unidirectional prediction image is directly output to the multi-frame motion compensation unit 201. Here, the first unidirectional prediction image corresponding to the first is the first prediction image, and the second unidirectional prediction image is the second prediction image.

  Two unidirectional motion compensation units 203 may be prepared, and each may generate two unidirectional prediction images. In this case, when the motion information (prediction parameter) indicates unidirectional prediction, the unidirectional motion compensation unit 203 directly outputs the first unidirectional prediction image as the first prediction image to the multi-frame motion compensation unit 201. Good.

  The multi-frame motion compensation unit 201 uses the first prediction image input from the memory 202, the second prediction image input from the unidirectional motion compensation unit 203, and the WP parameter information input from the motion evaluation unit 109. A prediction image is generated by performing weighted prediction. The multi-frame motion compensation unit 201 outputs a prediction image and inputs the prediction image to the subtraction unit 101 and the addition unit 106.

  FIG. 6 is a block diagram illustrating an example of the configuration of the multi-frame motion compensation unit 201 according to the first embodiment. As shown in FIG. 6, the multi-frame motion compensation unit 201 includes a default motion compensation unit 301, a weighted motion compensation unit 302, a WP parameter control unit 303, and WP selectors 304 and 305.

  The WP parameter control unit 303 outputs a WP application flag and weight information based on the WP parameter information input from the motion evaluation unit 109, inputs the WP application flag to the WP selectors 304 and 305, and weights the weight information. Input to the motion compensation unit 302. Specifically, when the WP parameter information is input from the motion evaluation unit 109, the WP parameter control unit 303 outputs the WP parameter information by separating it into a first WP application flag, a second WP application flag, and weight information. Then, the first WP application flag is input to the WP selector 304, the second WP application flag is input to the WP selector 305, and the weight information is input to the weighted motion compensation unit 302.

  Here, the WP parameter information includes a first WP application flag (specifically, flag information of the first WP application flag), a second WP adaptation flag (specifically, flag information of the second WP application flag), and a weight. Contains information information. The first WP application flag and the second WP application flag are parameters that can be set for each corresponding reference image and signal component, and whether to perform the default motion compensation prediction for the prediction of the first prediction image and the second prediction image, respectively. Information on whether to perform weighted motion compensation prediction is included. Here, the first WP application flag and the second WP application flag both indicate that default motion compensation prediction is performed when the flag is 1, and that weighted motion compensation prediction is performed when the flag is 1. .

  The weight information includes a first weighting factor value w0C, a second weighting factor value w1C, a parameter LWD indicating the fixed-point accuracy of the first weighting factor and the second weighting factor (hereinafter referred to as “fixed-point accuracy LWD”). Information) of the first offset o0C and the second offset o1C. The variable C means a signal component. For example, in the case of a YUV spatial signal, the luminance signal is C = Y, the Cr color difference signal is C = Cr, and the Cb color difference component is C = Cb.

  The first weighting factor is a weighting factor corresponding to the first predicted image, and is a parameter whose value is determined (fluctuates) by the fixed-point precision LWD. The second weighting coefficient is a weighting coefficient corresponding to the second predicted image, and is a parameter whose value is determined (fluctuates) by the fixed-point precision LWD. The fixed-point precision LWD is a parameter that controls the step size corresponding to the fractional precision of the first weight coefficient and the second weight coefficient. As the fixed-point precision LWD, different values may be used depending on the luminance and the color difference. However, here, for the sake of simplicity, description will be made without explicitly dividing each color signal. For example, when w0C is expressed as a real value, 1.0 (1 in binary notation) and when LWD is 5, the first weighting factor is 32 (100,000 in binary notation), and w1C is expressed as a real value. In the case of 2.0 (10 in binary notation) and LWD of 5, the second weight coefficient is 64 (1000000 in binary notation). The first offset o0C is an offset corresponding to the first predicted image, and the second offset o1C is an offset corresponding to the second predicted image.

  When the WP parameter information is input, the WP parameter control unit 303 checks whether the value of the weight information is within the specified range, and resets the value outside the range to a value within the range, Change the value of the WP application flag. For example, when w0C is expressed as a real value, when the LWD is 7, the first weight coefficient is 384. Here, 384 is outside the range of the first weighting factor and cannot be used, but 96 is within the range of the first weighting factor and can be used. In this case, the WP parameter control unit 303 sets the LWD to 5 and the first weighting factor to 96, so that when the w0C is expressed by a real value, the WP parameter control unit 303 keeps the first weighting factor within the range while maintaining 3.0. It may be set. At this time, the WP parameter control unit 303 may perform quantization processing. For example, when LWD is 7 and the first weighting factor is 385, the WP parameter control unit 303 performs quantization processing to set the first weighting factor to 384, and then sets LWD to 5 and w0C to real values. The first weighting factor may be reset to 96 by setting it to 3.0 when expressed. In addition, the WP parameter control unit 303 may change the value of the first WP application flag from 1 to 0 so that the weighted motion compensation prediction is not used. Although not limited to these methods, the WP parameter control unit 303 performs control so that the value of the weight information is not used beyond a prescribed range defined by specifications or the like.

  The WP selectors 304 and 305 switch the connection end of each predicted image based on the WP application flag input from the WP parameter control unit 303. When each WP application flag is 0, the WP selectors 304 and 305 connect each output terminal to the default motion compensation unit 301. Then, the WP selectors 304 and 305 output the first predicted image and the second predicted image and input them to the default motion compensation unit 301. On the other hand, when each WP application flag is 1, the WP selectors 304 and 305 connect each output terminal to the weighted motion compensation unit 302. The WP selectors 304 and 305 output the first predicted image and the second predicted image, and input them to the weighted motion compensation unit 302.

  The default motion compensation unit 301 performs an average value process (default motion compensation prediction) based on the two unidirectional prediction images (first prediction image and second prediction image) input from the WP selectors 304 and 305, and performs the prediction image. Is generated. Specifically, when the first WP application flag and the second WP application flag are 0, the default motion compensation unit 301 performs average value processing based on Expression (1).

  P [x, y] = Clip1 ((PL0 [x, y] + PL1 [x, y] + offset2) >> (shift2)) (1)

  Here, P [x, y] is a predicted image, PL0 [x, y] is a first predicted image, and PL1 [x, y] is a second predicted image. offset2 and shift2 are parameters of the rounding process in the average value process, and are determined by the internal calculation accuracy of the first predicted image and the second predicted image. Clip1 (X) is a function for clipping the variable X within a specific bit precision, and here, the clip is clipped within the bit precision of the predicted image. For example, when the bit accuracy of the predicted image is L = 8, a value outside the range of 0 to 255 is clipped from 0 to 255. More specifically, a value of 0 or less is set to 0, and a value greater than 255 is set to 255.

  If the bit accuracy of the prediction image is L and the bit accuracy of the first prediction image and the second prediction image is M (L ≦ M), shift2 is formulated by Equation (2), and offset2 is formulated by Equation (3). Is done.

  shift2 = (ML + 1) (2)

  offset2 = (1 << (shift2-1) (3)

  For example, when the bit accuracy of the predicted image is 8 and the bit accuracy of the first predicted image and the second predicted image is 14, shift2 = 7 from Equation (2), and offset2 = (1 << from Equation (3). 6).

  When the prediction mode indicated by the motion information (prediction parameter) is unidirectional prediction, the default motion compensation unit 301 uses only the first prediction image and calculates a final prediction image based on Expression (4). calculate.

  P [x, y] = Clip1 ((PLX [x, y] + offset1) >> (shift1)) (4)

  Here, PLX [x, y] indicates a unidirectional prediction image (first prediction image), and X is an identifier indicating either 0 or 1 in the reference list. For example, when the reference list is 0, PL0 [x, y] is obtained, and when the reference list is 1, PL1 [x, y] is obtained. offset1 and shift1 are parameters of the rounding process, and are determined by the internal calculation accuracy of the first predicted image. When the bit accuracy of the predicted image is L and the bit accuracy of the first predicted image is M, shift1 is formulated by Equation (5), and offset1 is formulated by Equation (6).

  shift1 = (ML) (5)

  offset1 = (1 << (shift1-1) (6)

  For example, when the bit accuracy of the predicted image is 8 and the bit accuracy of the first predicted image is 14, shift1 = 6 from Equation (5) and offset1 = (1 << 5) from Equation (6).

  The weighted motion compensation unit 302 is based on the two unidirectional prediction images (first prediction image and second prediction image) input from the WP selectors 304 and 305 and the weight information input from the WP parameter control unit 303. Performs weighted motion compensation (weighted motion compensation prediction).

  Here, the weight coefficient will be further described. FIG. 7 is a reference diagram for explaining the weighting coefficient, and shows an example of a change in gradation value of a moving image having a change in pixel value in the time direction. In the example shown in FIG. 7, the encoding target frame is Frame (t), the previous frame in time is Frame (t−1), and the next frame in time is Frame (t + 1). As shown in FIG. 7, in a fade image that changes from white to black, the brightness (gradation value) of the image decreases with time. The value of the weighting coefficient means the degree of change in the pixel value as described in FIG. 7, and when expressed as a real value when there is no change in the pixel value (when the change in the pixel value is 0), the value is 1.0. I take the.

  Here, the case where there is no change in the pixel value will be further described. For example, assuming a moving image in which the same still image is temporally continuous, the luminance change between the screens is zero. In this case, the weighted motion compensation unit 302 is equivalent to performing the default motion compensated prediction because the pixel value change is 0 even if the weighted motion compensated prediction is performed. In such a case, that is, when there is no change in pixel value, the weighted motion compensation unit 302 realizes the default motion compensated prediction by weighted motion compensated prediction by selecting the reference value of the weight coefficient. Here, the reference value of the weighting factor can be derived according to the fixed-point precision and becomes (1 << LWD).

  In general, changes in pixel values of a moving image such as a fade effect and a dissolve effect do not change so much for each frame, so that the value of the weight coefficient is biased to around 1.0 when expressed in real values. . In the first embodiment, since the value of the weighting factor is quantized to a fixed-point precision indicated by a power of 2, even if a slight change in the average pixel value occurs between the two images, 1 / If the change is less than 128 precision, the value of the weighting factor is quantized to 1.0 when expressed as a real value. For this reason, in 1st Embodiment, even if a pixel value change arises, it can be handled as a case where there is substantially no pixel value change. In the following, for the sake of simplicity, the description will be made assuming that there is no change in the pixel value, that is, the case where the value of the weighting coefficient is expressed as a real value of 1.0. In addition, when there is no pixel value change, it corresponds to a pixel value change being below a predetermined value (a value sufficiently smaller than the accuracy of the weighting factor).

  For this reason, when the first WP application flag and the second WP application flag are 1, the weighted motion compensation unit 302 performs the weighting process based on Expression (7).

  P [x, y] = Clip1 (((PL0 [x, y] * w0C + PL1 [x, y] * w1C + ((o0C + o1C + 1) << LWD)) >> (LWD + 1))) (7)

  Note that the weighted motion compensation unit 302 realizes a rounding process by controlling the LWD as expressed by Equation (8) when the calculation accuracy of the first predicted image, the second predicted image, and the predicted image is different.

  LWD '= LWD + offset1 (8)

  The rounding process can be realized by replacing the LWD in Expression (7) with LWD ′ in Expression (8). For example, when the bit accuracy of the prediction image is 8 and the bit accuracy of the first prediction image and the second prediction image is 14, by resetting the LWD, the same calculation accuracy as shift2 in Equation (1) can be obtained. A batch rounding process can be realized.

  When the prediction mode indicated by the motion information (prediction parameter) is unidirectional prediction, the weighted motion compensation unit 302 uses only the first predicted image and uses the final predicted image based on Equation (9). Is calculated.

  P [x, y] = Clip1 ((PLX [x, y] * wXC + (1 << (LWD-1))) >> (LWD) + oXC) (9)

  Here, PLX [x, y] indicates a unidirectional prediction image (first prediction image), wXC indicates a weighting factor corresponding to unidirectional prediction, and X indicates either 0 or 1 in the reference list. OXC indicates an offset corresponding to unidirectional prediction. For example, when the reference list is 0, PL0 [x, y], w0C, o0C, and when the reference list is 1, PL1 [x, y], w1C, o1C.

  Note that the weighted motion compensation unit 302 controls the LWD as in Expression (8) in the same manner as in bidirectional prediction when the calculation accuracy of the first predicted image, the second predicted image, and the predicted image is different. Implement rounding.

  The rounding process can be realized by replacing the LWD in Expression (9) with LWD ′ in Expression (8). For example, when the bit accuracy of the predicted image is 8 and the bit accuracy of the first predicted image is 14, a batch rounding process with the same calculation accuracy as shift1 of Equation (4) is realized by resetting the LWD It becomes possible to do.

  In the case of unidirectional prediction, since various parameters (second WP adaptation flag, second weighting factor, and second offset information) corresponding to the second predicted image are not used, they are set to predetermined initial values. May be.

  Returning to FIG. 1, the motion evaluation unit 109 performs motion evaluation between a plurality of frames based on the input image and the reference image input from the predicted image generation unit 107, outputs motion information and WP parameter information, and stores the motion information. The prediction image generation unit 107 and the encoding unit 110 are input, and the WP parameter information is input to the prediction image generation unit 107 and the index setting unit 108.

  For example, the motion evaluation unit 109 calculates an error by calculating a difference value starting from a plurality of reference images corresponding to the same position as the input image of the prediction target pixel block, shifts the position with a fractional accuracy, and minimizes the error. Optimal motion information is calculated by a technique such as block matching for searching for a block. In the case of bidirectional prediction, the motion evaluation unit 109 performs block matching including default motion compensation prediction as shown in Equation (1) and Equation (4) using motion information derived by unidirectional prediction. Thus, motion information for bidirectional prediction is calculated.

  At this time, the motion evaluation unit 109 can calculate WP parameter information by performing block matching including weighted motion compensated prediction as represented by Equation (7) and Equation (9). Note that the WP parameter information may be calculated using a method of calculating a weighting factor or an offset using a pixel gradient of the input image, or a method of calculating a weighting factor or an offset by accumulating prediction errors when encoded. Good. The WP parameter information may be a fixed value determined in advance for each encoding device.

  Here, referring to FIG. 7 again, a method for calculating the weighting coefficient, the fixed-point precision of the weighting coefficient, and the offset from a moving image with temporally changing pixel values will be described. As described above, in a fade image that changes from white to black as shown in FIG. 7, the pixel value (gradation value) decreases with time. The motion evaluation unit 109 can calculate a weighting coefficient by calculating this inclination.

  The fixed-point precision of the weighting factor is information indicating the accuracy of the slope, and the motion evaluation unit 109 can calculate an optimum value from the temporal distance of the reference image and the degree of change in the image value. For example, in FIG. 7, when the value of the weighting coefficient between Frame (t-1) and Frame (t + 1) is expressed in real number accuracy, it is 0.75, and if it is 1/4 accuracy, 3/4 can be expressed. Therefore, the motion evaluation unit 109 sets the fixed point precision to 2 (1 << 2). Since the value of the fixed-point precision affects the code amount when the weighting coefficient is encoded, an optimal value may be selected in consideration of the code amount and the prediction accuracy. Note that the fixed-point precision value may be a predetermined fixed value.

  Further, when the inclinations do not match, the motion evaluation unit 109 can calculate the offset value by obtaining a correction value (deviation amount) corresponding to the intercept of the linear function. For example, in FIG. 7, when the value of the weighting coefficient between Frame (t−1) and Frame (t + 1) is expressed in real number precision, it is 0.60, and the fixed point precision is 1 (1 << 1). There is a high possibility that the weighting coefficient is set to 1 (that is, 0.50 when the value of the weighting coefficient is expressed with real number precision). In this case, since the decimal point accuracy of the weighting coefficient is deviated from 0.60, which is the optimum value, by 0.10, the motion evaluation unit 109 calculates the correction value for this amount from the maximum value of the pixel and uses it as an offset value. Set. When the maximum value of the pixel is 255, the motion evaluation unit 109 may set a value such as 25 (255 × 0.1).

  In the first embodiment, the motion evaluation unit 109 is illustrated as a function of the encoding device 100. However, the motion evaluation unit 109 is not an essential component of the encoding device 100. For example, the motion evaluation unit 109 is encoded. It is good also as an apparatus outside the conversion apparatus 100. In this case, the motion information and WP parameter information calculated by the motion evaluation unit 109 may be loaded into the encoding device 100.

  The index setting unit 108 receives the WP parameter information input from the motion evaluation unit 109, checks the reference list (list number) and the reference image (reference number), outputs the index information, and outputs the index information to the encoding unit 110. input. The index setting unit 108 generates index information by mapping the WP parameter information input from the motion evaluation unit 109 to syntax elements described later. At this time, the index setting unit 108 derives a selection range of the weighting factor and confirms that the weighting factor is included in the selection range.

  Here, derivation of the selection range of the weight coefficient will be described.

  In the first embodiment, as described above, it is assumed that there is no change in the pixel value and the weight coefficient value is expressed as a real value of 1.0. In this case, the weight coefficient / weight coefficient reference value is used. = 1 holds. Since the reference value of the weighting factor is (1 << LWD) as described above, in the first embodiment, the weighting factor is (1 << LWD), which is the same value as the reference value of the weighting factor.

  H. In H.264, etc., an index such as a weighting factor and an offset is defined to take a signed 8-bit value from −128 to 127, and a fixed-point precision is defined to take a value from 0 to 7. . For this reason, in the first embodiment, there is a case where the weighting coefficient is outside the specified range.

  FIG. 2 is a reference diagram showing a selection range of weighting factors in H.264, and shows weighting factors (1 << LWD) when the fixed-point precision LWD takes a value from 0 to 7, respectively. As is clear from FIG. 8, the weighting factor takes a positive value near 0 when the value of the fixed-point precision LWD becomes small, but becomes 128 when the value of the fixed-point precision LWD becomes 7, H.264 is outside the specified range.

  In this way, H.C. According to the H.264 standard, the range of the weighting factor that is desired to be used is outside the standard range, which may not be practical. In the unidirectional prediction, even if a weighting factor corresponding to the negative direction is selected, the prediction pixel value output in the unidirectional prediction is highly likely to be clipped to the bit range of the input image and become zero. The weighting factor corresponding to the direction cannot be substantially selected. On the other hand, in bidirectional prediction, in order to realize extrapolation prediction, it is possible to use one unidirectional prediction weighting factor as a negative value and the other weighting factor as a positive value. In many cases, the negative value does not require the same accuracy as the positive value as the range of the weighting factor.

  Therefore, in the first embodiment, the index setting unit 108 derives a selection range of weighting factors by assigning values in a negative direction and a positive direction with the reference value of the weighting factor as a substantial center, and the derived weighting factor. Confirm that the weighting factor is included in the selected range.

  FIG. 9 is an explanatory diagram illustrating an example of a selection range of weighting factors according to the first embodiment. In the example shown in FIG. 9, unlike the weighting factor selection range described in FIG. 8, the weighting factor reference value (1 << LWD) is arranged so as to be located at the approximate center of the selection range, and the weighting factor reference A value obtained by subtracting 128 from the value (−128+ (1 << LWD)) is the minimum value of the selection range, and a value obtained by adding 127 to the reference value of the weighting coefficient (127+ (1 << LWD)) is the maximum value of the selection range. It is a value.

  The index setting unit 108 sets the selection range of the weighting factor using Expression (10) and Expression (11). In addition, the minimum value of the selection range is formulated by Formula (10), and the maximum value of the selection range is formulated by Formula (11).

  min_wXC = −128 + (1 << LWD) (10)

  max_wXC = 127 + (1 << LWD) (11)

  10A and 10B are explanatory diagrams illustrating specific examples of the weight coefficient selection range according to the first embodiment. FIG. 10A illustrates the weight coefficient selection range when the value of the fixed-point precision LWD is 7, FIG. 10B shows the selection range of the weighting coefficient when the value of the fixed point precision LWD is 5. In the example shown in FIG. 10A, the reference value 128 of the weighting coefficient is arranged so as to be positioned substantially at the center of the selection range, the minimum value of the selection range is 0, and the maximum value of the selection range is 255. In the example shown in FIG. 10B, the reference value 32 of the weighting coefficient is arranged so as to be positioned substantially at the center of the selection range, the minimum value of the selection range is −96, and the maximum value of the selection range is 159. .

  FIG. FIG. 12 is a reference diagram illustrating a minimum value and a maximum value of a weight coefficient selection range in H.264, and FIG. 12 is an explanatory diagram illustrating an example of a minimum value and a maximum value of a weight coefficient selection range according to the first embodiment. As shown in FIG. In H.264, the minimum value and the maximum value of the selection range of the weighting factor do not depend on the reference value of the weighting factor and are constant. On the other hand, as shown in FIG. 12, in the first embodiment, the minimum value and the maximum value of the weight coefficient selection range vary depending on the reference value of the weight coefficient.

  As shown in FIG. 12, when the selection range of the weighting factor is set with the reference value of the weighting factor approximately as the center, the range that the weighting factor can take is from −127 to 255, and the signed 9-bit accuracy is required. . For this reason, in the first embodiment, the encoding unit 110 described later updates a weighting coefficient set as an index, that is, a value to be encoded, to a difference value between the weighting coefficient and the reference value of the weighting coefficient. As shown in FIG. 9, when the reference value of the weighting factor is subtracted from the derived weighting factor selection range, it can be seen that the range of the difference value of the weighting factor takes a signed 8-bit value from −128 to 127. In other words, when the weight coefficient selection range is set around the weight coefficient reference value, the weight coefficient selection range varies depending on the weight coefficient reference value. By subtracting the value, the range of the difference value of the weight coefficient becomes constant without depending on the reference value of the weight coefficient. As described above, in the first embodiment, the weighting factor is replaced with the difference value of the weighting factor, so that the selection range of the weighting factor can be expanded and the selection range of signed 8-bit precision can be defined.

  If the index setting unit 108 confirms that the weighting factor is not included in the derived weighting factor selection range, the index setting unit 108 may perform clip processing using the maximum value or the minimum value of the weighting factor selection range. Good. In this case, the index setting unit 108 may clip at the minimum value if the weighting factor is smaller than the minimum value of the selection range, and may clip at the maximum value if the weighting factor is larger than the maximum value of the selection range. By introducing such clip processing, the encoding target value such as the difference value of the weighting coefficient takes a value within a predetermined bit accuracy without setting a specific range constraint. The configuration of the circuit scale used by can be clarified.

  In the first embodiment, it is assumed that the selection range of the weighting factor has a signed 8-bit accuracy. However, the accuracy of the selection range of the weighting factor is not limited to this. It is good also as a 9-bit precision. In this case, the selection range of the weighting factor is −256 to 255, but −128 in Expression (10) may be replaced with −256, and 127 in Expression (11) may be replaced with 255.

  In the first embodiment, the example in which the index setting unit 108 derives the weight coefficient selection range has been described. However, the present invention is not limited to this, and the encoding unit 110 may perform the derivation. That is, the index setting unit 108 may be a derivation unit, and the encoding unit 110 may be a derivation unit.

  13A and 13B are diagrams illustrating an example of WP parameter information input to the index setting unit 108 according to the first embodiment. An example of WP parameter information at the time of P-slice is as shown in FIG. 13A, and an example of WP parameter information at the time of B-slice is as shown in FIGS. 13A and 13B. The list number is an identifier indicating the prediction direction, and takes a value of 0 when unidirectional prediction is used, and two values of 0 and 1 can be used when bidirectional prediction is used. The reference numbers are values corresponding to 1 to N indicated in the frame memory 206. Since the WP parameter information is held for each reference list and reference image, the information necessary for the B-slice is 2N when the number of reference images is N.

  FIG. 14 is a flowchart illustrating an example of a process for deriving a selection range of weighting factors according to the first embodiment. Here, the case where the index setting unit 108 performs the process of deriving the selection range of the weighting factor will be described. However, as described above, the encoding unit 110 may perform the processing.

  First, the index setting unit 108 derives the fixed point precision LWD of the weighting factor (Step S02). Here, the index setting unit 108 may derive the fixed point precision LWD of the weighting coefficient from the WP parameter information or the index information.

  Subsequently, the index setting unit 108 derives a reference value (1 << LWD) of the weighting factor using the derived fixed point precision LWD (step S03).

  Subsequently, the index setting unit 108 subtracts 128 from the derived weight coefficient reference value (1 << LWD) to derive the minimum value of the weight coefficient selection range (step S04).

  Subsequently, the index setting unit 108 adds 127 to the derived reference value (1 << LWD) of the weighting factor to derive the maximum value of the weighting factor selection range (step S05).

  Then, the index setting unit 108 confirms that the weighting factor is included in the derived weighting factor selection range.

  Returning to FIG. 1, the encoding unit 110 selects the quantized transform coefficient input from the quantization unit 103, the motion information input from the motion evaluation unit 109, the index information input from the index setting unit 108, and the weighting factor. Encoding processing is performed on the range and various encoding parameters such as quantization information specified by the encoding control unit 111 to generate encoded data. The encoding process corresponds to, for example, Huffman encoding or arithmetic encoding.

  The encoding parameter is a parameter required for decoding, such as prediction information indicating a prediction method, information on quantization transform coefficients, information on quantization, and the like. For example, the encoding control unit 111 has an internal memory (not shown), the encoding parameters are held in the internal memory, and the encoding parameters of the adjacent already encoded pixel block are used when encoding the pixel block. You can For example, H.M. In the H.264 intra prediction, the prediction information of the pixel block can be derived from the prediction information of the encoded adjacent block.

  The encoding unit 110 outputs the generated encoded data according to an appropriate output timing managed by the encoding control unit 111. The output encoded data is, for example, multiplexed with various information such as a multiplexing unit (not shown) and temporarily stored in an output buffer (not shown). (Storage medium) or transmission system (communication line).

  The encoding unit 110 includes an entropy encoding unit 110A and an index reconstruction unit 110B.

  The entropy encoding unit 110A performs encoding processing such as variable length encoding and arithmetic encoding on the input information. For example, H.M. In H.264, context-adaptive variable-length coding (CAVLC) and context-adaptive arithmetic coding (CABAC: Context based Adaptive Binary Coding) are used.

  In order to reduce the code length of the syntax element of the index information input from the index setting unit 108, the index reconstruction unit 110B performs a prediction process according to the characteristics of the parameter of the syntax element, and uses the value of the syntax element as it is (directly). Value) and the difference between the predicted values and output to the entropy encoding unit 110A. A specific example of the prediction process will be described later. In addition, when the encoding unit 110 performs the derivation of the weight coefficient selection range, it is performed by the index reconstruction unit 110B.

  FIG. 15 is a diagram illustrating an example of a syntax 500 used by the encoding device 100 according to the first embodiment. A syntax 500 indicates a structure of encoded data generated by the encoding apparatus 100 by encoding an input image (moving image data). When decoding the encoded data, the decoding apparatus described later refers to the same syntax structure as that of the syntax 500 and interprets the syntax of the moving image data.

  The syntax 500 includes three parts: a high level syntax 501, a slice level syntax 502, and a coding tree level syntax 503. The high level syntax 501 includes syntax information of a layer higher than the slice. The syntax information includes, for example, tile-shaped division information described in the example illustrated in FIG. 2B. A slice refers to a rectangular area or a continuous area included in a frame or a field. The slice level syntax 502 includes information necessary for decoding each slice. The coding tree level syntax 503 includes information necessary for decoding each coding tree (ie, each coding tree block). Each of these parts includes more detailed syntax.

  The high level syntax 501 includes sequence and picture level syntaxes such as a sequence parameter set syntax 504, a picture parameter set syntax 505, and an adaptation parameter set syntax 506.

  The slice level syntax 502 includes a slice header syntax 507, a breadweight table syntax 508, a slice data syntax 509, and the like. The tread weight table syntax 508 is called from the slice header syntax 507.

  The coding tree level syntax 503 includes a coding tree unit syntax 510, a transform unit syntax 511, a prediction unit syntax 512, and the like. The coding tree unit syntax 510 may have a quadtree structure. Specifically, the coding tree unit syntax 510 can be recursively called as a syntax element of the coding tree unit syntax 510. That is, one coding tree block can be subdivided with a quadtree. The coding tree unit syntax 510 includes a transform unit syntax 511. The transform unit syntax 511 is called in each coding tree unit syntax 510 at the extreme end of the quadtree. The transform unit syntax 511 describes information related to inverse orthogonal transformation and quantization. In these syntaxes, information related to weighted motion compensation prediction may be described.

  FIG. 16 is a diagram illustrating an example of the picture parameter set syntax 505 according to the first embodiment. The weighted_pred_flag is a syntax element indicating, for example, whether the weighted compensated prediction of the first embodiment related to P-slice is valid or invalid. When weighted_pred_flag is 0, the weighted motion compensation prediction of the first embodiment in the P-slice is invalid. Accordingly, the WP application flag included in the WP parameter information is always set to 0, and the WP selectors 304 and 305 connect the respective output terminals to the default motion compensation unit 301. On the other hand, when weighted_pred_flag is 1, the weighted motion compensation prediction of the first embodiment in the P-slice is valid.

  As another example, when weighted_pred_flag is 1, in the syntax of lower layers (slice header, coding tree block, transform unit, prediction unit, etc.) The validity or invalidity of the weighted motion compensated prediction of one embodiment may be specified.

  For example, weighted_bipred_idc is a syntax element indicating whether the weighted compensated prediction of the first embodiment related to B-slice is valid or invalid. When weighted_bipred_idc is 0, the weighted motion compensated prediction of the first embodiment in the B-slice is invalid. Accordingly, the WP application flag included in the WP parameter information is always set to 0, and the WP selectors 304 and 305 connect the respective output terminals to the default motion compensation unit 301. On the other hand, when weighted_bipred_idc is 1, the weighted motion compensated prediction of the first embodiment in the B-slice is valid.

  As another example, when weighted_bipred_idc is 1, in the syntax of lower layers (slice header, coding tree block, transform unit, etc.), the syntax of the first embodiment is determined for each local region inside the slice. The validity or invalidity of the weighted motion compensation prediction may be defined.

  FIG. 17 is a diagram illustrating an example of the slice header syntax 507 according to the first embodiment. slice_type indicates the slice type (I-slice, P-slice, B-slice, etc.) of the slice. pic_parameter_set_id is an identifier indicating which picture parameter set syntax 505 is to be referred to. num_ref_idx_active_override_flag is a flag indicating whether or not the number of valid reference images is updated. When this flag is 1, num_ref_idx_l0_active_minus1 and num_ref_idx_l1_active_minus1 that define the number of reference images in the reference list can be used. pred_weight_table () is a function indicating the predicate weight table syntax used for weighted motion compensation prediction. When the above weighted_pred_flag is 1 and P-slice, and when weighted_bipred_idc is 1 and B-slice, this function is Called.

  FIG. 18 is a diagram illustrating an example of the tread weight table syntax 508 according to the first embodiment. luma_log2_weight_denom represents the fixed-point precision (LWD) of the weighting factor of the luminance signal in the slice, and is a value corresponding to the LWD in Expression (7) or Expression (9). delta_chroma_log2_weight_denom represents the fixed point precision of the weight coefficient of the color difference signal in the slice, and a derivation method will be described later. chroma_format_idc is an identifier representing a color space, and MONO_IDX is a value indicating a monochrome video. num_ref_common_active_minus1 indicates a value obtained by subtracting 1 from the number of reference images included in the common list in the slice.

  luma_weight_l0_flag and luma_weight_l1_flag indicate WP adaptation flags in the luminance signals corresponding to List 0 and List 1, respectively. When this flag is 1, the weighted motion compensated prediction of the luminance signal of the first embodiment is effective over the entire slice. chroma_weight_l0_flag and chroma_weight_l1_flag indicate WP adaptation flags in the color difference signals corresponding to List 0 and List 1, respectively. When this flag is 1, the weighted motion compensation prediction of the chrominance signal of the first embodiment is effective over the entire slice. luma_weight_l0 [i] and luma_weight_l1 [i] are weighting factors of the luminance signal corresponding to the reference number i managed in list 0 and list 1, respectively. luma_offset_l0 [i] and luma_offset_l1 [i] are offsets of the luminance signal corresponding to the i-th reference number managed in list 0 and list 1, respectively. These are values corresponding to w0C, w1C, o0C, and o1C, respectively, of Equation (7) or Equation (9). However, C = Y.

  chroma_weight_l0 [i] [j] and chroma_weight_l1 [i] [j] are the weighting coefficients of the color difference signals corresponding to the i-th reference number managed in list 0 and list 1, respectively. chroma_offset_l0 [i] [j] and chroma_offset_l1 [i] [j] are offsets of the color difference signal corresponding to the reference number i managed in list 0 and list 1, respectively. These are values corresponding to w0C, w1C, o0C, and o1C, respectively, of Equation (7) or Equation (9). However, C = Cr or Cb. j indicates a component of a color difference component. For example, in the case of a YUV 4: 2: 0 signal, j = 0 indicates a Cr component and j = 1 indicates a Cb component. In this description, j = 0 may be used as the Cb component and j = 1 may be used as the Cr component.

  Here, the detail of the prediction method of each syntax element relevant to the weighted prediction in a syntax structure is demonstrated. The prediction of syntax elements is performed by the index reconstruction unit 110B. In the example shown in FIG. 18, syntax elements into which prediction is introduced are shown with a delta prefix.

  First, a prediction method between signals of luma_log2_weight_denom and chroma_log2_weight_denom indicating the fixed point precision of the weight coefficient will be described. The index reconstruction unit 110B performs prediction processing between the luma_log2_weight_denom and chroma_log2_weight_denom signals using Equation (12), and performs restoration processing using Equation (13). Here, as shown in FIG. 18, since luma_log2_weight_denom is defined first, chroma_log2_weight_denom is predicted from the value of luma_log2_weight_denom.

  delta_chroma_log2_weight_denom = (chroma_log2_weight_denom-luma_log2_weight_denom)… (12)

  chroma_log2_weight_denom = (luma_log2_weight_denom + delta_chroma_log2_weight_denom)… (13)

  Since the fading effect generally has few cases of causing different temporal changes depending on the color space, the fixed-point accuracy for each signal component has a strong correlation between the luminance component and the color difference component. For this reason, it is possible to reduce the amount of information indicating the fixed-point precision by performing prediction in the color space in this way.

  In Equation (12), the luminance component is subtracted from the color difference component, but the color difference component may be subtracted from the luminance component. In this case, the equation (13) may be modified according to the equation (12).

  Next, a description will be given of prediction methods for luma_weight_lx [i] and chroma_weight_lx [i] [j] representing the weighting coefficients of the luminance and chrominance signals, respectively. Here, x is an identifier indicating 0 or 1. The values of luma_weight_lx [i] and chroma_weight_lx [i] [j] change according to the values of luma_log2_weight_denom and chroma_log2_weight_denom, respectively. For example, when the value of luma_log2_weight_denom is 3, luma_weight_lx [i] when there is no change in pixel value is (1 << 3). On the other hand, when the value of luma_log2_weight_denom is 5, luma_weight_lx [i] assuming that there is no change in brightness is (1 << 5).

  For this reason, the index reconstruction unit 110B performs the prediction process using the weighting coefficient when there is no change in the pixel value as the reference coefficient (default value). Specifically, the index reconstruction unit 110B performs prediction processing of luma_weight_lx [i] using Equations (14) to (15), and performs restoration processing using Equation (16). Similarly, the index reconstruction unit 110B performs the prediction process of chroma_weight_lx [i] using Expressions (17) to (18), and performs the restoration process using Expression (19).

  delta_luma_weight_lx [i] = (luma_weight_lx [i]-default_luma_weight_lx) (14)

  default_luma_weight_lx = (1 << luma_log2_weight_denom)… (15)

  luma_weight_lx [i] = (default_luma_weight_lx + delta_luma_weight_lx [i]) (16)

  delta_chroma_weight_lx [i] [j] = (chroma_weight_lx [i] [j]-default_chroma_weight_lx) (17)

  default_chroma_weight_lx = (1 << chroma_log2_weight_denom)… (18)

  chroma_weight_lx [i] [j] = (default_chroma_weight_lx + delta_chroma_weight_lx [i] [j]) (19)

  Here, default_luma_weight_lx and default_chroma_weight_lx are reference values (default values) with no pixel value change in the luminance component and the color difference component, respectively.

  An image including a fade effect is often faded at a specific fade change point, and the other images are often normal natural images or images without a fade effect. In this case, the weighting coefficient is often used when there is no change in the pixel value. Therefore, by deriving the initial value when there is no change in the pixel value from the fixed-point precision and using it as the predicted value, the code amount of the weighting coefficient can be reduced.

  Next, a prediction method of chroma_offset_lx [i] [j] representing the offset of the color difference signal will be described. In the YUV color space, the color difference component expresses the color by the amount of deviation from the median. For this reason, the amount of change from the pixel value change in consideration of the median value using the weighting coefficient can be used as the predicted value. Specifically, the index reconstruction unit 110B performs a prediction process of chroma_offset_lx [i] [j] using Expressions (20) to (21), and performs a restoration process using Expression (22).

  delta_chroma_offset_lx [i] [j] = (chroma_offset_lx [i] [j] + ((MED * chroma_weight_lx [i] [j]) >> chroma_log2_weight_denom)-MED)… (20)

  MED = (MaxChromaValue >> 1) (21)

  Here, MaxChromaValue indicates the maximum pixel value from which a color difference signal can be obtained. For example, in the case of an 8-bit signal, MaxChromaValue is 255 and MED is 128.

  chroma_offset_lx [i] [j] = (delta_chroma_offset_lx [i] [j]-((MED * chroma_weight_lx [i] [j]) >> chroma_log2_weight_denom) + MED)… (22)

  By using a signal characteristic of the color difference signal and introducing a prediction value that takes into account the amount of deviation from the median value, the code amount of the offset value of the color difference signal is reduced compared to when the offset value is encoded as it is. it can.

  FIG. 19 is an explanatory diagram illustrating an example of a relationship between values of syntax elements according to the first embodiment, and illustrates a relationship between values of luma_log2_weight_denom, default_luma_weight_lx, luma_weight_lx [i], and delta_luma_weight_lx [i]. As shown in FIG. 19, delta_luma_weight_lx [i], which is a syntax element encoded by the entropy encoding unit 110A, that is, a possible range of the difference value of the weight coefficient is fixed to a range of −128 to 127. The signed 8-bit precision.

  As described above, in the first embodiment, the selection range of the weighting factor is derived by assigning values in the negative direction and the positive direction with the reference point of the weighting factor at which the pixel value change is 0 as a substantial center. Confirm that the weighting factor is included in the selection range of the weighting factor. For this reason, according to the first embodiment, the H.264 standard. Compared with H.264 or the like, the selection range of the weighting coefficient can be expanded, and a positive value having a high selection frequency can be easily obtained. According to the first embodiment, since the signed 8-bit value in which the difference value of the weighting coefficient to be encoded is −128 to 127 is taken as a fixed value, the weighted coefficient selection range is expanded and the signed value is added. A selection range with 8-bit precision can be defined.

  As described above, in the first embodiment, since the range of syntax to be encoded (difference value of weighting coefficient) can be fixed, the specification is compared with a configuration in which the encoder dynamically changes these ranges. It can be simplified. For example, when the syntax to be encoded is a weighting factor and the selection range of the weighting factor varies depending on the reference value of the weighting factor, the reference value of the weighting factor is associated with the minimum value and the maximum value of the selection range of the weighting factor. A configuration is required in which a table is prepared and the weighting coefficient selection range is referred to each time after derivation, or a weighting factor selection range is calculated and derived each time. In such a case, a configuration in which the table is loaded into the memory and referenced each time, and an arithmetic circuit for calculating the selection range of the weighting factor each time are required, which increases the hardware scale. On the other hand, in the first embodiment, since the range of syntax to be encoded (difference value of weighting coefficient) can be fixed, the hardware scale is reduced without being restricted by the hardware configuration as described above. be able to.

  Furthermore, in the first embodiment, the difference value of the weighting coefficient whose range is fixed to signed 8-bit precision is encoded, but the difference value of the weighting coefficient takes a value near the center of the range (near 0). Therefore, the code length at the time of encoding can be shortened, and encoding efficiency can be improved. H. In H.264, the weighting coefficient is encoded with a signed exponent Golomb code (se (v)). This code is an effective code for a symbol whose value to be encoded increases exponentially with reference to 0. For this reason, it is common to set the reference value with the highest frequency of use at the center of the range. In the first embodiment, a case where a pixel value change between pictures in a general moving image is 0 is set as a reference value of the weighting coefficient, and prediction from the reference value is also introduced in prediction of the selection range of the weighting coefficient. ing. As a result, since the exponent Golomb code and the selection range of the prediction and weighting coefficients match, the code amount reduction effect is high, and the coefficient range is determined around the reference value. The distance from the reference value of the negative value is approximately the same, and there is an effect that encoding can be performed with a shorter code length compared to the conventional method.

  Note that syntax elements not defined in the present embodiment may be inserted between the rows of the syntax tables illustrated in FIGS. 16 to 18 of the first embodiment, and descriptions regarding other conditional branches are included. May be. Further, the syntax table may be divided into a plurality of tables, or a plurality of syntax tables may be integrated. Moreover, the term of each illustrated syntax element can be changed arbitrarily.

(Second Embodiment)
In the second embodiment, a decoding device that decodes encoded data encoded by the encoding device of the first embodiment will be described. In the second embodiment, as in the first embodiment, the description will be made on the assumption that there is no change in pixel value, that is, the case where the value of the weighting factor is expressed as a real value of 1.0.

  FIG. 20 is a block diagram illustrating an example of the configuration of the decoding device 800 according to the second embodiment.

  The decoding device 800 decodes encoded data stored in an input buffer (not shown) into a decoded image and outputs the decoded image to an output buffer (not shown) as an output image. The encoded data is output from, for example, the encoding device 100 of FIG. 1 and the like, and is input to the decoding device 800 via a storage system, a transmission system, or a buffer (not shown).

  As illustrated in FIG. 20, the decoding device 800 includes a decoding unit 801, an inverse quantization unit 802, an inverse orthogonal transform unit 803, an addition unit 804, a predicted image generation unit 805, and an index setting unit 806. Prepare. The inverse quantization unit 802, the inverse orthogonal transform unit 803, the addition unit 804, and the predicted image generation unit 805 are respectively the inverse quantization unit 104, the inverse orthogonal transform unit 105, the addition unit 106, the predicted image generation unit 107, FIG. Are substantially the same or similar elements. Note that the decoding control unit 807 shown in FIG. 20 controls the decoding device 800 and can be realized by, for example, a CPU.

  The decoding unit 801 performs decoding based on the syntax for each frame or field in order to decode the encoded data. The decoding unit 801 includes an entropy decoding unit 801A and an index reconstruction unit 801B.

  The entropy decoding unit 801A sequentially entropy-decodes the code string of each syntax, such as motion information including a prediction mode, a motion vector, and a reference number, index information for weighted motion compensated prediction, a quantization transform coefficient, and the like. The encoding parameter of the encoding target block is reproduced. Note that entropy decoding is also referred to as parsing processing. Here, the encoding parameters are all parameters necessary for decoding such as information on transform coefficients and information on quantization in addition to the above.

  Specifically, the entropy decoding unit 801A has a function of performing decoding processing such as variable length decoding processing and arithmetic decoding processing on the input encoded data. For example, H.M. In H.264, context-adaptive variable-length coding (CAVLC: Context based Adaptive Length Coding) and context-adaptive arithmetic coding (CABAC: Context based Adaptive Binary Coding) are used. Is decoded into meaningful syntax elements. These processes are also called decryption processes.

  The index reconstruction unit 801B restores the decrypted index information and reconstructs the index information. Specifically, the index reconfiguration unit 801B performs prediction processing according to the characteristics of the syntax element parameters to reduce the code length of the decoded index information syntax element, restores the syntax element, and Reconstruct information. A specific example of the prediction process will be described later.

  The decoding unit 801 outputs the motion information, the index information, and the quantized transform coefficient, inputs the quantized transform coefficient to the inverse quantization unit 802, inputs the index information to the index setting unit 806, and converts the motion information into the predicted image. Input to the generation unit 805.

  The inverse quantization unit 802 performs an inverse quantization process on the quantized transform coefficient input from the decoding unit 801 to obtain a restored transform coefficient. Specifically, the inverse quantization unit 802 performs inverse quantization according to the quantization information used in the decoding unit 801. More specifically, the inverse quantization unit 802 multiplies the quantized transform coefficient by the quantization step size derived from the quantization information to obtain a restored transform coefficient. The inverse quantization unit 802 outputs the restored transform coefficient and inputs it to the inverse orthogonal transform unit 803.

  The inverse orthogonal transform unit 803 performs inverse orthogonal transform corresponding to the orthogonal transform performed on the encoding side on the reconstructed transform coefficient input from the inverse quantization unit 802 to obtain a reconstructed prediction error. The inverse orthogonal transform unit 803 outputs the restoration prediction error and inputs it to the addition unit 804.

  The adding unit 804 adds the restored prediction error input from the inverse orthogonal transform unit 803 and the corresponding predicted image to generate a decoded image. The adding unit 804 outputs the decoded image and inputs the decoded image to the predicted image generation unit 805. The adding unit 804 outputs the decoded image to the outside as an output image. The output image is then temporarily stored in an external output buffer (not shown) or the like. For example, in accordance with the output timing managed by the decoding control unit 807, the output image is sent to a display device system or a video device system (not shown). Is output.

  The index setting unit 806 receives the index information input from the decoding unit 801, converts it into WP parameter information, outputs it, and inputs it to the predicted image generation unit 805. Specifically, the index setting unit 806 receives the index information decoded by the entropy decoding unit 801A and reconfigured by the index reconfiguration unit 801B. Then, the index setting unit 806 confirms the list of reference images and the reference number, converts them into WP parameter information, and outputs the converted WP parameter information to the predicted image generation unit 805. When the index setting unit 806 converts the index information into WP parameter information, the index setting unit 806 derives a selection range of the weighting factor and confirms that the weighting factor is included in the selection range. The derivation of the weighting factor selection range is the same as in the first embodiment, and a detailed description thereof will be omitted. Further, the derivation of the weighting factor selection range may be performed by the index reconstruction unit 801B instead of the index setting unit 806. That is, the index setting unit 806 may be a derivation unit, and the index reconstruction unit 801B (decoding unit 801) may be a derivation unit.

  Further, the WP parameter information includes information on the first WP application flag, the second WP adaptation flag, and the weight information, as in the first embodiment. Also, the weight information includes the first weighting factor value w0C, the second weighting factor value w1C, the first weighting factor and the fixed weight precision LWD of the second weighting factor, the first offset o0C, and the first weighting factor as in the first embodiment. Contains information on two offsets o1C.

  The predicted image generation unit 805 generates a predicted image using the motion information input from the decoding unit 801, the WP parameter information input from the index setting unit 806, and the decoded image input from the addition unit 804.

  Here, the details of the predicted image generation unit 805 will be described with reference to FIG. Like the predicted image generation unit 107, the predicted image generation unit 805 includes a multi-frame motion compensation unit 201, a memory 202, a unidirectional motion compensation unit 203, a prediction parameter control unit 204, a reference image selector 205, and a frame memory 206. And a reference image control unit 207.

  The frame memory 206 stores the decoded image input from the addition unit 106 as a reference image under the control of the reference image control unit 207. The frame memory 206 has a plurality of memory sets FM1 to FMN (N ≧ 1) for temporarily storing reference images.

  The prediction parameter control unit 204 prepares a plurality of combinations of reference image numbers and prediction parameters as a table based on the motion information input from the decoding unit 801. Here, the motion information indicates a motion vector indicating a shift amount of motion used in motion compensation prediction, a reference image number, information on a prediction mode such as unidirectional / bidirectional prediction, and the like. The prediction parameter refers to information regarding a motion vector and a prediction mode. Then, the prediction parameter control unit 204 selects and outputs a combination of a reference image number and a prediction parameter used for generating a prediction image based on the motion information, inputs the reference image number to the reference image selector 205, and outputs the prediction parameter. Is input to the unidirectional motion compensation unit 203.

  The reference image selector 205 is a switch for switching which output terminal of the frame memories FM1 to FMN included in the frame memory 206 is connected according to the reference image number input from the prediction parameter control unit 204. For example, if the reference image number is 0, the reference image selector 205 connects the output end of FM1 to the output end of the reference image selector 205. If the reference image number is N-1, the reference image selector 205 refers to the output end of the FMN. Connected to the output terminal of the image selector 205. The reference image selector 205 outputs the reference image stored in the frame memory to which the output terminal is connected among the frame memories FM1 to FMN included in the frame memory 206, and inputs the reference image to the unidirectional motion compensation unit 203. Note that in the decoding device 800, since the reference image is not used except by the predicted image generation unit 805, the reference image does not have to be output to the outside of the predicted image generation unit 805.

  The unidirectional motion compensation unit 203 performs a motion compensation prediction process according to the prediction parameter input from the prediction parameter control unit 204 and the reference image input from the reference image selector 205, and generates a unidirectional prediction image. Since motion compensation prediction has already been described with reference to FIG. 5, description thereof will be omitted.

  The unidirectional motion compensation unit 203 outputs a unidirectional prediction image and temporarily stores it in the memory 202. Here, when the motion information (prediction parameter) indicates bidirectional prediction, the multi-frame motion compensation unit 201 performs weighted prediction using two types of unidirectional prediction images. The first unidirectional prediction image corresponding to the first is stored in the memory 202, and the second unidirectional prediction image is directly output to the multi-frame motion compensation unit 201. Here, the first unidirectional prediction image corresponding to the first is the first prediction image, and the second unidirectional prediction image is the second prediction image.

  Two unidirectional motion compensation units 203 may be prepared, and each may generate two unidirectional prediction images. In this case, when the motion information (prediction parameter) indicates unidirectional prediction, the unidirectional motion compensation unit 203 directly outputs the first unidirectional prediction image as the first prediction image to the multi-frame motion compensation unit 201. Good.

  The multi-frame motion compensation unit 201 uses the first prediction image input from the memory 202, the second prediction image input from the unidirectional motion compensation unit 203, and the WP parameter information input from the motion evaluation unit 109. A prediction image is generated by performing weighted prediction. The multi-frame motion compensation unit 201 outputs a prediction image and inputs the prediction image to the addition unit 804.

  Here, the details of the multi-frame motion compensation unit 201 will be described with reference to FIG. Similar to the predicted image generation unit 107, the multi-frame motion compensation unit 201 includes a default motion compensation unit 301, a weighted motion compensation unit 302, a WP parameter control unit 303, and WP selectors 304 and 305.

  The WP parameter control unit 303 outputs a WP application flag and weight information based on the WP parameter information input from the index setting unit 806, inputs the WP application flag to the WP selectors 304 and 305, and weights the weight information. Input to the motion compensation unit 302.

  Specifically, when the WP parameter information is input from the index setting unit 806, the WP parameter control unit 303 separates the WP parameter information into a first WP application flag, a second WP application flag, and weight information, and outputs them. The first WP application flag is input to the WP selector 304, the second WP application flag is input to the WP selector 305, and the weight information is input to the weighted motion compensation unit 302.

  When the WP parameter information is input, the WP parameter control unit 303 checks whether the value of the weight information is within a specified range. For example, when w0C is expressed as a real value, when the LWD is 7, the first weight coefficient is 384. Here, 384 is outside the range of the first weighting coefficient and cannot be used. In this case, since the WP parameter control unit 303 is data that violates the standard, information indicating that the standard is violated may be notified to the decoding control unit 807 to stop the decoding process. Further, the WP parameter control unit 303 may perform the clipping process within the range of the first weighting coefficient and proceed with the decoding process. In addition, the WP parameter control unit 303 may change the value of the first WP application flag from 1 to 0 and perform default motion compensation prediction.

  The WP selectors 304 and 305 switch the connection end of each predicted image based on the WP application flag input from the WP parameter control unit 303. When each WP application flag is 0, the WP selectors 304 and 305 connect each output terminal to the default motion compensation unit 301. Then, the WP selectors 304 and 305 output the first predicted image and the second predicted image and input them to the default motion compensation unit 301. On the other hand, when each WP application flag is 1, the WP selectors 304 and 305 connect each output terminal to the weighted motion compensation unit 302. The WP selectors 304 and 305 output the first predicted image and the second predicted image, and input them to the weighted motion compensation unit 302.

  The default motion compensation unit 301 performs an average value process based on the two unidirectional predicted images (first predicted image and second predicted image) input from the WP selectors 304 and 305 to generate a predicted image. Specifically, when the first WP application flag and the second WP application flag are 0, the default motion compensation unit 301 performs average value processing based on Expression (1).

  When the prediction mode indicated by the motion information (prediction parameter) is unidirectional prediction, the default motion compensation unit 301 uses only the first prediction image and calculates a final prediction image based on Expression (4). calculate.

  The weighted motion compensation unit 302 is based on the two unidirectional prediction images (first prediction image and second prediction image) input from the WP selectors 304 and 305 and the weight information input from the WP parameter control unit 303. Performs weighted motion compensation. Specifically, when the first WP application flag and the second WP application flag are 1, the weighted motion compensation unit 302 performs the weighting process based on Expression (7).

  Note that the weighted motion compensation unit 302 performs rounding processing by controlling the LWD, which is the fixed-point precision, as in Expression (8) when the calculation accuracy of the first prediction image, the second prediction image, and the prediction image is different. To realize.

  Also, when the prediction mode indicated by the motion information (prediction parameter) is unidirectional prediction, the weighted motion compensation unit 302 uses only the first prediction image and uses the final prediction image based on Equation (9). Is calculated.

  Note that the weighted motion compensation unit 302, when the calculation accuracy of the first predicted image, the second predicted image, and the predicted image is different, calculates the LWD that is the fixed-point accuracy as shown in Equation (8) as in the bidirectional prediction. Rounding processing is realized by controlling to.

  The fixed-point precision of the weighting factor has already been described with reference to FIG. In the case of unidirectional prediction, since various parameters (second WP adaptation flag, second weighting factor, and second offset information) corresponding to the second predicted image are not used, they are set to predetermined initial values. May be.

  Decoding section 801 uses syntax 500 shown in FIG. A syntax 500 indicates a structure of encoded data to be decoded by the decoding unit 801. The syntax 500 has already been described with reference to FIG. Also, the picture parameter set syntax 505 has already been described with reference to FIG. 16 except that encoding is decoding, and thus description thereof will be omitted. Also, the slice header syntax 507 has already been described with reference to FIG. 17 except that encoding is decoding, and thus description thereof is omitted. Also, the treadweight table syntax 508 has already been described with reference to FIG. 18 except that the encoding is decoding, and thus the description thereof will be omitted.

  Here, the detail of the prediction method of each syntax element relevant to the weighted prediction in a syntax structure is demonstrated. The prediction of syntax elements is performed by the index reconstruction unit 801B. The syntax configuration that explicitly indicates the prediction method of the second embodiment is the same as that of the first embodiment.

  For the prediction method between the signals of luma_log2_weight_denom and chroma_log2_weight_denom indicating the fixed-point precision of the weighting coefficient, restoration processing is performed using Equation (13).

  For the prediction method of luma_weight_lx [i] and chroma_weight_lx [i] [j] representing the weighting coefficients of the luminance and chrominance signals, restoration processing is performed using Equations (16) and (19).

  The plurality of prediction methods described above can be used not only independently but also in combination. For example, by combining Formula (13), Formula (15), Formula (19), and the like, it is possible to efficiently reduce the code amount of syntax elements in the index information.

  As described above, in the second embodiment, the weight coefficient selection range is derived by assigning values in the negative direction and the positive direction with the reference point of the weight coefficient at which the pixel value change is 0 as the approximate center. Confirm that the weighting factor is included in the selection range of the weighting factor. For this reason, according to the second embodiment, the H.264 standard. Compared with H.264 or the like, the selection range of the weighting coefficient can be expanded, and a positive value having a high selection frequency can be easily obtained. Further, according to the second embodiment, a signed 8-bit value in which a difference value of a weighting factor to be decoded is −128 to 127 is set as a fixed value, so that a signed 8 A selection range of bit precision can be defined.

  As described above, in the second embodiment, since the range of syntax to be decoded (difference value of weighting factors) can be fixed, the decoder determines whether or not the decoded encoded data is within a predetermined specification range. Can be easily checked and the specifications can be simplified. For example, when the syntax to be decoded is a weighting factor, and the selection range of the weighting factor varies depending on the reference value of the weighting factor, the table in which the reference value of the weighting factor is associated with the minimum value and the maximum value of the selection range of the weighting factor Is prepared and is referred to each time after the weight coefficient selection range is derived. In such a case, a configuration in which the table is loaded into the memory and referred to each time is required, and the hardware scale increases. On the other hand, in the second embodiment, since the range of syntax to be decoded (difference value of weighting factors) can be fixed, the hardware scale is reduced without being restricted by the hardware configuration as described above. Can do.

  Furthermore, in the second embodiment, the difference value of the weighting coefficient whose range is fixed to signed 8-bit precision is decoded, but the difference value of the weighting coefficient takes a value near the center (near 0) of the range. The code length at the time of decoding can be shortened, and the decoding efficiency can be improved. H. In H.264, etc., the weighting coefficient is decoded by a signed exponent Golomb code (se (v)). This code is an effective code for symbols whose decoding value increases exponentially with reference to 0. Generally, the reference value having the highest use frequency is set at the center of the range. In the second embodiment, when a pixel value change between pictures is 0 in a general moving image, the reference value of the weighting factor is used, and prediction from the reference value is also introduced in the prediction of the selection range of the weighting factor. ing. As a result, since the exponent Golomb code and the selection range of the prediction and weighting coefficients match, the code amount reduction effect is high, and the coefficient range is determined around the reference value. The distance from the reference value of the negative value is approximately the same, and there is an effect that decoding can be performed with a shorter code length compared to the conventional method.

(Modification 1 of the first embodiment)
In the first embodiment, the derivation of the selection range of the weighting factor in the encoding apparatus 100 has been described. In the first modification, the derivation of the selection range of the offset in the encoding apparatus 100 will be described.

  As described in Equations (20) to (22), in the YUV color space, the color difference component expresses a color by the amount of deviation from the median value. For this reason, the amount of change from the pixel value change in consideration of the median value using the weighting coefficient can be used as the predicted value. This predicted value indicates the reference value of the offset when the influence of the weighting coefficient is excluded. That is, the index setting unit 108 can derive an offset selection range by assigning a range of values that can be obtained with the predicted value (offset reference value) approximately as the center, and the offset is included in the offset selection range derived. Can be confirmed.

  For example, when the LWD is 2 and the weighting factor is 5, the reference value of the weighting factor is (1 << 2), that is, 4. On the other hand, since the value of the weighting factor is 5, the pixel value changes. Since the color difference signal expresses a color by a deviation amount from the median value, the index setting unit 108 obtains an offset reference value by eliminating the influence of the weighting factor. The offset reference value is formulated by Equation (23).

  Pred = (MED − ((MED * chroma_weight_lx [i] [j]) >> chroma_log2_weight_denom)) (23)

  Pred indicates the reference value of the offset of the color difference signal, MED indicates the median value of the color difference signal (128 when 8 bits), and the term on the right side indicates the amount of deviation from the median due to the influence of the weighting coefficient. Note that Equation (23) corresponds to a value obtained by inverting the sign of the rightmost term of Equation (20). As shown in Equation (23), the reference value for the offset of the color difference signal is determined according to the weight coefficient of the color difference signal and the fixed point precision.

  Note that Equation (23) can be transformed as Equation (24).

  Pred = ((1 << (BitDepth-1))-((chroma_weight_lx [i] [j]) << (BitDepth-1-chroma_log2_weight_denom)) (24)

  Here, BitDepth represents the pixel depth of the color difference signal, and BitDepth is 8 for an 8-bit signal. Since the MED in Expression (23) is a value expressed by a power of 2, if the shift in the right side is rewritten using BitDepth, it can be expressed as Expression (24).

  FIG. 21 is an explanatory diagram illustrating an example of a selection range of offset of the color difference signal according to the first modification. In the example shown in FIG. 21, Pred is arranged so as to be positioned approximately at the center of the selection range, (Pred) − (1 << OR) is the minimum value of the selection range, and (Pred) + (1 << OR ) -1 is the maximum value of the selection range. OR means the bit precision of the offset. For example, H.264 is 8. As shown in FIG. 21, the selection range of the chrominance signal offset is defined within a predetermined bit accuracy with the reference value of the chrominance signal offset as a substantial center. Although detailed description is omitted, the difference value of the offset of the color difference signal to be encoded (the difference value between the offset of the color difference signal and the reference value of the offset of the color difference signal) is defined as a fixed value of the offset bit accuracy. it can. For example, if the accuracy is 8 bits, the difference value of the offset of the color difference signal is an 8-bit fixed value from −128 to 127. For example, if the accuracy is 9 bits, the difference value of the offset of the color difference signal is a 9-bit fixed value from −256 to 255.

  As a result, it is possible to solve the problem that the range of values to be encoded cannot be determined unless the reference value is restored. In the first modification, the example in which the index setting unit 108 derives the selection range of the color difference signal offset has been described. However, the present invention is not limited to this, and the encoding unit 110 may perform the derivation.

  FIG. 22 is a flowchart illustrating an example of the derivation process of the color selection signal offset selection range according to the first modification. Here, the case where the index setting unit 108 performs the process of deriving the offset selection range of the color difference signal will be described. However, as described above, the encoding unit 110 may perform the processing.

  First, the index setting unit 108 derives the fixed point precision LWD of the weighting factor (Step S12). Here, the index setting unit 108 may derive the fixed point precision LWD of the weighting coefficient from the WP parameter information or the index information.

  Subsequently, the index setting unit 108 derives a weighting factor Wxc (S13). Here, the index setting unit 108 may derive the weighting coefficient Wxc from the WP parameter information or from the index information.

  Subsequently, the index setting unit 108 derives the reference value of the offset of the color difference signal from the equation (23) using the fixed point precision LWD and the weighting factor Wxc of the derived weighting factor (Step S14).

  Subsequently, the index setting unit 108 subtracts (1 << OR) from the derived color difference signal offset reference value to derive the minimum value of the color difference signal offset selection range (step S15).

  Subsequently, the index setting unit 108 adds (1 << OR) -1 to the derived color difference signal offset reference value to derive the maximum value of the color difference signal offset selection range (step S16).

  Then, the index setting unit 108 confirms that the offset of the color difference signal is included in the selected selection range of the offset of the color difference signal. When the index setting unit 108 confirms that the color difference signal offset is not included in the color difference signal offset selection range, the index setting unit 108 performs clip processing using the maximum value or the minimum value of the color difference signal offset selection range. May be performed. In this case, the index setting unit 108 clips at the minimum value if the offset of the color difference signal is smaller than the minimum value of the selection range, and clips at the maximum value if the offset of the color difference signal is larger than the maximum value of the selection range. Good. By introducing such clip processing, the encoding target value such as the difference value of the offset of the chrominance signal takes a value within a predetermined bit accuracy without providing a specific range constraint. The circuit scale configuration used by the hardware can be clarified.

  As described with reference to FIG. 18, since the weighting factor value and the fixed-point precision information are encoded before the offset information, the weighting factor value can be derived when the offset reference value is derived. .

  In addition, the selection range of the first offset and the second offset for the color difference signal described in Modification 1 may be applied separately from the selection range of the first weight coefficient and the second weight coefficient described in the first embodiment. . For example, the selection range of the first weighting factor and the second weighting factor is H.264. The selection range of the first offset and the second offset for the color difference signal may be the same as in H.264.

  According to the first modification, since the range of syntax to be encoded (offset difference value) can be fixed, the specification can be simplified as compared with a configuration in which the encoder dynamically changes these ranges. . If the syntax to be decoded is an offset, and the offset selection range varies according to the offset reference value, prepare a table that associates the offset reference value with the minimum and maximum values of the offset selection range, A configuration in which the offset selection range is referred to after being derived or a configuration in which the offset selection range is calculated and derived each time is required. In such a case, a configuration in which the table is loaded into the memory and referred to each time and an arithmetic circuit for calculation are required, and the hardware scale increases. On the other hand, when the range of the syntax (offset difference value) to be encoded is fixed as in the first modification, the hardware scale can be reduced because the hardware configuration is not restricted.

(Modification 1 of 2nd Embodiment)
In the second embodiment, the derivation of the weight coefficient selection range in the decoding device 800 has been described. In the first modification of the second embodiment, the derivation of the offset selection range in the decoding device 800 will be described. In the first modification of the second embodiment, the index setting unit 806 can derive a selection range of offsets by assigning a range of values that can take values with the predicted value (offset reference value) approximately as the center, and the offset is derived. It can be confirmed that it is included in the selected range of offset. Note that the derivation of the offset selection range is the same as that of the first modification of the first embodiment, and thus detailed description thereof is omitted. The offset selection range may be derived not by the index setting unit 806 but by the index reconstruction unit 801B.

  According to the first modification of the second embodiment, since the range of syntax to be encoded (offset difference value) can be a fixed value, compared to a configuration in which the decoder dynamically changes these ranges, Specifications can be simplified. If the syntax to be decoded is an offset, and the offset selection range varies according to the offset reference value, prepare a table that associates the offset reference value with the minimum and maximum values of the offset selection range, A configuration in which the offset selection range is referred to after being derived or a configuration in which the offset selection range is calculated and derived each time is required. In such a case, a configuration in which the table is loaded into the memory and referred to each time and an arithmetic circuit for calculation are required, and the hardware scale increases. On the other hand, when the range of the syntax (offset difference value) to be encoded is fixed as in the first modification, the hardware scale can be reduced because the hardware configuration is not restricted.

(Modification 2 of the first embodiment)
In the first embodiment, the derivation of the selection range of the weighting factor in the encoding device 100 has been described. However, in Modification 2, the selection of the weighting factor is performed when the selection range of the weighting factor in the encoding device 100 is derived. An example of shifting the range will be described.

  In the first embodiment, as described with reference to FIG. 9, the approximate center of the weight coefficient selection range is used as the reference value for the weight coefficient. In addition, as described with reference to FIG. 7 and the like, when there is no average pixel value change between images, the weight coefficient value is expressed as a real value of 1.0, and the range in which the weight coefficient is negative is unidirectional. Not selected for weighted prediction. From this, it can be seen that the selection range of the weight coefficient in actual operation has the highest selection frequency near the reference value, and the negative range is rarely used. Therefore, in the second modification, the index setting unit 108 shifts the weighting factor selection range to the positive side when the weighting factor selection range is derived.

  FIG. 23 is an explanatory diagram illustrating an example of a selection range of weighting factors according to the second modification. In the example shown in FIG. 23, unlike the weight coefficient selection range described in FIG. 9, a new reference value ((1 << LWD) + SHFT) obtained by adding the shift value SHFT to the weight coefficient reference value (1 << LWD). ) Is located at the approximate center of the selection range, and the value obtained by subtracting 128 from this value (−128+ (1 << LWD) + SHFT) is the minimum value of the selection range, and 127 is added to this value. The value (127+ (1 << LWD) + SHFT) is the maximum value of the selection range. At this time, the maximum value may be larger than 255 depending on the value of the fixed-point precision LWD. However, the index setting unit 108 may perform the clipping process with the maximum value of 255, or the SHFT that can be taken for each fixed point. The value of may be changed. In the second modification, the example in which the index setting unit 108 derives the selection range of the weighting coefficient has been described. However, the present invention is not limited to this, and the encoding unit 110 may perform the derivation.

  FIG. 24 is a flowchart illustrating an example of a process for deriving a selection range of weighting factors according to the second modification. Here, the case where the index setting unit 108 performs the process of deriving the selection range of the weighting factor will be described. However, as described above, the encoding unit 110 may perform the processing.

  First, the index setting unit 108 derives the fixed point precision LWD of the weighting factor (Step S22). Here, the index setting unit 108 may derive the fixed point precision LWD of the weighting coefficient from the WP parameter information or the index information.

  Subsequently, the index setting unit 108 derives a reference value ((1 << LWD) + SHFT) of the weighting factor using the derived fixed point precision LWD and the shift value SHFT (step S23).

  Subsequently, the index setting unit 108 subtracts 128 from the derived weight coefficient reference value ((1 << LWD) + SHFT) to derive the minimum value of the weight coefficient selection range (step S24).

  Subsequently, the index setting unit 108 adds 127 to the derived reference value of the weighting factor ((1 << LWD) + SHFT) to derive the maximum value of the selection range of the weighting factor (step S25).

  Then, the index setting unit 108 confirms that the weighting factor is included in the derived weighting factor selection range. If the index setting unit 108 confirms that the weighting factor is not included in the derived weighting factor selection range, the index setting unit 108 may perform clip processing using the maximum value or the minimum value of the weighting factor selection range. Good. In this case, the index setting unit 108 may clip at the minimum value if the weighting factor is smaller than the minimum value of the selection range, and may clip at the maximum value if the weighting factor is larger than the maximum value of the selection range. By introducing such clip processing, the encoding target value such as the difference value of the weighting coefficient takes a value within a predetermined bit accuracy without setting a specific range constraint. The configuration of the circuit scale used by can be clarified.

  As described above, in the second modification, values are assigned to the negative direction and the positive direction with the reference point shifted by a predetermined value in consideration of the change of the weighting factor in the selection range of the weighting factor. Thus, the range of values to be encoded can be fixed.

(Modification 2 of the second embodiment)
In the second embodiment, the derivation of the selection range of the weight coefficient in the decoding device 800 has been described. However, in the second modification of the second embodiment, when the selection range of the weight coefficient in the decoding device 800 is derived, the weight is selected. An example of shifting the coefficient selection range will be described. In the second modification of the second embodiment, the index setting unit 806 shifts the weighting factor selection range to the positive side when the weighting factor selection range is derived. Note that the derivation of the selection range of the weight coefficient is the same as that of the second modification of the first embodiment, and thus detailed description thereof is omitted. Further, the derivation of the weighting factor selection range may be performed by the index reconstruction unit 801B instead of the index setting unit 806.

  As described above, in Modification 2 of the second embodiment, the negative direction and the positive direction are centered on the reference point shifted by a predetermined value in consideration of the change in the weighting factor in the selection range of the weighting factor. By assigning a value to, the range of values to be decoded can be fixed.

(Modification 3 of the first embodiment)
In Modification 3, a derivation method different from the derivation method of the selection range of the weighting factor in the encoding device 100 of the first embodiment will be described.

  In Modification 3, luma_weight_lx [i] in Equations (14) to (16) has a fixed selection range, and delta_luma_weight_lx [i] has a dynamic selection range according to the LWD. Note that the selection range of the weighting factor (luma_weight_lx [i]) in the fifth modification is the same as that in FIG.

  FIG. 25 is an explanatory diagram illustrating an example of the range of difference values of the weighting factors to be encoded in the third modification. In the example shown in FIG. 25, the difference value (delta_luma_weight_lx [i]) of the weight coefficient takes a signed 9-bit value in order to perform addition / subtraction of the signed 8-bit signal. On the other hand, since the reference value of the weighting coefficient takes a value that increases according to the fixed-point precision, the difference value of the weighting coefficient tends to be negative as the value of the fixed-point precision increases.

  FIG. 26 is an explanatory diagram illustrating an example of a relationship between values of syntax elements according to the third modification, and illustrates a relationship between values of luma_log2_weight_denom, default_luma_weight_lx, luma_weight_lx [i], and delta_luma_weight_lx [i]. The delta_luma_weight_lx [i], which is a syntax element encoded by the entropy encoding unit 110A, that is, the range that can be taken by the difference value of the weight coefficient, has a larger value of luma_log2_weight_denom representing fixed-point precision, as shown in FIG. It can be seen that there is a tendency to bias toward the negative side. On the other hand, the value of the decoded weight coefficient (luma_weight_lx [i]) is found to be in a fixed range from −128 to 127.

  As described above, in the third modification, the difference value range to be encoded is determined so that the decoded weight coefficient value has a fixed selection range. The same selection range as H.264 can be determined.

(Modification 3 of 2nd Embodiment)
In Modification 3 of the second embodiment, a derivation method different from the derivation method of the weight coefficient selection range in the decoding device 800 of the second embodiment will be described. However, since the method for deriving the selection range of the weighting factor in the third modification of the second embodiment is the same as that in the third modification of the first embodiment, detailed description thereof is omitted.

  As described above, in the third modification of the second embodiment, even when the prediction method is changed by defining the range of the difference value to be decoded so that the value of the weighting coefficient has a fixed selection range, the The same selection range as H.264 can be determined.

(Modification 4 of the first embodiment)
In Modification 4, an example in which the weighting coefficient selection range is shifted when the weighting coefficient selection range is derived in Modification 3 of the first embodiment will be described.

  In the fourth modification, the index setting unit 108 shifts the range of the difference value of the weighting factor to the positive side, which is substantially equivalent to shifting the selection range of the weighting factor after decoding to the positive side. is there.

  FIG. 27 is an explanatory diagram illustrating an example of a range of difference values of weighting factors according to the fourth modification. Compared to the third modification of the first embodiment, the range of the difference value of the weighting coefficient taking the value in the 8-bit range from −128 to 127 is shifted to the positive side by SHFT.

  FIG. 28 shows the selection range of the weighting factor after decoding of the fourth modification. From FIG. 28, it can be seen that in the fourth modification, the range of the weight coefficient is shifted to the positive side by the shift of the difference value. Even with such a configuration, it is possible to select a reference value when LWD is 7, which could not be selected conventionally.

(Modification 4 of the second embodiment)
In Modification 4 of the second embodiment, an example will be described in which the weighting coefficient selection range is shifted when the weighting coefficient selection range is derived in Modification 3 of the second embodiment. However, since the method for shifting the selection range of the weighting factor in the fourth modification of the second embodiment is the same as that in the fourth modification of the first embodiment, detailed description thereof is omitted. Even with such a configuration, it is possible to select a reference value when LWD is 7, which could not be selected conventionally.

(Modification 5 of the first embodiment)
In Modification 5, an example in which the difference value of the weighting coefficient is subjected to wrap processing in Modification 3 and Modification 4 of the first embodiment will be described.

  As described with reference to FIG. 25, the difference value of the weight coefficient is a signed 9-bit signal (−256 to 126), and is biased toward the negative side when the fixed point precision increases. Since the weight coefficient is generally entropy-encoded using an exponential Golomb code or the like, if the positive / negative balance is biased, the encoding efficiency may decrease. The range of the difference value of the weighting factor varies depending on the fixed-point precision, but the range when the fixed-point precision is fixed is within 8 bits. For example, when the LWD is 7, the range of the difference value is from −256 to −1. When shifting to the 0 reference, an 8-bit value from 0 to 255 is obtained. For this reason, the index reconstruction unit 110B wraps the signed 9 bits into unsigned 8 bits according to the fixed-point precision. In this case, the positive value takes the conventional value, and the negative value is connected to the end of the positive value.

  FIG. 29 is a flowchart illustrating an example of the wrapping process of the difference value of the weighting coefficient according to the fifth modification.

  First, the index reconstruction unit 110B derives the fixed-point precision LWD of the weighting coefficient from the index information (Step S32).

  Subsequently, the index reconstruction unit 110B derives a weighting factor from the index information (Step S33).

  Subsequently, the index reconstruction unit 110B derives the reference value (1 << LWD) of the weighting factor using the derived fixed point precision LWD (step S34).

  Subsequently, the index reconstruction unit 110B derives the difference value of the weighting coefficient from the mathematical formula (14) (step S35).

  Subsequently, the index reconstruction unit 110B performs wrap processing based on the reference value (1 << LWD) of the weighting factor, and connects the negative value behind the maximum positive value without changing the positive value. To generate an unsigned 8-bit code (step S36).

  Then, the code generated by the index reconstruction unit 110B is entropy encoded by the entropy encoding unit 110A.

  As described above, in Modified Example 5, by performing wrap processing on a signed 9-bit value, it becomes possible to always perform encoding with 8 bits without a sign, such as a signed 9-bit exponential Golomb encoding unit. No need to have hardware.

(Modification 5 of the second embodiment)
In Modification 5 of the second embodiment, an example will be described in which the difference value of the weighting coefficient is subjected to wrap processing in Modification 3 and Modification 4 of the second embodiment.

  As described with reference to FIG. 25, the difference value of the weight coefficient is a signed 9-bit signal (−256 to 126), and is biased toward the negative side when the fixed point precision increases. Since the weight coefficient is generally entropy-encoded using an exponential Golomb code or the like, if the positive / negative balance is biased, the encoding efficiency may decrease. The range of the difference value of the weighting factor varies depending on the fixed-point precision, but the range when the fixed-point precision is fixed is within 8 bits. For example, when the LWD is 7, the range of the difference value is from −256 to −1. When shifting to the 0 reference, an 8-bit value from 0 to 255 is obtained. Therefore, the index reconstruction unit 801B wraps the signed 9 bits into unsigned 8 bits according to the fixed point precision. In this case, the positive value takes the conventional value, and the negative value is connected to the end of the positive value.

  FIG. 30 is a flowchart illustrating an example of the restoration process of the difference value of the weighting coefficient according to the fifth modification of the second embodiment.

  First, the entropy decoding unit 801A decodes the encoded data to derive the fixed-point precision LWD of the weighting factor (Step S42).

  Subsequently, the entropy decoding unit 801A next decodes the encoded data to decode an unsigned 8-bit code indicating the difference value of the weighting coefficient (step S43).

  Subsequently, the index reconstruction unit 801B derives the reference value (1 << LWD) of the weighting factor using the derived fixed point precision LWD (step S44).

  Subsequently, the index reconfiguration unit 801B restores the unsigned 8-bit code to a signed 9-bit difference value using the derived reference value (1 << LWD) of the weighting factor (step S45). Here, the difference value is restored by connecting the code that is equal to or higher than the reference value to the negative side while keeping the value below the reference value from the decoded data. The difference value of the restored weighting factor is derived as described above, and the weighting factor is restored using Equation (15).

  As described above, in Modification 5 of the second embodiment, by performing a wrap process on a signed 9-bit value, it is possible to always encode with an unsigned 8 bit, and a signed 9-bit exponential Golomb. There is no need to have hardware such as an encoding unit.

(Modification 6)
In the first to second embodiments, an example is described in which a frame is divided into rectangular blocks of 16 × 16 pixel size and the like, and encoding / decoding is sequentially performed from the upper left block to the lower right side of the screen (FIG. See 2A). However, the encoding order and the decoding order are not limited to this example. For example, encoding and decoding may be performed in order from the lower right to the upper left, or encoding and decoding may be performed so as to draw a spiral from the center of the screen toward the screen end. Furthermore, encoding and decoding may be performed in order from the upper right to the lower left, or encoding and decoding may be performed so as to draw a spiral from the screen end toward the center of the screen. In this case, since the position of the adjacent pixel block that can be referred to changes depending on the encoding order, the position may be changed to a usable position as appropriate.

  In the first to second embodiments described above, the prediction target block size such as a 4 × 4 pixel block, an 8 × 8 pixel block, a 16 × 16 pixel block, and the like has been described as an example, but the prediction target block is a uniform block. It does not have to be a shape. For example, the prediction target block size may be a 16 × 8 pixel block, an 8 × 16 pixel block, an 8 × 4 pixel block, a 4 × 8 pixel block, or the like. Moreover, it is not necessary to unify all the block sizes within one coding tree block, and a plurality of different block sizes may be mixed. When a plurality of different block sizes are mixed in one coding tree block, the code amount for encoding or decoding the division information increases as the number of divisions increases. Therefore, it is desirable to select the block size in consideration of the balance between the code amount of the division information and the quality of the locally decoded image or the decoded image.

  In the first to second embodiments, for the sake of simplification, there is a part in which a comprehensive description is described regarding color signal components without distinguishing between prediction processing and selection range derivation methods for luminance signals and color difference signals. . However, when the prediction process and the selection range derivation method are different between the luminance signal and the color difference signal, the same or different prediction methods may be used. If different prediction methods are used between the luminance signal and the color difference signal, the prediction method selected for the color difference signal can be encoded or decoded in the same manner as the luminance signal.

  In the first to second embodiments, for the sake of simplification, the luminance signal and the weighted motion compensation prediction process for the color difference signal are not distinguished from each other, and the comprehensive explanation is described regarding the color signal component. However, when the weighted motion compensation prediction process is different between the luminance signal and the color difference signal, the same or different weighted motion compensation prediction process may be used. If a weighted motion compensation prediction process that is different between the luminance signal and the color difference signal is used, the weighted motion compensation prediction process selected for the color difference signal can be encoded or decoded in the same manner as the luminance signal.

  In the first and second embodiments, syntax elements not defined in this embodiment can be inserted between the rows of the table shown in the syntax configuration, and other conditional branch descriptions are included. It does not matter. Alternatively, the syntax table can be divided and integrated into a plurality of tables. Moreover, it is not always necessary to use the same term, and it may be arbitrarily changed depending on the form to be used.

  As described above, each embodiment and each modification defines a range of values that can be taken by syntax elements when performing weighted motion compensation prediction, and the range of values that accompanies the range is within a predetermined bit accuracy range. By setting and setting a configuration that gives a short code length to values that are actually frequently used, a highly efficient weighted motion compensation prediction process is performed while solving the problem of encoding redundant information of syntax elements. Realize. Therefore, according to each embodiment and each modification, encoding efficiency is improved, and consequently subjective image quality is also improved.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

  For example, it is possible to provide a program that realizes the processing of each of the above embodiments by storing it in a computer-readable storage medium. The storage medium may be a computer-readable storage medium such as a magnetic disk, optical disk (CD-ROM, CD-R, DVD, etc.), magneto-optical disk (MO, etc.), semiconductor memory, etc. For example, the storage format may be any form.

  Further, the program for realizing the processing of each of the above embodiments may be stored on a computer (server) connected to a network such as the Internet and downloaded to the computer (client) via the network.

DESCRIPTION OF SYMBOLS 100 Coding apparatus 101 Subtraction part 102 Orthogonal transformation part 103 Quantization part 104 Inverse quantization part 105 Inverse orthogonal transformation part 106 Adder part 107 Predictive image generation part 108 Index setting part 109 Motion evaluation part 110 Coding part 110A Entropy coding part 110B Index reconstruction unit 111 Coding control unit 201 Multiple frame motion compensation unit 202 Memory 203 Unidirectional motion compensation unit 204 Prediction parameter control unit 205 Reference image selector 206 Frame memory 207 Reference image control unit 301 Default motion compensation unit 302 Weighted motion Compensation unit 303 WP parameter control unit 304, 305 WP selector 800 decoding device 801 decoding unit 801A entropy decoding unit 801B index reconstruction unit 802 inverse quantization unit 803 inverse orthogonal transform unit 804 Calculation unit 805 predicted image generator 806 index setting unit 807 the decoding control unit

Claims (25)

An electronic device to which encoded data is input,
Decoding the first fixed-point precision of the luminance weighting factor from the encoded data;
Determining the same value as the value obtained by left-shifting “1” by the number of bits determined by the first fixed-point precision as the first reference value of the luminance weighting coefficient;
From the encoded data, a value that is the same as the difference between the luminance weighting factor and the first reference value, and can take a fixed range from −128 to +127 regardless of the first fixed-point precision. Decoding a first difference value of the luminance weighting coefficient;
The luminance weighting factor, which is a value that can take a range from (the first reference value−128) to (the first reference value + 127) by adding the first reference value and the first difference value. Decide
A value obtained by multiplying the median by the color difference weighting factor from the median of the maximum pixel values and shifting right by the number of bits determined by the second fixed point precision of the color difference weighting factor. By subtracting, the second reference value of the color difference offset is calculated,
From the encoded data, the color difference offset is the same value as the difference between the color difference offset and the second reference value, and can take a range from −2 ⁿ to +2 ⁿ −1. 2 Decode the difference value,
By adding the second reference value and the second difference value, the value can take a range from (the second reference value −2 ⁿ ) to (the second reference value +2 ⁿ −1). An electronic apparatus comprising a processing unit that determines an offset of the color difference. The processor is
Decoding a third difference value, which is the same value as the difference between the first fixed-point precision and the second fixed-point precision, from the encoded data;
The electronic device according to claim 1, wherein the second fixed-point precision is determined by adding the first fixed-point precision and the third difference value. The processor is
A value that is the same as a value obtained by left-shifting “1” by the number of bits determined by the second fixed-point precision is determined as a third reference value of the color difference weighting coefficient;
The encoded data is a value that is the same as the difference between the color difference weighting factor and the third reference value, and can take a fixed range from −128 to +127 regardless of the second fixed-point precision. Decoding a fourth difference value of the color difference weighting coefficient;
By adding the third reference value and the fourth difference value, the color difference weighting factor is a value that can take a range from (the second reference value−128) to (the second reference value + 127). The electronic device according to claim 1, wherein the electronic device is determined.   The value obtained by left-shifting “1” by the number of bits determined by the first fixed-point precision is a luminance weight coefficient used when there is no change in luminance between the reference image and the target image. The electronic device according to any one of claims 1 to 3, wherein the electronic device has the same value as the value. The processor is
Decoding the luminance offset from the encoded data;
The predicted luminance value is obtained by at least using the integration of the luminance weight coefficient with the luminance value of the reference image, the right shift of the number of bits determined by the first fixed-point precision, and the addition of the luminance offset. Decide
Prediction value of color difference using at least the integration of the color difference value of the reference image with the weight coefficient of the color difference, the right shift of the number of bits determined by the second fixed-point precision, and the addition of the offset of the color difference The electronic device according to claim 1, wherein the electronic device is determined. The processor is
Decoding transform coefficients from the encoded data;
Using at least an inverse transform to the transform coefficients to determine a prediction error value;
The electronic apparatus according to claim 5, wherein a decoded image is generated using the predicted luminance value, the predicted color difference value, and the prediction error value. The processor is
Decoding a motion vector from the encoded data;
The electronic device according to claim 4, wherein the reference image is generated based on the motion vector. A receiving unit for receiving the encoded data via a communication line;
A buffer for temporarily storing at least a part of the encoded data,
The processing unit reads at least a part of the encoded data from the buffer and sequentially decodes parameters included in the encoded data,
The electronic device according to claim 1, wherein the processing unit is a CPU (Central Processing Unit). A receiving unit for receiving the encoded data via a communication line;
A buffer for temporarily storing at least a part of the encoded data,
The processing unit reads at least a part of the encoded data from the buffer and sequentially decodes parameters included in the encoded data,
The electronic device according to claim 1, wherein the processing unit is a DSP (Digital Signal Processor) or an FPGA (Field Programmable Gate Array). Inputting encoded data; and
Decoding the first fixed point precision of the luminance weighting factor from the encoded data;
Determining a value equal to a value obtained by left-shifting “1” by the number of bits determined by the first fixed-point precision as a first reference value of the luminance weighting factor;
From the encoded data, a value that is the same as the difference between the luminance weighting factor and the first reference value, and can take a fixed range from −128 to +127 regardless of the first fixed-point precision. Decoding a first difference value of the luminance weighting factor;
The luminance weighting factor, which is a value that can take a range from (the first reference value−128) to (the first reference value + 127) by adding the first reference value and the first difference value. A step of determining
A value obtained by multiplying the median by the color difference weighting factor from the median of the maximum pixel values and shifting right by the number of bits determined by the second fixed point precision of the color difference weighting factor. Calculating a second reference value of the color difference offset by subtracting;
From the encoded data, the color difference offset is the same value as the difference between the color difference offset and the second reference value, and can take a range from −2 ⁿ to +2 ⁿ −1. 2 decoding the difference value;
By adding the second reference value and the second difference value, the value can take a range from (the second reference value −2 ⁿ ) to (the second reference value +2 ⁿ −1). Determining an offset of the color difference;
A decoding method including: Decoding a third difference value that is the same value as the difference between the first fixed-point precision and the second fixed-point precision from the encoded data;
Determining the second fixed-point precision by adding the first fixed-point precision and the third difference value;
The decoding method according to claim 10, further comprising: Determining the same value as a value obtained by left-shifting “1” by the number of bits determined by the second fixed-point precision as a third reference value of the color difference weighting factor;
The encoded data is a value that is the same as the difference between the color difference weighting factor and the third reference value, and can take a fixed range from −128 to +127 regardless of the second fixed-point precision. Decoding a fourth difference value of the color difference weighting coefficient;
By adding the third reference value and the fourth difference value, the color difference weighting factor is a value that can take a range from (the second reference value−128) to (the second reference value + 127). A step of determining
The decoding method according to claim 10 or 11, further comprising:   The value obtained by left-shifting “1” by the number of bits determined by the first fixed-point precision is a luminance weight coefficient used when there is no change in luminance between the reference image and the target image. The decoding method according to any one of claims 10 to 12, which is the same value as the value. Decoding a luminance offset from the encoded data;
The predicted luminance value is obtained by at least using the integration of the luminance weight coefficient with the luminance value of the reference image, the right shift of the number of bits determined by the first fixed-point precision, and the addition of the luminance offset. A step to determine;
Prediction value of color difference using at least the integration of the color difference value of the reference image with the weight coefficient of the color difference, the right shift of the number of bits determined by the second fixed-point precision, and the addition of the offset of the color difference A step of determining
The decoding method according to any one of claims 10 to 13, further comprising: Decoding transform coefficients from the encoded data;
Determining a prediction error value using at least an inverse transform to the transform coefficient;
Generating a decoded image using the predicted luminance value, the predicted color difference value, and the prediction error value;
The decoding method according to claim 14, further comprising: Decoding a motion vector from the encoded data;
Generating the reference image based on the motion vector;
The decoding method according to claim 13 or 14, further comprising: A receiving unit receiving the encoded data via a communication line;
A buffer temporarily storing at least a portion of the encoded data;
A processing unit further comprising: reading at least a part of the encoded data from the buffer and sequentially decoding parameters included in the encoded data;
The decoding method according to any one of claims 10 to 16, wherein the processing unit is a CPU (Central Processing Unit). A receiving unit receiving the encoded data via a communication line;
A buffer temporarily storing at least a portion of the encoded data;
A processing unit further comprising: reading at least a part of the encoded data from the buffer and sequentially decoding parameters included in the encoded data;
The decoding method according to any one of claims 10 to 16, wherein the processing unit is a DSP (Digital Signal Processor) or an FPGA (Field Programmable Gate Array). Inputting encoded data; and
Decoding the first fixed point precision of the luminance weighting factor from the encoded data;
Determining a value equal to a value obtained by left-shifting “1” by the number of bits determined by the first fixed-point precision as a first reference value of the luminance weighting factor;
From the encoded data, a value that is the same as the difference between the luminance weighting factor and the first reference value, and can take a fixed range from −128 to +127 regardless of the first fixed-point precision. Decoding a first difference value of the luminance weighting factor;
The luminance weighting factor, which is a value that can take a range from (the first reference value−128) to (the first reference value + 127) by adding the first reference value and the first difference value. A step of determining
A value obtained by multiplying the median by the color difference weighting factor from the median of the maximum pixel values and shifting right by the number of bits determined by the second fixed point precision of the color difference weighting factor. Calculating a second reference value of the color difference offset by subtracting;
From the encoded data, the color difference offset is the same value as the difference between the color difference offset and the second reference value, and can take a range from −2 ⁿ to +2 ⁿ −1. 2 decoding the difference value;
By adding the second reference value and the second difference value, the value can take a range from (the second reference value −2 ⁿ ) to (the second reference value +2 ⁿ −1). Determining an offset of the color difference;
A program that causes a computer to execute. Decoding a third difference value that is the same value as the difference between the first fixed-point precision and the second fixed-point precision from the encoded data;
Determining the second fixed-point precision by adding the first fixed-point precision and the third difference value;
The program according to claim 19, further causing the computer to execute. Determining the same value as a value obtained by left-shifting “1” by the number of bits determined by the second fixed-point precision as a third reference value of the color difference weighting factor;
The encoded data is a value that is the same as the difference between the color difference weighting factor and the third reference value, and can take a fixed range from −128 to +127 regardless of the second fixed-point precision. Decoding a fourth difference value of the color difference weighting coefficient;
By adding the third reference value and the fourth difference value, the color difference weighting factor is a value that can take a range from (the second reference value−128) to (the second reference value + 127). A step of determining
21. The program according to claim 19 or 20, for causing the computer to further execute.   The value obtained by left-shifting “1” by the number of bits determined by the first fixed-point precision is a luminance weight coefficient used when there is no change in luminance between the reference image and the target image. The program according to any one of claims 19 to 21, wherein the program has the same value as the value. Decoding a luminance offset from the encoded data;
The predicted luminance value is obtained by at least using the integration of the luminance weight coefficient with the luminance value of the reference image, the right shift of the number of bits determined by the first fixed-point precision, and the addition of the luminance offset. A step to determine;
Prediction value of color difference using at least the integration of the color difference value of the reference image with the weight coefficient of the color difference, the right shift of the number of bits determined by the second fixed-point precision, and the addition of the offset of the color difference A step of determining
23. The program according to any one of claims 19 to 22, for causing the computer to further execute. Decoding transform coefficients from the encoded data;
Determining a prediction error value using at least an inverse transform to the transform coefficient;
Generating a decoded image using the predicted luminance value, the predicted color difference value, and the prediction error value;
24. The program according to claim 23, further causing the computer to execute. Decoding a motion vector from the encoded data;
Generating the reference image based on the motion vector;
24. The program according to claim 22 or 23, further causing the computer to execute.