JP6679778B2 - Image coding device, image coding method and program, image decoding device, image decoding method and program - Google Patents

Description

  The present invention relates to an image encoding device, an image encoding method and program, an image decoding device, an image decoding method and program, and more particularly to an encoding method / decoding method for a quantization matrix in an image.

As a compression recording method of a moving image, H.264 is used. H.264 / MPEG-4 AVC (hereinafter H.264) is known. (Non-patent document 1)
H. In H.264, each element of the quantization matrix can be changed to an arbitrary value by encoding the scaling_list information. According to the description in Section 7.3.2.1.1.1 of Non-Patent Document 1, each element of the quantization matrix takes an arbitrary value by adding the difference value delta_scale to the immediately preceding element. You can

ITU-TH. H.264 (03/2010) Advanced video coding for general audiovisual services

  H. In H.264, each element in the quantization matrix is scanned in the direction from the element corresponding to the low frequency component in the upper left of the two-dimensional quantization matrix to the element corresponding to the high frequency component in the lower right. For example, when the two-dimensional quantization matrix shown in FIG. 6A is encoded, the one-dimensional quantization matrix shown in FIG. 6B is obtained by using the scanning method called zigzag scanning shown in FIG. Place in a matrix. Then, the difference value between the element to be coded in the matrix and the immediately preceding element is calculated, and the matrix of the difference values shown in FIG. 6D is obtained. Subsequently, the delta_scale which is the above-mentioned difference value is encoded using a method called a signed Exp-Golomb code shown in FIG. For example, if the difference between the immediately preceding element and the element to be encoded is 0, the binary code 1 is encoded, and if the difference is -2, the binary code 00101 is encoded.

  However, H. In the zigzag scanning used in H.264, since each element of the quantization matrix is scanned in the diagonal direction, there is a problem that the code amount of the quantization matrix increases depending on the characteristics of the quantization matrix.

  Therefore, the present invention has been made to solve the above-mentioned problems, and an object thereof is to realize highly efficient quantization matrix encoding / decoding.

  In order to solve the above problems, the image coding apparatus of the present invention has the following configuration. That is, an encoding device that encodes image data, and has n rows and n columns (n is an integer of 4 or more) that can be represented by a two-dimensional array used when quantizing image data to be encoded. An acquisition unit that acquires a difference value between predetermined elements in each element in the quantization matrix, the acquisition unit being the (p + 1) th row and the first column (p is 2 or more and n-1) in the quantization matrix. The difference value between the element corresponding to the following integer) and the element corresponding to the 1st row and the pth column in the quantization matrix is acquired, and the nth row and the qth column (q is 2 or more and n− The difference value between the element corresponding to the (q-1) row and the n-th column in the quantization matrix is acquired and corresponds to the n-th row and the n-th column in the quantization matrix. Elements and the quantization mat Obtaining a difference value between the element corresponding to the (n-1) th row and the n-th column in the box.

  Further, the image decoding device of the present invention has the following configuration. That is, an image decoding device for decoding image data from a bit stream, which is a quantization matrix used when deriving transform coefficients from quantized coefficients, having n rows and n columns (n is an integer of 4 or more) Decoding means for decoding a difference value between predetermined elements in each element included in the quantization matrix from the bit stream, a first difference value among a plurality of difference values decoded by the decoding means, and a predetermined initial value Is added to derive the first element, and the r-th (r is an integer of 2 or more) difference value in the plurality of difference values and the r-1th element are added to obtain the r-th element. By deriving a plurality of elements by deriving a plurality of elements, and the quantization capable of expressing each of the plurality of elements derived by the first deriving means in a two-dimensional array Ma Associating means for associating each element in the lix with the associating means, wherein the associating means associates the first element of the plurality of elements with the first row and the p-th column (p is 2 or more and n− An integer of 1 or less), the second element next to the first element is made to correspond to the element at the (p + 1) th row and the first column in the quantization matrix, and When the third element is made to correspond to the element of the (q-1) th row and the nth column (q is an integer of 2 or more and n-1 or less) in the quantization matrix, a fourth element next to the third element is assigned. The element corresponds to the element in the n-th row and the q-th column in the quantization matrix, and the penultimate element of the plurality of elements corresponds to the element in the (n-1) -th row and the n-th column in the quantization matrix. The last element of the plurality of elements To correspond to the n-th row and n-th column of elements in the Coca matrix.

  According to the present invention, it is possible to reduce the amount of code for quantization matrix coding and perform highly efficient coding / decoding.

FIG. 3 is a block diagram showing the configuration of the image encoding device according to the first embodiment. FIG. 3 is a block diagram showing the configuration of an image decoding device according to the second embodiment. FIG. 3 is a block diagram showing the arrangement of an image coding apparatus according to the third embodiment. Block diagram showing the configuration of an image decoding apparatus according to a fourth embodiment The figure which shows an example of the encoding table of (a), (b) positive / negative symmetry and positive / negative asymmetry. The figure which shows an example of (a)-(e) quantization matrix and difference matrix. Diagram showing an example of encoding a quantization matrix The figure which shows an example of (a), (b) bit stream structure. The flowchart which shows the image coding process in the image coding apparatus which concerns on Embodiment 1. The flowchart which shows the image decoding process in the image decoding apparatus which concerns on Embodiment 2. The flowchart which shows the image coding process in the image coding apparatus which concerns on Embodiment 3. The flowchart which shows the image decoding process in the image decoding apparatus which concerns on Embodiment 4. (A)-(e) The figure which shows the example of the scanning method of the coefficient of a quantization matrix, and the difference calculation method. Block diagram showing a configuration example of computer hardware applicable to the image encoding device and the decoding device of the present invention The figure which shows the example of encoding of the quantization matrix in Embodiment 5 and Embodiment 6. (A)-(c) The figure which shows the example of the scanning method and difference calculation method of the coefficient of a quantization matrix in Embodiment 5 and Embodiment 6. (A)-(c) The figure which shows an example of the quantization matrix and difference matrix in Embodiment 5 and Embodiment 6. (A), (b) The figure which shows the example of the scanning method of the coefficient of the quantization matrix in Embodiment 7 and Embodiment 8. (A), (b) The figure which shows the example of the quantization matrix and difference matrix in Embodiment 7 and Embodiment 8. (A)-(d) The figure which shows the example of the scanning method of the coefficient of the quantization matrix in Embodiment 7 and Embodiment 8.

  Hereinafter, the present invention will be described in detail based on its preferred embodiments with reference to the accompanying drawings. The configurations shown in the following embodiments are merely examples, and the present invention is not limited to the illustrated configurations.

  Further, in the present invention, in the present invention, the scanning method of the two-dimensional matrix shown in FIG. 13B is called horizontal scanning, and the scanning method of the two-dimensional matrix shown in FIG. 13D is vertical scan. I call it.

<Embodiment 1>
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an image coding apparatus according to this embodiment.

  In FIG. 1, reference numeral 101 is a block division unit that divides an input image into a plurality of blocks.

  Reference numeral 102 denotes a prediction unit that predicts each block divided by the block division unit 101 in block units, determines a prediction method, calculates a difference value according to the prediction method, and calculates a prediction error. Intra prediction is performed for intra frames in the case of still images or moving images, and motion compensation prediction is also performed for moving images. Intra prediction is generally realized by selecting an optimal prediction method from among a plurality of reference methods in order to calculate a prediction value from surrounding pixel data.

  Reference numeral 103 denotes a transform unit that performs orthogonal transform on the prediction error of each block, and performs orthogonal transform in units of the input block size or the block size obtained by subdividing the input block size to calculate orthogonal transform coefficients. Hereinafter, the block that performs the orthogonal transform is referred to as a transform block. The orthogonal transform is not particularly limited, but discrete cosine transform, Hadamard transform, or the like may be used. Further, in the present embodiment, for the sake of explanation, it is assumed that the prediction error in block units of 8 × 8 pixels is vertically and horizontally divided into two, and orthogonal transformation is performed in transform block units of 4 × 4 pixels. The shape is not limited to this. Orthogonal transform may be performed using a transform block having the same size as the block, or orthogonal transform may be performed using a transform block obtained by dividing the block into units finer than the vertical and horizontal two-division.

  A quantization matrix holding unit 106 generates and stores a quantization matrix. The generation method of the stored quantization matrix is not particularly limited, but the user may input the quantization matrix or calculate it from the characteristics of the input image, and use the one specified in advance as the initial value. But of course it's good. In the present embodiment, it is assumed that a two-dimensional quantization matrix corresponding to a conversion block of 4 × 4 pixels as shown in FIG. 6A is generated and stored.

  A quantization unit 104 quantizes the orthogonal transform coefficient by the quantization matrix stored in the quantization matrix holding unit 106. A quantization coefficient can be obtained by this quantization.

  Reference numeral 105 denotes a coefficient coding unit that codes the quantized coefficient thus obtained to generate quantized coefficient code data. The encoding method is not particularly limited, but Huffman code, arithmetic code, or the like can be used.

  A quantization matrix scanning unit 109 scans the two-dimensional quantization matrix stored in the quantization matrix holding unit 106 to calculate the difference between each element and arranges the difference in a one-dimensional matrix. In this patent, the difference arranged in this one-dimensional matrix is referred to as a difference matrix.

  A quantization matrix encoding unit 107 encodes the difference matrix that has become a one-dimensional matrix by the quantization matrix scanning unit 109 to generate quantization matrix code data. An integration unit 108 generates header information, a code relating to prediction and conversion, and integrates the quantized coefficient code data generated by the coefficient coding unit 105 and the quantization matrix code data generated by the quantization matrix coding unit 107. It is an encoding unit. The code related to prediction and conversion is, for example, a code such as the selected prediction method or the state of division of a conversion block.

  An image encoding operation in the image encoding device will be described below. In the present embodiment, moving image data is input in frame units, but still image data for one frame may be input. Further, in the present embodiment, for the sake of simplicity of description, only the intra prediction coding process will be described, but the present invention is not limited to this and is also applicable to the inter prediction coding process. In the present embodiment, for the sake of explanation, the block division unit 101 is described as being divided into blocks of 8 × 8 pixels, but the present invention is not limited to this.

  Next, the elements of the quantization matrix are encoded prior to encoding the image. First, the quantization matrix holding unit 106 generates a quantization matrix. The quantization matrix is determined according to the block size to be encoded. The method of determining the elements of the quantization matrix is not particularly limited. For example, a predetermined initial value may be used or may be set individually. Alternatively, the setting may be generated according to the characteristics of the image.

  The quantization matrix holding unit 106 holds the quantization matrix thus generated. FIG. 6A is an example of a quantization matrix corresponding to a conversion block of 4 × 4 pixels. A thick frame 600 represents a quantization matrix. In order to simplify the explanation, it is assumed that the configuration is 16 pixels corresponding to a conversion block of 4 × 4 pixels, and that each square in the thick frame represents an element. In the present embodiment, it is assumed that the quantization matrix shown in FIG. 6A is held in a two-dimensional shape, but the elements in the quantization matrix are not limited to this. For example, when a transform block size of 8 × 8 pixels is also used in addition to this embodiment, another quantization matrix corresponding to the 8 × 8 pixel transform block is held.

  The quantization matrix scanning unit 109 sequentially reads the quantization matrix stored in the two-dimensional shape from the quantization matrix holding unit 106, scans each element, calculates the difference, and arranges it in a one-dimensional matrix. In this embodiment, the vertical scan shown in FIG. 13D is used, and the difference between each element and the immediately preceding element is calculated in the scanning order, but the scanning method and the difference calculation method are limited to this. Not done. The horizontal scan shown in FIG. 13B may be used to calculate the difference between each element and the immediately preceding element in the scanning order. Also, while using the scanning method of FIG. 13B, only the difference of the leftmost element is calculated as the difference from the upper element as shown in FIG. 13C, and otherwise the same as in FIG. 13B. You may calculate the difference with the last element. Further, while using the scanning method of FIG. 13D, as shown in FIG. 13E, only the difference between the elements at the upper end is calculated as the difference from the element on the left, and other than that, FIG. Similarly, the difference from the immediately preceding element may be calculated. In the present embodiment, the two-dimensional shape quantization matrix shown in FIG. 6A is scanned using the vertical scan shown in FIG. 13D, and the difference between each element and the immediately preceding element is calculated. The difference matrix shown in FIG. 6E is generated. Further, the difference value corresponding to the first element of the matrix is calculated as a difference from a predetermined initial value, and the initial value is set to 8 in the present embodiment, but it is not limited to this and may take an arbitrary value. The value of the first element itself may be used.

  The quantization matrix coding unit 107 sequentially reads the difference matrix from the quantization matrix scanning unit 109 and codes the difference matrix to generate quantization matrix code data. In the present embodiment, it is assumed that the coding is performed using the coding table shown in FIG. 5A, but the coding table is not limited to this, and for example, the coding as shown in FIG. You may use a table.

  FIG. 7 shows the encoding matrix of FIG. 5A obtained by calculating the difference matrix of the quantization matrix shown in FIG. 6A using each scanning method of FIGS. 13A and 13D. An example when encoded with is shown. The row of elements in FIG. 7 shows a scan of each element of the quantization matrix of FIG. 6A, and the row of difference values shows the predetermined initial value 8 and the difference value with the immediately preceding element. . The zigzag scan row shows the code when the zigzag scan of the conventional method shown in FIG. 13A is used, and a total of 68 bits are required. On the other hand, the row of the vertical scan shows the code when the vertical scan shown in FIG. 13D is used, and a total of 60 bits are required. In the present embodiment, the same quantization is performed with a smaller code amount. The matrix can be encoded. The coded data of the quantization matrix thus generated is input to the integrated coding unit 108. The integrated coding unit 108 codes the header information necessary for coding the image data, and integrates the coded data of the quantization matrix.

  Subsequently, the image data is encoded. The image data for one frame is input to the block division unit 101 and divided into blocks of 8 × 8 pixels. The divided image data is input to the prediction unit 102.

  The prediction unit 102 performs block-by-block prediction and generates a prediction error. The transform unit 103 divides the prediction error generated by the prediction unit 102 into transform block sizes, performs orthogonal transform, and generates orthogonal transform coefficients. Then, the orthogonal transform coefficient is input to the quantization unit 104. In the present embodiment, it is assumed that the prediction error in block units of 8 × 8 pixels is divided into transform block units of 4 × 4 pixels and orthogonal transform is performed.

  Returning to FIG. 1, the quantization unit 104 quantizes the orthogonal transform coefficient output from the transform unit 103 using the quantization matrix stored in the quantization matrix holding unit 106 to generate a quantized coefficient. The generated quantized coefficient is input to the coefficient coding unit 105.

  The coefficient coding unit 105 codes the quantized coefficient generated by the quantization unit 104, generates quantized coefficient code data, and outputs the quantized coefficient code data to the integrated coding unit 108. The integrated coding unit 108 generates a code relating to prediction and conversion in block units, integrates the coded data in the header, the coded data in block units, and the quantized coefficient coded data generated in the coefficient coding unit 105 into bits. Generate and output a stream.

  FIG. 8A is an example of the bit stream output in the first embodiment. The sequence header includes the coded data of the quantization matrix, and is composed of the coding result of each element. However, the coding position is not limited to this. Of course, the picture header section and other header sections may be encoded. When the quantization matrix is changed in one sequence, it is possible to update it by newly encoding the quantization matrix. At this time, all the quantization matrices may be rewritten, or a part of them may be changed by designating the scanning method of the rewritten quantization matrix and the conversion block size.

  FIG. 9 is a flowchart showing an image coding process in the image coding apparatus according to the first embodiment. First, in step S901, the quantization matrix holding unit 106 generates a quantization matrix.

  In step S902, the quantization matrix scanning unit 109 scans the quantization matrix generated in step S901, calculates the difference between each element, and generates a difference matrix. In the present embodiment, the quantization matrix shown in FIG. 6A is scanned by the scanning method shown in FIG. 13D to generate the difference matrix shown in FIG. 6E. The conversion matrix and scanning method are not limited to these.

  In step S903, the quantization matrix coding unit 107 codes the difference matrix generated in step S902. In the present embodiment, the quantization matrix encoding unit 107 is assumed to encode the difference matrix shown in FIG. 6E using the encoding table shown in FIG. The encoding table according to the present invention is not limited to this.

  In step S904, the integrated encoding unit 108 encodes and outputs the header portion of the bitstream. In step S905, the block division unit 101 divides the input image in frame units into block units. In step S906, the prediction unit 102 performs prediction in block units and generates a prediction error.

  In step S907, the transform unit 103 divides the prediction error generated in step S906 into transform block sizes, performs orthogonal transform, and generates orthogonal transform coefficients. In step S908, the quantization unit 104 quantizes the orthogonal transform coefficient generated in step S907 using the quantization matrix generated in step S901 and stored in the quantization matrix holding unit 106 to obtain the quantized coefficient. To generate.

  In step S909, the coefficient coding unit 105 codes the quantized coefficient generated in step S908 to generate quantized coefficient coded data. In step S910, the image coding apparatus determines whether or not coding of all transform blocks in the block has been completed. If completed, the process proceeds to step S911, and if not completed, the next The process returns to step S907 for the conversion block.

  In step S911, the image coding apparatus determines whether or not coding of all blocks is completed, and if completed, stops all operations and ends the processing. The process returns to step S905 for the block.

  With the above configuration and operation, in particular, the bitstream having a smaller code amount of the quantization matrix can be generated by the processing of unidirectionally scanning the quantization matrix shown in step S902 to calculate the difference matrix.

  It should be noted that, in the present embodiment, a frame using only intra prediction has been described as an example, but it is obvious that a frame that can use inter prediction can also be applied.

  Furthermore, in the present embodiment, the blocks are 8 × 8 pixels and the conversion blocks are 4 × 4 pixels for the sake of explanation, but the present invention is not limited to this. For example, the block size can be changed to 16 × 16 pixels, 32 × 32 pixels, or the like, and the shape of the block is not limited to a square, and may be a rectangle of 16 × 8 pixels.

  Further, although the conversion block size is half the block size in each of the vertical and horizontal directions, it may be the same size or may be a size smaller than the vertical and horizontal halves.

  In the present embodiment, the difference matrix is once generated and encoded. However, the quantization matrix encoding unit 107 directly calculates the difference value from the quantization matrix using a predetermined scanning method and encodes the difference value. It is also possible to adopt a configuration that realizes. In that case, the quantization matrix scanning unit 109 can be omitted.

Further, when different quantization matrices are provided and adapted depending on the scanning method of the orthogonal transform coefficient, the scanning method of the elements of the quantization matrix may be determined according to the scanning method of the orthogonal transform coefficient.
Further, although the case where there is one quantization matrix has been described in the present embodiment, the present invention is not limited to this. For example, in the case of providing different quantization matrices depending on luminance and chromaticity, a common quantization matrix scanning method may be used or they may be provided individually.

<Embodiment 2>
FIG. 2 is a block diagram showing the configuration of the image decoding apparatus according to the second embodiment of the present invention. In this embodiment, decoding of the bitstream generated in the first embodiment will be described.

  In FIG. 2, 201 is a decoding / separating unit that decodes the header information of the input bitstream, separates the necessary code from the bitstream, and outputs the code to the subsequent stage. The decoding / separating unit 201 performs an operation reverse to that of the integrated encoding unit 108 in FIG. A quantization matrix decoding unit 206 decodes the quantization matrix coded data from the header information of the bit stream to generate a difference matrix.

  Reference numeral 208 denotes a quantization matrix reverse scanning unit that reversely scans the difference matrix generated by the quantization matrix decoding unit 206 to reproduce the quantization matrix. The quantization matrix reverse scanning unit 208 performs an operation reverse to that of the quantization matrix scanning unit 109 of FIG. A quantization matrix holding unit 207 stores the quantization matrix reproduced by the quantization matrix reverse scanning unit 208.

  On the other hand, 202 is a coefficient decoding unit that decodes the quantized coefficient code from the code separated by the decoding / separation unit 201 and reproduces the quantized coefficient. Reference numeral 203 denotes an inverse quantization unit that inversely quantizes the quantized coefficient using the quantization matrix stored in the quantization matrix holding unit 207 and reproduces the orthogonal transform coefficient. An inverse transform unit 204 performs an inverse orthogonal transform that is the inverse of the transform unit 103 in FIG. 1 and reproduces a prediction error. Reference numeral 205 denotes a prediction reconstructing unit that reproduces the block image data from the prediction error and the decoded image data.

  The image decoding operation in the image decoding apparatus will be described below. In the present embodiment, the moving image bitstream generated in the first embodiment is input in frame units, but a still image bitstream for one frame may be input. Further, in the present embodiment, only the intra-prediction decoding process is described for ease of description, but the present invention is not limited to this and is also applicable to the inter-prediction decoding process.

  In FIG. 2, the input bit stream for one frame is input to the decoding / separation unit 201, the header information necessary for reproducing the image is decoded, and the code used in the subsequent stage is separated and output. . The quantization matrix coded data included in the header information is input to the quantization matrix decoding unit 206, and the one-dimensional difference matrix is reproduced. At this time, in the present embodiment, it is assumed that the difference value of each element of the quantization matrix is decoded by using the decoding table shown in FIG. Is not limited to this. The reproduced difference matrix is input to the quantization matrix inverse scanning unit 208.

  The difference matrix input to the quantization matrix inverse scanning unit 208 calculates each element of the quantization matrix from each difference value, and inversely scans it to reproduce a two-dimensional quantization matrix. The reproduced quantization matrix is input to and stored in the quantization matrix holding unit 207. Further, among the codes separated by the decoding / separation unit 201, the quantized coefficient code data is input to the coefficient decoding unit 202. First, the coefficient decoding unit 202 decodes the quantized coefficient code data for each transform block, reproduces the quantized coefficient, and outputs the quantized coefficient to the inverse quantization unit 203.

  The inverse quantization unit 203 inputs the quantized coefficient reproduced by the coefficient decoding unit 202 and the quantization matrix stored in the quantization matrix holding unit 207. Then, inverse quantization is performed using the quantization matrix to reproduce the orthogonal transform coefficient and output to the inverse transform unit 204. The inverse transform unit 204 inputs the orthogonal transform coefficient, performs the inverse orthogonal transform that is the inverse of the transform unit 103 in FIG. 1, reproduces the prediction error, and outputs the prediction error to the prediction reconstruction unit 205. The prediction reconstructing unit 205 performs prediction from the decoded surrounding pixel data to the input prediction error, reproduces block-unit image data, and outputs the image data.

  FIG. 10 is a flowchart showing an image decoding process in the image decoding device according to the second embodiment.

  First, in step S1001, the decoding / separation unit 201 decodes the header information and separates the code for output to the subsequent stage. In step S1002, the quantization matrix decoding unit 206 decodes the quantization matrix coded data included in the header information by using the decoding table shown in FIG. 5A, and obtains the difference matrix necessary for the quantization matrix reproduction. To generate. In step S1003, the quantization matrix inverse scanning unit 208 calculates each element of the quantization matrix from the difference matrix generated in step S1002, and inversely scans it to reproduce a two-dimensional quantization matrix.

  In step S1004, the coefficient decoding unit 202 decodes the quantized coefficient code data in units of transform block, and reproduces the quantized coefficient. In step S1005, the dequantization unit 203 dequantizes the quantized coefficient regenerated in step S1004 using the quantized matrix regenerated in step S1003, and regenerates the orthogonal transform coefficient. In step S1006, the inverse transform unit 204 performs inverse orthogonal transform on the orthogonal transform coefficient regenerated in step S1005 to regenerate the prediction error. In step S1007, the image decoding apparatus determines whether or not decoding of all conversion blocks in the block is completed. If completed, the process proceeds to step S1008, and if not completed, the next conversion block is completed. Then, the process returns to step S1004.

  In step S1008, the prediction reconstruction unit 205 performs prediction from the decoded surrounding pixel data, adds the prediction to the prediction error reproduced in step S1006, and reproduces the decoded image of the block. In step S1009, the image decoding apparatus determines whether or not decoding of all blocks is completed, and if completed, stops all operations and ends processing, and if not, completes the next block. Target, the process returns to step S1003.

  With the above configuration and operation, it is possible to perform decoding of the bit stream generated in the first embodiment, in which the code amount of the quantization matrix is smaller, and obtain a reproduced image. In addition, the block size, the conversion block size, and the block shape are not limited to the same as in the first embodiment.

  Further, in the present embodiment, the difference value of each element of the quantization matrix is decoded using the decoding table shown in FIG. 5A, but the decoding table used is not limited to this.

  Further, when the quantization matrix code data is included in the bit stream of one sequence a plurality of times, the quantization matrix can be updated. The decoding / separation unit 201 detects the quantization matrix coded data, and the quantization matrix decoding unit 206 decodes it to generate a difference matrix. The quantization matrix inverse scanning unit 208 inversely scans the generated difference matrix to reproduce the quantization matrix. Then, the reproduced quantization matrix data is replaced with the corresponding quantization matrix in the quantization matrix holding unit 207. At this time, all the quantization matrices may be rewritten, or a part of them may be changed by discriminating the rewritten quantization matrix.

  Further, although the present embodiment has been described by taking the method of accumulating the code data for one frame and then performing the processing as an example, the present invention is not limited to this. For example, an input method such as a block unit or a slice unit composed of a plurality of blocks may be used. Further, the packet may be divided into fixed-length packets or the like instead of the block or the like.

  Further, in the present embodiment, although the configuration is such that the difference matrix is once generated and then the quantization matrix is reproduced, the quantization matrix decoding unit 206 decodes the difference value and then directly quantizes it using a predetermined scanning method. A structure for reproducing the matrix may be adopted. In that case, the quantization matrix reverse scanning unit 208 can be omitted.

  Further, when different quantization matrices are provided and adapted depending on the scanning method of the orthogonal transform coefficient, the scanning method of the elements of the quantization matrix may be determined according to the scanning method of the orthogonal transform coefficient.

<Embodiment 3>
FIG. 3 is a block diagram showing the image coding apparatus of this embodiment. In FIG. 3, parts having the same functions as those in FIG. 1 of the first embodiment are given the same numbers, and description thereof will be omitted.

  A scanning control information generation unit 321 generates quantization matrix scanning method information indicating how to scan each quantization matrix. A scanning method 309 determines a scanning method based on the quantization matrix scanning method information generated by the scanning control information generation unit 321, scans the quantization matrix stored in the quantization matrix holding unit 106, and calculates a difference value. It is a quantization matrix scanning unit that generates a difference matrix.

  Similar to the integrated coding unit 108 in FIG. 1, 308 is an integrated coding unit that generates header information, a code relating to prediction, and conversion. The integrated coding unit 108 is a quantization matrix from the scan control information generation unit 321. The difference is that the scanning method information is input and encoded.

  An image encoding operation in the image encoding device will be described below.

  The coding control information generation unit 321 first generates quantization matrix scanning method information indicating how to scan each quantization matrix to calculate a difference value. In the present embodiment, if the quantization matrix scanning method information is 0, the quantization matrix is scanned using the scanning method shown in FIG. 13A, and each element and the immediately previous element in the scanning order are scanned. It is assumed that the difference value is calculated and the difference matrix is generated. If it is 1, the quantization matrix is scanned using the scanning method shown in FIG. 13B, the difference value between each element and the immediately preceding element in the scanning order is calculated, and the difference matrix is generated. I shall. If it is 2, the quantization matrix is scanned using the scanning method shown in FIG. 13D, the difference value between each element and the immediately preceding element in the scanning order is calculated, and the difference matrix is generated. I shall. The scanning method and the difference calculating method of each element of the quantization matrix are not limited to these, and methods other than those shown in FIGS. 13A, 13B, and 13D may be used, and for example, FIG. The difference calculation method shown in (c) and (e) may be used. The combination of the quantization matrix scanning method information and the quantization matrix scanning method is not limited to this. The method of generating the quantization matrix scanning method information is not particularly limited, but it may be input by the user, may be specified in advance as a fixed value, or may be stored in the quantization matrix holding unit 106. Of course, it may be calculated from the characteristics of the quantization matrix. The generated quantization matrix scanning method information is input to the quantization matrix scanning unit 309 and the integrated coding unit 308.

  The quantization matrix scanning unit 309 scans each quantization matrix stored in the quantization matrix holding unit 106 based on the input quantization matrix scanning method information, calculates a difference value, and generates a difference matrix. And outputs it to the quantization matrix coding unit 107.

  The integrated encoding unit 308 encodes the quantization matrix scanning method information generated by the scanning control information generating unit 321, generates a quantization matrix scanning information code, incorporates it into header information, and outputs it. The encoding method is not particularly limited, but Huffman code, arithmetic code, or the like can be used. FIG. 8B shows an example of a bitstream including the quantization matrix scanning information code. The quantization matrix coding information code may be put in any of headers such as sequences and pictures, but it exists before each quantization matrix code data.

  FIG. 11 is a flowchart showing the image coding process in the image coding apparatus according to the third embodiment. In FIG. 11, parts that perform the same functions as those in FIG. 9 of the first embodiment are given the same numbers and description thereof is omitted.

  In step S1151, the scan control information generation unit 321 determines the quantization matrix scanning method in the subsequent step S1152, and generates quantization matrix scanning method information. In step S1152, the quantization matrix scanning unit 309 scans the quantization matrix generated in step S901 to calculate a difference value based on the quantization matrix scanning method determined in step S1151 to generate a difference matrix. To do. In step S1153, the quantization matrix coding unit 107 codes the difference matrix generated in step S1152. In step S1154, the quantization matrix scanning method information is coded to generate a quantization matrix scanning information code, which is incorporated in the header part and output like the other codes.

  With the above configuration and operation, each quantization matrix is scanned by the optimum scanning method, and it is possible to generate a bitstream with a smaller code amount of the quantization matrix. When different quantization matrices are provided and adapted depending on the scanning method of the orthogonal transformation coefficient, the scanning method of the elements of the quantization matrix may be determined according to the scanning method of the orthogonal transformation coefficient. When another scanning method is used, a flag indicating that and a quantization matrix scanning method information to be used may be encoded.

  Further, although the case where there is one quantization matrix has been described in the present embodiment, the present invention is not limited to this. For example, when providing quantization matrices that differ in luminance and chromaticity, information of a common quantization matrix scanning method may be encoded and used, or may be individually provided and encoded.

  Further, the scanning control information generating unit 321 may generate the scanning method by referring to the quantization matrix generated by the quantization matrix holding unit 106. As described above, it is possible to prepare a plurality of scanning methods in advance and select them to be used as the quantization matrix scanning information, or the order of the scanned elements may be encoded separately. In the case of FIG. 13 (a), the order of 1, 2, 6, 7, 3, 5, 8, 13, 4, 9, 12, 14, 10, 11, 15, 16 is encoded and sent. I don't mind.

<Embodiment 4>
FIG. 4 is a block diagram showing the image decoding apparatus of this embodiment. In FIG. 4, parts that perform the same functions as those in FIG. 2 of the second embodiment are given the same numbers, and description thereof is omitted. In this embodiment, decoding of the bitstream generated in the third embodiment will be described.

  Reference numeral 401 is a decoding / separation unit that decodes the header information of the input bitstream, separates the necessary code from the bitstream, and outputs the code to the subsequent stage. 2 is different from the decoding / separating unit 201 in FIG. 2 in that the quantization matrix scanning information code is separated from the header information of the bit stream and is output to the subsequent stage.

  A scanning control information decoding unit 421 decodes the quantization matrix scanning method information code separated by the decoding / separation unit 401 and reproduces the information of the quantization matrix scanning method. A quantization matrix reverse scanning unit 408 reproduces a quantization matrix by inversely scanning the difference matrix generated by the quantization matrix decoding unit 206 based on the information of the quantization matrix scanning method.

  The image decoding operation in the image decoding apparatus will be described below.

  In FIG. 4, the input bit stream for one frame is input to the decoding / separation unit 401, the header information necessary for reproducing the image is decoded, and the code used in the subsequent stage is separated and output. . The quantization matrix scanning method information code included in the header information is input to the scanning control information decoding unit 421 to reproduce the quantization matrix scanning method information. Then, the reproduced quantization matrix scanning method information is input to the quantization matrix inverse scanning unit 408. On the other hand, the quantization matrix coded data included in the header information is input to the quantization matrix decoding unit 206.

  The quantization matrix decoding unit 206 decodes the quantization matrix code data and reproduces the difference matrix. The reproduced difference matrix is input to the quantization matrix inverse scanning unit 408. The quantization matrix inverse scanning unit 408 inversely scans the difference matrix input from the quantization matrix decoding unit 206 based on the information of the quantization matrix scanning method, adds the difference in element units, and adds the quantization matrix. To play. The reproduced quantization matrix is stored in the quantization matrix holding unit 207.

  FIG. 12 is a flowchart showing an image decoding process in the image decoding device according to the fourth embodiment. Portions having the same functions as those in FIG. 10 of the second embodiment are given the same numbers, and the description thereof will be omitted.

  In step S1001, the decoding / separation unit 401 decodes the header information. In step S1251, the scanning control information decoding unit 421 decodes the quantization matrix scanning method information code included in the header information and reproduces the quantization matrix scanning method information. In step S1253, the quantization matrix inverse scanning unit 408 inversely scans the difference matrix reproduced in step S1252 by using the information about the scanning method of the quantization matrix reproduced in step S1251, and reproduces the quantization matrix. .

  With the above configuration and operation, each quantization matrix generated in the third embodiment is scanned by the optimum scanning method, the bit stream having a smaller code amount of the quantization matrix is decoded, and the reproduced image can be obtained. it can.

  When different quantization matrices are provided and adapted depending on the scanning method of the orthogonal transformation coefficient, the scanning method of the elements of the quantization matrix may be determined according to the scanning method of the orthogonal transformation coefficient. When another scanning method is used, a flag indicating that and a quantization matrix scanning method information to be used may be encoded.

<Fifth Embodiment>
In this embodiment, the image coding apparatus has the same configuration as that of FIG. 1 of the first embodiment. However, the operation of the quantization matrix scanning unit 109 is different. Therefore, the coding other than the quantization matrix scanning unit 109 is similar to that of the first embodiment, and the description thereof is omitted.

  The quantization matrix scanning unit 109 sequentially reads the quantization matrix stored in the two-dimensional shape from the quantization matrix holding unit 106, calculates the difference between each element and the predicted value, scans the calculated difference, and performs primary calculation. Place it in the original matrix. The difference calculation method is different from that of the quantization matrix scanning unit 109 of the first embodiment.

  In the present embodiment, the predicted value is calculated by referring to the left and upper elements as shown in FIG. 16C, and the calculated predicted value is calculated using the horizontal scan shown in FIG. 16A. It shall be scanned and placed in a one-dimensional matrix. Regarding the calculation method of the predicted value, in the present embodiment, the one with the larger value of the left element and the upper element is set as the predicted value, but the present invention is not limited to this, and the smaller value is also the predicted value, and the average value of the two elements May be used as the predicted value. The left element is used as a prediction value when encoding the uppermost element in the matrix, and the upper element is used as the prediction value when encoding the leftmost element. Further, the difference value corresponding to the first element of the matrix is calculated as a difference from a predetermined initial value, and the initial value is set to 8 in the present embodiment, but it is not limited to this and may take an arbitrary value. The value of the first element itself may be used. The scanning method is not limited to the horizontal scanning, and may be the vertical scanning shown in FIG. 16B, as long as it is a unidirectional scanning method.

  The flowchart showing the image coding process in this embodiment is the same as that in FIG. 9 of the first embodiment. However, the operation of step S902 is different. Therefore, the encoding operation other than step S902 is the same as that of the first embodiment, and the description thereof is omitted.

  In step S902, the quantization matrix scanning unit 109 calculates the difference of each element with respect to the quantization matrix generated in step S901 and scans the calculated difference to generate a difference matrix. In this embodiment, the case where the quantization matrix shown in FIG. 17A is generated in step S901 will be described as an example. The generated quantization matrix calculates the two-dimensional difference value matrix shown in FIG. 17B using the maximum value of the upper or left element shown in FIG. 16C as the predicted value. Then, the calculated difference value matrix is scanned by the horizontal scanning shown in FIG. 16A to generate the difference matrix shown in FIG. 17C. Of course, the difference value calculation method is not limited to the maximum value as long as the upper and left elements are used, the minimum value or the average value may be used, and the scanning method is not limited to the horizontal scanning and may be a unidirectional scanning method. .

  FIG. 15 shows a difference value calculated from the quantization matrix shown in FIG. 17A with the maximum value of the upper or left element shown in FIG. An example in which scanning is performed using the scanning method and encoding is performed using the encoding table of FIG. The row of the difference value in FIG. 15 shows a value obtained by horizontally scanning the difference between the predetermined initial value 8 and the maximum value of the left or upper element, and is equivalent to the difference matrix in FIG. 17C. The code line shows the code when the difference value is coded using the coding table of FIG. 5A, and requires a total of 50 bits. This indicates that the quantization matrix can be encoded with a code amount smaller than 68 bits of the conventional method shown in FIG. 7 and 60 bits of the first embodiment.

  With the above configuration and operation, it is possible to generate a bitstream in which the code amount of the quantization matrix is even smaller.

  In the present embodiment, the predicted value is calculated using the left and upper elements, but the present invention is not limited to this, and the upper left element may be used, for example. Also, other elements may be used. At that time, the median value may be used in addition to the maximum value, the minimum value, and the average value, and the present invention is not limited thereto.

<Sixth Embodiment>
In this embodiment, the image decoding apparatus has the same configuration as that of the second embodiment shown in FIG. However, the operation of the quantization matrix reverse scanning unit 208 is different. Therefore, the decoding other than the quantization matrix inverse scanning unit 208 is similar to that of the second embodiment, and the description thereof is omitted. Also, in this embodiment, decoding of the bitstream generated in the sixth embodiment will be described.

  The quantization matrix reverse scanning unit 208 performs the reverse operation of the quantization matrix scanning unit 109 of the sixth embodiment. The difference matrix input to the quantization matrix inverse scanning unit 208 reversely scans each difference value to reproduce a two-dimensional difference value matrix, and then calculates each element of the quantization matrix to perform two-dimensional quantization. Play the matrix. In the present embodiment, the difference matrix is reversely scanned using the horizontal scan shown in FIG. 16A to reproduce a two-dimensional difference value matrix, and the left or upper element is reproduced as shown in FIG. 16C. And each element of the quantization matrix is calculated from the difference value and the two-dimensional quantization matrix is reproduced. The reverse scanning method is not limited to the horizontal scanning, and may be the vertical scanning shown in FIG. 16B, as long as it is a unidirectional scanning method. Regarding the method of calculating each element of the quantization matrix, in the present embodiment, the one with the larger value of the left element and the upper element is the predicted value, and the sum of it and the difference value is each element of the quantization matrix. , But is not limited to this. The smaller value of the left element and the upper element may be used as the predicted value, or the average value of the two elements may be used as the predicted value, and the sum of the predicted value and the difference value may be calculated for each element of the quantization matrix. The value. The left element is used as a prediction value when reproducing the uppermost element in the matrix, and the upper element is used as the prediction value when encoding the leftmost element, and the sum of the difference value is used as the element value. Further, the difference value corresponding to the first element of the matrix is calculated as a difference from a predetermined initial value, and the initial value is set to 8 in the present embodiment, but it is not limited to this and may take an arbitrary value. The value of the first element itself may be used. The scanning method is not limited to the horizontal scanning, and may be the vertical scanning shown in FIG. 16B, as long as it is a unidirectional scanning method.

  The flowchart showing the image decoding process in this embodiment is the same as that in FIG. 10 of the second embodiment. However, the operation of step S1003 is different. Therefore, the decoding operation other than step S1003 is the same as that of the second embodiment, and the description thereof is omitted.

  In step S1003, the quantization matrix inverse scanning unit 208 inversely scans each difference value from the difference matrix generated in step S1002 to reproduce a two-dimensional difference value matrix, and then calculates each element of the quantization matrix. Then, the two-dimensional quantization matrix is reproduced. In the present embodiment, the difference matrix shown in FIG. 17C shown in the sixth embodiment will be described as an example. The difference matrix is inversely scanned by the horizontal scanning shown in FIG. 16A, and the two-dimensional difference value matrix shown in FIG. 17B is calculated. Then, the larger one of the elements above or to the left of each element is set as the prediction value, and the sum of each prediction value and each difference value is set as the value of each element of the quantization matrix. The reverse scanning method is not limited to horizontal scanning and may be a unidirectional reverse scanning method, and the method of reproducing each element of the quantization matrix also uses the smaller one of the left or upper elements or the average value as the prediction value and the difference value. Of course, the sum of can be used as the value of each element.

  With the above configuration and operation, it is possible to perform decoding of the bit stream generated in the sixth embodiment with a smaller code amount of the quantization matrix, and obtain a reproduced image.

  In the present embodiment, the predicted value is calculated using the left and upper elements, but the present invention is not limited to this, and the upper left element may be used, for example. Also, other elements may be used. At that time, the median value may be used in addition to the maximum value, the minimum value, and the average value, and the present invention is not limited thereto.

<Embodiment 7>
In this embodiment, the image coding apparatus has the same configuration as that of FIG. 1 of the first embodiment. However, the operation of the quantization matrix scanning unit 109 is different. Therefore, the coding other than the quantization matrix scanning unit 109 is similar to that of the first embodiment, and the description thereof is omitted.

  The quantization matrix scanning unit 109 sequentially reads the quantization matrix stored in the two-dimensional shape from the quantization matrix holding unit 106, calculates the difference between each element and the predicted value, scans the calculated difference, and performs primary calculation. Place it in the original matrix. The difference calculation method is different from that of the quantization matrix scanning unit 109 of the first embodiment.

  In the present embodiment, the diagonal unidirectional scan as shown in FIG. 18A is used, and the difference from the immediately preceding element is calculated for each element in the scanning order. Not limited to. 18A shown in FIG. 18B may be a diagonal unidirectional scan symmetrical with respect to a diagonal line, and a unidirectional scanning method may be used.

  The flowchart showing the image coding process in this embodiment is the same as that in FIG. 9 of the first embodiment. However, the operation of step S902 is different. Therefore, the encoding operation other than step S902 is the same as that of the first embodiment, and the description thereof is omitted.

  In step S902, the quantization matrix scanning unit 109 scans the quantization matrix generated in step S901, calculates the difference between each element, and generates a difference matrix. In the present embodiment, the quantization matrix shown in FIG. 19A is scanned by the scanning method shown in FIG. 18A to generate the difference matrix shown in FIG. 19B. The conversion matrix and scanning method are not limited to these.

  With the above configuration and operation, the scanning method is shared by the encoding method using the diagonal scanning method shown in FIG. 19A without using the zigzag scanning shown in FIG. Bitstream can be generated with the same efficiency while saving memory.

  Alternatively, as shown in FIGS. 20A to 20D, the quantization matrix may be divided into some small matrices and unidirectional scanning may be performed in the small matrices. By doing so, the scanning method of the 4 × 4 quantization matrix can be applied to a larger quantization matrix, and the memory indicating the order can be omitted.

<Embodiment 8>
In this embodiment, the image decoding apparatus has the same configuration as that of the second embodiment shown in FIG. However, the operation of the quantization matrix reverse scanning unit 208 is different. Therefore, the decoding other than the quantization matrix inverse scanning unit 208 is similar to that of the second embodiment, and the description thereof is omitted. Also, in this embodiment, decoding of the bitstream generated in the seventh embodiment will be described.

  The quantization matrix inverse scanning unit 208 performs an operation reverse to that of the quantization matrix scanning unit 109 of the seventh embodiment. The difference matrix input to the quantization matrix inverse scanning unit 208 calculates each element of the quantization matrix from each difference value, and inversely scans it to reproduce a two-dimensional quantization matrix.

  In the present embodiment, each element of the quantization matrix is calculated from each difference value of the difference matrix, and it is reverse-scanned using the scanning method shown in FIG. 18A to reproduce the two-dimensional quantization matrix. To do. The reverse scanning method is not limited to the method shown in FIG. 16A, and may be a diagonal unidirectional scan symmetrical with respect to FIG. 18A shown in FIG. 18B in a diagonal direction. The reverse scanning method may be used.

  The flowchart showing the image decoding process in this embodiment is the same as that in FIG. 10 of the second embodiment. However, the operation of step S1003 is different. Therefore, the decoding operation other than step S1003 is the same as that of the second embodiment, and the description thereof is omitted.

  In step S1003, the quantization matrix inverse scanning unit 208 calculates each element of the quantization matrix from the difference matrix generated in step S1002, and inversely scans it to reproduce a two-dimensional quantization matrix. In the present embodiment, the respective elements of the quantization matrix are calculated from the difference matrix shown in FIG. 19B used in the seventh embodiment, and the calculated respective elements are calculated by the reverse scanning method shown in FIG. Reverse scanning is performed to reproduce the quantization matrix shown in FIG. The difference matrix and the inverse scanning method used are not limited to these.

  With the above configuration and operation, it is possible to decode the bit stream generated in the seventh embodiment while saving the memory used by sharing the scanning method, and obtain the reproduced image.

<Embodiment 9>
In the above embodiment, each processing unit shown in FIGS. 1 to 4 is described as being configured by hardware. However, the processing performed by each processing unit shown in FIGS. 1 to 4 may be configured by a computer program.

  FIG. 14 is a block diagram showing an example of the hardware configuration of a computer applicable to the image display device according to each of the above embodiments.

  The CPU 1401 controls the entire computer using the computer programs and data stored in the RAM 1402 and the ROM 1403, and also executes the processes described above as the processes performed by the image processing apparatus according to each of the above embodiments. That is, the CPU 1401 functions as each processing unit shown in FIGS.

  The RAM 1402 has an area for temporarily storing a computer program or data loaded from the external storage device 1406, data acquired from the outside via the I / F (interface) 1407, and the like. Further, the RAM 1402 has a work area used when the CPU 1401 executes various processes. That is, the RAM 1402 can be allocated as, for example, a frame memory or can appropriately provide other various areas.

  The ROM 1403 stores setting data of the computer, a boot program, and the like. The operation unit 1404 is composed of a keyboard, a mouse and the like, and can be operated by the user of the computer to input various instructions to the CPU 1401. The output unit 1405 outputs the processing result by the CPU 1401. The output unit 1405 can be configured by a display device such as a liquid crystal display to display the processing result.

  The external storage device 1406 is a large capacity information storage device represented by a hard disk drive device. The external storage device 1406 stores an OS (operating system) and a computer program for causing the CPU 1401 to realize the functions of the units illustrated in FIGS. 1 to 4. Furthermore, each image data to be processed may be stored in the external storage device 1406.

  The computer programs and data stored in the external storage device 1406 are appropriately loaded into the RAM 1402 under the control of the CPU 1401 and are processed by the CPU 1401. The I / F 1407 can be connected to a network such as a LAN or the Internet, and other devices such as a projection device and a display device. The computer acquires and sends various information via the I / F 1407. You can 1408 is a bus connecting the above-mentioned units.

  The operation having the above-described configuration is controlled mainly by the CPU 1401 which is the operation described in the above flowchart.

<Other embodiments>
The object of the present invention is also achieved by supplying a storage medium storing a code of a computer program that realizes the above-described functions to a system, and the system reading and executing the code of the computer program. In this case, the computer program code itself read from the storage medium realizes the functions of the above-described embodiments, and the storage medium storing the computer program code constitutes the present invention. Further, it includes a case where an operating system (OS) running on a computer performs some or all of actual processing based on the instructions of the code of the program, and the processing realizes the above-described functions. .

  Further, it may be realized in the following forms. That is, the computer program code read from the storage medium is written in the memory provided in the function expansion card inserted in the computer or the function expansion unit connected to the computer. Then, based on the instructions of the code of the computer program, a case where the CPU or the like included in the function expansion card or the function expansion unit performs some or all of the actual processing to realize the above-described functions is also included.

  When the present invention is applied to the storage medium, the storage medium stores the code of the computer program corresponding to the above-described flowchart.

Claims (13)

An encoding device for encoding image data,
For each element in a quantization matrix of n rows and n columns (n is an integer of 4 or more) that can be represented by a two-dimensional array used when quantizing image data to be encoded, a difference value between predetermined elements is calculated. Acquisition means to acquire,
Have
The acquisition means is
The difference value between the element corresponding to the (p + 1) th row and the 1st column (p is an integer of 2 or more and n-1 or less) in the quantization matrix and the element corresponding to the 1st row and the pth column of the quantization matrix is calculated. Acquired,
Difference between an element corresponding to the nth row and the qth column (q is an integer of 2 or more and n-1 or less) in the quantization matrix and an element corresponding to the (q-1) th row and the nth column in the quantization matrix Get the value,
An image coding apparatus, wherein a difference value between an element corresponding to the nth row and the nth column in the quantization matrix and an element corresponding to the (n-1) th row and the nth column in the quantization matrix is acquired. . The image coding apparatus according to claim 1, wherein the number of rows in the quantization matrix is equal to the number of columns in the quantization matrix. The image coding apparatus according to claim 1, further comprising: a generation unit that generates header information indicating a size of the quantization matrix. The said acquisition means acquires the difference value of the element corresponding to the 1st row and 1st column in the said quantization matrix, and a predetermined initial value. The claim 1 characterized by the above-mentioned. Image coding device. A prediction unit that performs a prediction process on an image to be encoded in order to generate a prediction error;
Prediction means for orthogonally transforming the prediction error to generate transform coefficients;
Quantizing means for quantizing the transform coefficient using the quantization matrix to generate a quantized coefficient;
Have
The encoding means encodes the quantized coefficient.
The image encoding device according to claim 1, wherein the image encoding device is an image encoding device. An image decoding device for decoding image data from a bitstream,
A quantization matrix used when deriving a transform coefficient from a quantized coefficient, which is a difference value between predetermined elements in each element included in the quantization matrix of n rows and n columns (n is an integer of 4 or more). Decoding means for decoding from the bitstream,
The first difference value of the plurality of difference values decoded by the decoding means and a predetermined initial value are added to derive the first element, and the r-th (r is 2) of the plurality of difference values. First deriving means for deriving a plurality of elements by adding the difference value of (the above integer) and the r-1th element to derive the rth element,
Associating means for associating each of the plurality of elements derived by the first deriving means with each element in the quantization matrix that can be represented by a two-dimensional array,
The associating means is
When the first element of the plurality of elements is made to correspond to the element of the first row and the p-th column (p is an integer of 2 or more and n-1 or less) in the quantization matrix, the next element of the first element is The second element is made to correspond to the element in the (p + 1) th row and first column in the quantization matrix,
When the third element of the plurality of elements is made to correspond to the element of the (q-1) th row and the nth column (q is an integer of 2 or more and n-1 or less) in the quantization matrix, the third element The fourth element following the element is made to correspond to the element in the n-th row and the q-th column in the quantization matrix,
The penultimate element of the plurality of elements corresponds to the element at the (n-1) th row and nth column in the quantization matrix;
The image decoding device, wherein the last element of the plurality of elements is made to correspond to the element in the n-th row and the n-th column in the quantization matrix. The image decoding device according to claim 6, wherein the number of rows in the quantization matrix is equal to the number of columns in the quantization matrix. The image decoding device according to claim 6 or 7, wherein the size of the quantization matrix is based on information included in header information. Second deriving means for deriving the transform coefficient from the quantized coefficient using the quantization matrix,
Third derivation means for deriving a prediction error from the conversion coefficient,
9. An image generation unit that performs prediction based on decoded pixels to derive a predicted image and that generates the predicted image and an image based on the prediction error. The image decoding device according to item 1. A coding method for coding image data, comprising:
For each element in a quantization matrix of n rows and n columns (n is an integer of 4 or more) that can be represented by a two-dimensional array used when quantizing image data to be encoded, a difference value between predetermined elements is calculated. Acquisition process to acquire,
Have
In the acquisition step,
The difference value between the element corresponding to the (p + 1) th row and the 1st column (p is an integer of 2 or more and n-1 or less) in the quantization matrix and the element corresponding to the 1st row and the pth column of the quantization matrix is calculated. Acquired,
Difference between an element corresponding to the nth row and the qth column (q is an integer of 2 or more and n-1 or less) in the quantization matrix and an element corresponding to the (q-1) th row and the nth column in the quantization matrix Get the value,
An image coding method, wherein a difference value between an element corresponding to the nth row and the nth column in the quantization matrix and an element corresponding to the (n-1) th row and the nth column in the quantization matrix is obtained. . An image decoding method for decoding image data from a bitstream,
A quantization matrix used when deriving a transform coefficient from a quantized coefficient, which is a difference value between predetermined elements in each element included in the quantization matrix of n rows and n columns (n is an integer of 4 or more). A decoding step of decoding from the bitstream,
The first difference value of the plurality of difference values decoded by the decoding process is added to a predetermined initial value to derive the first element, and the r-th (r is 2) of the plurality of difference values. A first deriving step of deriving a plurality of elements by deriving the r-th element by adding the difference value of (the above integer) and the (r-1) -th element,
An associating step of associating each of the plurality of elements derived by the first deriving step with each element in the quantization matrix that can be represented by a two-dimensional array,
In the associating step,
When the first element of the plurality of elements is made to correspond to the element of the first row and the p-th column (p is an integer of 2 or more and n-1 or less) in the quantization matrix, the next element of the first element is The second element is made to correspond to the element in the (p + 1) th row and first column in the quantization matrix,
When the third element of the plurality of elements is made to correspond to the element of the (q-1) th row and the nth column (q is an integer of 2 or more and n-1 or less) in the quantization matrix, the third element The fourth element following the element is made to correspond to the element in the n-th row and the q-th column in the quantization matrix,
The penultimate element of the plurality of elements corresponds to the element at the (n-1) th row and nth column in the quantization matrix;
The image decoding method, wherein the last element of the plurality of elements is made to correspond to the element in the n-th row and the n-th column in the quantization matrix.   A program that causes a computer to function as each unit of the image coding apparatus according to claim 1.   A program that causes a computer to function as each unit of the image decoding apparatus according to claim 6.

Images