JP2013223097A - Image decoding device, image decoding method, and image decoding program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem that, to reduce block distortion caused by quantization, it was necessary to set deblocking filtering strength appropriately according to quantization parameters.SOLUTION: A quantization parameter derivation unit 205 decodes quantization parameter information corresponding to a block to be inverse-quantized, and also decodes quantization parameter information corresponding to a block to be decoded without compression to derive block quantization parameters. An inverse quantization and inverse orthogonal transformation unit 209 inverse-quantizes to decode the image signal of a block which was compressed before being encoded on the basis of the quantization parameters. An intra-PCM decoding unit 207 decodes the image signal of a block which was encoded without compression. A deblocking filter unit 211 determines filtering strength by using quantization parameters set for each of blocks on both sides of the block boundary and applies filtering to the decoded image signals of blocks on both sides of the block boundary.

Description

  The present invention relates to an image decoding technique, and more particularly to an image decoding technique using filtering.

  Representative examples of compression encoding systems for moving images include MPEG-2 video, MPEG-4 visual, MPEG-4 AVC / H. There are standards such as H.264. In these standards, a picture is divided into a plurality of rectangular blocks and encoded / decoded in units of blocks. Since encoding is performed by performing intra prediction, inter prediction, orthogonal transformation, and quantization in units of blocks, distortion occurs at the block boundaries. This distortion is called block distortion. Block distortion is a mode such as intra / inter between two blocks on the top and bottom or left and right across the block boundary, images referred to in inter prediction, motion vectors used in inter prediction, quantization parameters at the time of quantization, etc. Due to the difference, distortion occurs. The process of removing or reducing the block distortion on the decoded image is called a deblocking filter method.

ISO / IEC 14496-10 Information technology-Coding of audio-visual objects-Part 10: Advanced Video Coding

  In order to reduce block distortion caused by quantization, it is necessary to appropriately set the strength of deblocking filtering according to the quantization parameter.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide an image encoding technique capable of appropriately setting the filtering strength in order to reduce block distortion and an image decoding technique corresponding thereto. It is in.

  In order to solve the above problem, an image decoding apparatus according to an aspect of the present invention is an image decoding apparatus that decodes an image signal in units of blocks, and decodes quantization parameter information corresponding to a block to be dequantized. In both cases, a quantization parameter deriving unit (205) for decoding quantization parameter information corresponding to a block to be decoded uncompressed and deriving a quantization parameter of the block, and an image signal of the block encoded by compression A first decoding unit (204, 209) that performs inverse quantization on the basis of the quantization parameter, and a second decoding unit (203, 207) that decodes an image signal of an uncompressed block; The filtering strength is determined using the quantization parameter set for each of the blocks on both sides of the block boundary, and the blocks on both sides of the block boundary are determined. Deblocking filter unit for performing a filtering process to the decoded image signal and a (211).

  Another aspect of the present invention is an image decoding method. This method is an image decoding method for decoding an image signal on a block-by-block basis, and decodes quantization parameter information corresponding to a block to be inversely quantized and also a quantization parameter corresponding to a block to be decoded uncompressed. A quantization parameter deriving step for decoding information and deriving a quantization parameter for the block; and a first decoding step for dequantizing and decoding the image signal of the block that has been compressed and encoded based on the quantization parameter A second decoding step for decoding an image signal of a block encoded without compression, and a quantization parameter set for each of the blocks on both sides of the block boundary to determine a filtering strength, A deblocking filter step for filtering the decoded image signals of the blocks on both sides of the boundary; Including.

  It should be noted that any combination of the above-described constituent elements and a conversion of the expression of the present invention between a method, an apparatus, a system, a recording medium, a computer program, etc. are also effective as an aspect of the present invention.

  According to the present invention, the filtering strength can be set appropriately to reduce block distortion.

It is a block diagram which shows the structure of the image coding apparatus of embodiment. It is a block diagram which shows the structure of the image decoding apparatus of embodiment. It is a figure explaining the tree block prescribed | regulated by this Embodiment, and an encoding block. It is a figure explaining the division mode prescribed | regulated by this Embodiment. It is a figure explaining an example of a quantization group block. It is a flowchart explaining the determination of a quantization parameter, and an encoding process procedure. It is a flowchart explaining the encoding process procedure of the intra PCM block of the 1st Example of this Embodiment. It is a flowchart explaining the encoding process procedure of the quantization parameter information and the conversion block in an encoding block. It is a flowchart explaining the decoding / derivation | leading-out process of a quantization parameter. It is a flowchart explaining the decoding process procedure of the intra PCM block of the 1st Example of this Embodiment. It is a flowchart explaining the decoding process procedure of the quantization parameter information and the transform block in an encoding block. FIG. 10 is a flowchart for describing a quantization parameter prediction value derivation processing procedure in steps S1109 in FIG. 6 and S1406 in FIG. 9; It is a figure explaining an example of the pixel of a block boundary. It is a flowchart explaining the process sequence of the deblocking filter part 111 of an image coding apparatus, and the deblocking filter part 211 of an image decoding apparatus. It is a flowchart explaining a deblocking filter process procedure. It is a flowchart explaining the deblocking filter process procedure for every encoding block. It is a figure explaining an example of the vertical boundary and horizontal boundary of the conversion block in an encoding block. It is a figure explaining an example of the vertical boundary and horizontal boundary of the prediction block in an encoding block. It is a flowchart which shows the signal filtering process procedure of step S2204 of FIG. It is a flowchart which shows the filtering process procedure of the block edge of step S2302 of FIG. It is a figure explaining the table | surface which matched the index indexB and the variable (beta). It is a diagram illustrating a table that associates index indexTc and variables t _c. It is a flowchart which shows the filtering process procedure of the signal for every line of step S3114 of FIG. It is a flowchart explaining the encoding process sequence of the intra PCM block of the 2nd Example of this Embodiment. It is a flowchart explaining the decoding process procedure of the intra PCM block of the 2nd Example of this Embodiment.

  In the present embodiment, with regard to video encoding, in particular, encoding is performed by performing intra prediction, inter prediction, orthogonal transform, and quantization, which will be described later, in units of blocks obtained by dividing a picture into rectangles having an arbitrary size and shape. Do.

  First, techniques used in the present embodiment and technical terms are defined.

(About tree blocks and coding blocks)
In the embodiment, as shown in FIG. 3, the picture is equally divided into square units of any same size. This unit is defined as a tree block, and is an encoding / decoding target block in a picture (an encoding target block in the encoding process and a decoding target block in the decoding process. Hereinafter, unless otherwise specified, It is used as a basic unit of address management for specifying. Except for monochrome, the tree block is composed of one luminance signal and two color difference signals. The size of the tree block can be freely set to a power of 2 depending on the picture size and the texture in the picture. The tree block recursively divides the luminance signal and chrominance signal in the tree block as necessary according to the texture in the picture into four parts (vertically and horizontally), if necessary. Small blocks can be made. Each block is defined as a coding block, and is a basic unit of processing when performing coding and decoding. Except for monochrome, the coding block is also composed of one luminance signal and two color difference signals. The maximum size of the coding block is the same as the size of the tree block. An encoded block having the minimum size of the encoded block is called a minimum encoded block, and can be freely set to a power of 2.

  In FIG. 3, the encoding block A is a single encoding block without dividing the tree block. The encoding block B is an encoding block formed by dividing a tree block into four. The coding block C is a coding block obtained by further dividing the block obtained by dividing the tree block into four. The coding block D is a coding block formed by further recursively dividing the block obtained by dividing the tree block into four parts twice, and is a coding block of the minimum size.

  In the description of the embodiment, the color difference format is 4: 2: 0, the size of the tree block is set to 64 × 64 pixels for the luminance signal, and 32 × 32 pixels for the color difference signal, and the minimum coding block size is set. The luminance signal is set to 8 × 8 pixels, and the color difference signal is set to 4 × 4 pixels. In FIG. 3, the size of the coding block A is 64 × 64 pixels for the luminance signal and 32 × 32 pixels for the color difference signal, and the size of the coding block B is 32 × 32 pixels for the luminance signal and 16 × 16 pixels for the color difference signal. Thus, the size of the coding block C is 16 × 16 pixels for the luminance signal and 8 × 8 pixels for the color difference signal, and the size of the coding block D is 8 × 8 pixels for the luminance signal and 4 × 4 pixels for the color difference signal. . When the color difference format is 4: 4: 4, the size of the luminance signal and the color difference signal of each coding block is equal. When the color difference format is 4: 2: 2, the size of the coding block A is 32 × 64 pixels for the color difference signal, the size of the coding block B is 16 × 32 pixels for the color difference signal, and the size of the coding block C is The color difference signal is 8 × 16 pixels, and the size of the coding block D, which is the smallest coding block, is 4 × 8 pixels.

(About prediction mode)
Encoded / decoded in units of coding blocks (used in encoded processing for encoded picture, predicted block, image signal, etc., and used in decoded processing for decoded picture, predicted block, image signal, etc.) (Used in this sense unless otherwise noted.) Intra mode (MODE_INTRA) for encoding within a picture to be encoded / decoded without using a picture, and a decoded picture signal of an encoded / decoded picture The inter mode (MODE_INTER) to be encoded is switched by performing inter prediction with reference to FIG. A mode for identifying the intra mode (MODE_INTRA) and the inter mode (MODE_INTER) is defined as a prediction mode (PredMode). The prediction mode (PredMode) has an intra mode (MODE_INTRA) or an inter mode (MODE_INTER) as a value, and can be selected and encoded.

(About split mode, prediction block, prediction unit)
When performing intra prediction and inter prediction by dividing a picture into blocks, prediction is performed by dividing an encoded block as necessary in order to reduce the unit for switching the method of intra prediction and inter prediction. A mode for identifying the division method of the luminance signal and the color difference signal of the coding block is defined as a division mode (PartMode). Further, a block divided as necessary is defined as a prediction block. As shown in FIG. 4, eight types of partition modes (PartMode) are defined according to the method of dividing the luminance signal of the coding block.

  A division mode (PartMode) that is regarded as one prediction block without dividing the luminance signal of the encoded block shown in FIG. 4A is defined as 2N × 2N division (PART_2Nx2N). The division modes (PartMode) for dividing the luminance signal of the coding block shown in FIGS. 4B, 4C, and 4D into two prediction blocks arranged vertically are 2N × N division (PART_2NxN) and 2N × nU, respectively. It is defined as division (PART_2NxnU) and 2N × nD division (PART_2NxnD). However, 2N × N division (PART_2NxN) is a division mode divided up and down at a ratio of 1: 1, and 2N × nU division (PART_2NxnU) is a division mode divided up and down at a ratio of 1: 3 and 2N × The nD division (PART_2NxnD) is a division mode in which division is performed at a ratio of 3: 1 up and down. The division modes (PartMode) for dividing the luminance signals of the coding blocks shown in FIGS. 4 (e), (f), and (g) into two prediction blocks arranged on the left and right are divided into N × 2N divisions (PART_Nx2N) and nL × 2N, respectively. It is defined as division (PART_nLx2N) and nR × 2N division (PART_nRx2N). However, N × 2N division (PART_Nx2N) is a division mode in which left and right are divided at a ratio of 1: 1, and nL × 2N division (PART_nLx2N) is a division mode in which division is performed at a ratio of 1: 3 in the left and right, and nR × 2N division (PART_nRx2N) is a division mode in which the image is divided in the ratio of 3: 1 to the left and right. The division mode (PartMode) in which the luminance signal of the coding block shown in FIG. 4 (h) is divided into four parts in the vertical and horizontal directions and defined as four prediction blocks is defined as N × N division (PART_NxN).

  Note that the color difference signal is also divided in the same manner as the vertical and horizontal division ratios of the luminance signal for each division mode (PartMode).

  When the prediction mode (PredMode) is inter mode (MODE_INTER), the partition mode (PartMode) is 2N × 2N partition (PART_2Nx2N), 2N × N partition (PART_2NxN), 2N × nU partition (PART_2NxnU), 2N × nD partition (PART_2NxnD) , N × 2N partition (PART_Nx2N), nL × 2N partition (PART_nLx2N), and nR × 2N partition (PART_nRx2N). Only coding block D, which is the smallest coding block, has a partition mode (PartMode) of 2N × 2N partition (PART_2Nx2N), 2N × N partition (PART_2NxN), 2N × nU partition (PART_2NxnU), 2N × nD partition (PART_2NxnD) In addition to N × 2N partition (PART_Nx2N), nL × 2N partition (PART_nLx2N), and nR × 2N partition (PART_nRx2N), N × N partition (PART_NxN) is defined. The reason why N × N division (PART_NxN) is not defined other than the smallest coding block is that, except for the smallest coding block, the coding block can be divided into four to represent a small coding block.

  When the prediction mode (PredMode) is the intra mode (MODE_INTRA), the division mode (PartMode) is 2N × 2N, except for the coding block D (the luminance signal is 8 × 8 pixels in this embodiment) which is the smallest coding block. Only the division (PART_2Nx2N) is defined, and only the coding block D, which is the smallest coding block, defines the N × N division (PART_NxN) in addition to the 2N × 2N division (PART_2Nx2N) as the division mode (PartMode). The reason why N × N division (PART_NxN) is not defined other than the smallest coding block is that, except for the smallest coding block, the coding block can be divided into four to represent a small coding block.

(Intra prediction and intra prediction mode)
In the intra prediction, the pixel value of the processing target block is predicted from the pixel values of the surrounding decoded blocks in the same picture. In the encoding apparatus and decoding apparatus according to the present embodiment, 34 intra prediction modes are selected and intra prediction is performed. The intra prediction mode (intraPredMode) includes vertical prediction (intra prediction mode intraPredMode = 0) that predicts in the vertical direction from the above decoded block, and horizontal prediction (intra prediction mode intraPredMode) that predicts in the horizontal direction from the left decoded block. = 1), average value prediction predicted by calculating an average value from surrounding decoded blocks (intra prediction mode intraPredMode = 2), average value prediction predicted at an angle of 45 degrees diagonally from surrounding decoded blocks In addition to (intraPredMode = 3), 31 types of angle prediction (intra prediction mode intraPredMode = 4... 34) for predicting diagonally at various angles from surrounding decoded blocks are defined.

  Intra prediction modes used when performing prediction from surrounding image signals that have been encoded / decoded in the intra mode (MODE_INTRA) are prepared for the luminance signal and the chrominance signal, respectively. The intra luminance prediction mode and the intra prediction mode for color difference signals are defined as the intra color difference prediction mode. In the encoding and decoding of the intra luminance prediction mode, when it is determined that the encoding side can predict from the intra luminance prediction mode of the surrounding block using the correlation with the intra luminance prediction mode of the surrounding block. If it is determined that it is better to set a different value for the intra luminance prediction mode than the information for identifying the block to be referenced and predict from the intra luminance prediction mode of the neighboring blocks, the intra luminance prediction mode is further set. A mechanism for encoding or decoding the value of is used. By predicting the intra luminance prediction mode of the block to be encoded / decoded from the intra luminance prediction modes of the neighboring blocks, the amount of code to be transmitted can be reduced. On the other hand, in the coding and decoding of the intra chrominance prediction mode, the intra luminance prediction mode is used on the encoding side by utilizing the correlation with the intra luminance prediction mode of the prediction block of the luminance signal at the same position as the prediction block of the chrominance signal. It is better to predict the value of the intra color difference prediction mode from the value of the intra luminance prediction mode and to set a unique value for the intra color difference prediction mode than to predict from the intra luminance prediction mode. When it is determined, a mechanism for encoding or decoding the value of the intra color difference prediction mode is used. By predicting the intra color difference prediction mode from the intra luminance prediction mode, the amount of code to be transmitted can be reduced.

(Intra PCM coding)
In the embodiment, in the intra mode (MODE_INTRA), in addition to intra prediction encoding that performs encoding using intra prediction that performs prediction from an encoded / decoded surrounding image signal, intra prediction, inter prediction, and orthogonal transform Intra-PCM coding that encodes an image signal as it is as a PCM signal without compression without using compression processing such as quantization is defined. In intra-PCM, the code amount becomes a fixed length depending on the size of the encoded block and the bit depth of the PCM signal. In addition, when the code amount becomes large when fine quantization is performed, it may be possible to perform encoding with a small code amount by selecting non-compressed intra PCM encoding. Further, lossless encoding can be handled by performing intra PCM encoding on all image signals. Intra PCM encoding is performed in units of encoded blocks. A coded block coded by intra PCM is defined as an intra PCM block.

(Conversion block)
As in the prior art, in this embodiment, the code amount is reduced by using orthogonal transform such as DCT (Discrete Cosine Transform), DST (Discrete Sine Transform), etc. that transforms a discrete signal into the frequency domain and its inverse transform. . The coding block recursively divides the luminance signal and the color difference signal in the coding block into four as necessary in order to optimize the coding process according to the texture in the picture, and in small block units, Conversion or inverse conversion can be performed. A block to be subjected to orthogonal transformation or inverse orthogonal transformation divided as necessary is defined as a transformation block.

(Quantization parameter)
In the embodiment, the quantization parameter for deriving the size of the quantization step when quantizing the orthogonally transformed coefficient is set and transmitted in units of the following quantization group blocks. By controlling the value of this quantization parameter on the encoding side, the size of the quantization step can be controlled, and the amount of code can be controlled and the subjective image quality can be controlled. By setting the quantization parameter to a small value, the quantization step is reduced and thin quantization is performed. In this case, although a large amount of code is required, image degradation can be suppressed to a low level, and encoding distortion such as block distortion and mosquito distortion is less noticeable. On the other hand, by setting the value of the quantization parameter large, the quantization step becomes large and coarse quantization is performed. In this case, encoding can be performed with a small amount of code, but image degradation becomes high, and encoding distortion such as block distortion and mosquito distortion becomes conspicuous. In the embodiment, the quantization parameter QP _Y can take a value from −QpBdOffset _Y to 51. However, the variable QpBdOffset _Y is a variable set based on the bit depth of the video signal to be encoded, and is derived from the following equation.
QpBdOffset _Y = 6 * bit_depth_luma_minus8
However, the variable bit_depth_luma_minus8 is a syntax element set based on the bit depth of the encoded luminance signal, and has a value obtained by subtracting 8 from the value of the bit depth of the encoded luminance signal. When the bit depth of the luminance signal is 8 bits, the values of bit_depth_luma_minus8 and QpBdOffset _Y are both 0, and when 10 bits, the value of bit_depth_luma_minus8 is 2 and the value of QpBdOffset _Y is 12.
Further, the quantization parameter QP _Y ′ = of the luminance signal actually used in the quantization and inverse quantization, and the quantization parameters QP _Cb ′, QP _Cr ′ of the color difference signals Cb, Cr are derived by the following equations.
QP _Y '= QP _Y + QpBdOffset _Y
QP _Cb '= QP _Y + cb_qp_offset + QpBdOffset _C
QP _Cr '= QP _Y + cr_qp_offset + QpBdOffset _C
However, the variable QpBdOffset _C is a variable set based on the bit depth of the color difference signal to be encoded, and is derived from the following equation.
QpBdOffset _C = 6 * bit_depth_chroma_minus8
The variable bit_depth_chroma_minus8 is a syntax element set based on the bit depth of the color difference signal to be encoded, and has a value obtained by subtracting 8 from the value of the bit depth of the color difference signal to be encoded. When the bit depth of the color difference signal is 8 bits, the values of bit_depth_chroma_minus8 and QpBdOffset _C are both 0, and when 10 bits, the value of bit_depth_luma_minus8 is 2 and the value of QpBdOffset _C is 12. Also, the quantization parameter QP _Y ′ of the luminance signal can take a value from 0 to 51 + QpBdOffset _Y. Further, the quantization parameters QP _Cb ′ and QP _Cr ′ of the color difference signals Cb and Cr can take values from 0 to 51 + QpBdOffset _C.

(Quantization group block)
In the embodiment, a quantization parameter is set for each coding block, and a syntax element cu_qp_delta, which will be described later, indicating information related to the difference of the quantization parameter is transmitted. However, the minimum size of a coding block that transmits a later-described syntax element cu_qp_delta indicating information on the difference between quantization parameters, that is, the minimum size of a coding block for setting a quantization parameter (hereinafter referred to as MinCUDQPSize) is set. For a coding block smaller than the set minimum size MinCUDQPSize, a plurality of coding blocks grouped to the set minimum size share the value of the quantization parameter. A quantization group block, which is a unit for setting a quantization parameter, is defined, and a quantization parameter is set for each quantization group block. The quantization group block is composed of one encoded block having a set minimum size MinCUDQPSize or larger or a plurality of encoded blocks having a set minimum size MinCUDQPSize.

(Position of tree block, coding block, prediction block, transform block)
The position of each block including the tree block, the encoding block, the prediction block, and the transform block of the present embodiment has the position of the pixel of the luminance signal at the upper left of the luminance signal picture as the origin (0, 0). The pixel position of the upper left luminance signal included in each block area is represented by two-dimensional coordinates (x, y). The direction of the coordinate axis is a right direction in the horizontal direction and a downward direction in the vertical direction, respectively, and the unit is one pixel unit of the luminance signal. Of course, the luminance signal and the color difference signal have the same image size (number of pixels) and the color difference format is 4: 4: 4. Of course, the luminance signal and the color difference signal have a different color size format of 4: 4. Even in the case of 2: 0, 4: 2: 2, the position of each block of the color difference signal is represented by the coordinates of the pixel of the luminance signal included in the block area, and the unit is one pixel of the luminance signal. In this way, not only can the position of each block of the color difference signal be specified, but also the relationship between the positions of the luminance signal block and the color difference signal block can be clarified only by comparing the coordinate values. In the embodiment, when the coordinates of the defined luminance signal block and the x component and y component of the coordinate of the color difference signal are the same regardless of the type of the color difference format, the shape and size of the block. Only define these blocks in the same position.

  FIG. 1 is a block diagram illustrating a configuration of an image encoding device according to an embodiment. The image encoding apparatus according to the embodiment includes an image memory 101, a quantization parameter determination unit 102, an intra prediction unit 103, an intra PCM encoding unit 104, an inter prediction unit 105, an encoding method determination unit 106, and a residual signal generation unit. 107, orthogonal transform / quantization unit 108, inverse quantization / inverse orthogonal transform unit 109, decoded image signal superimposing unit 110, deblocking filter unit 111, encoded information storage memory 112, first decoded image memory 113, second A decoded image memory 114, a first encoded bit string generation unit 115, a second encoded bit string generation unit 116, a third encoded bit string generation unit 117, an encoded bit string multiplexing unit 118, and a switch 119.

  The image memory 101 temporarily stores image signals to be encoded supplied in time order. The image signals to be encoded stored in the image memory 101 are rearranged in the encoding order, divided into a plurality of encoding block units in a plurality of combinations according to settings, and further divided into prediction block units. Then, it is supplied to the intra prediction unit 103, the inter prediction unit 105, and the residual signal generation unit 107.

  The quantization parameter determination unit 102 determines a quantization parameter for each quantization group block from the viewpoint of code amount control, adaptive quantization, and the like. The determined quantization parameter is supplied to the encoding method determination unit 106, the orthogonal transform / quantization unit 108, and the inverse quantization / inverse orthogonal transform unit 109, and is also stored in the encoded information storage memory 112.

  The intra prediction unit 103 is a prediction block unit to be encoded from a decoded image signal stored in the first decoded image memory 113 in a prediction block unit corresponding to each division mode (PartMode) in a plurality of encoding block units. Each of the luminance signal and the color difference signal is subjected to intra prediction according to the plurality of intra luminance prediction modes and the intra color difference prediction mode to obtain an intra prediction signal.

  A prediction residual signal is obtained by subtracting an intra prediction signal in units of prediction blocks for each pixel from a signal to be encoded supplied in units of prediction blocks. An evaluation value for evaluating the amount of code and the amount of distortion is calculated using the prediction residual signal, and the most suitable mode from the viewpoint of the amount of code and the amount of distortion among a plurality of intra prediction modes for each prediction block. And the intra prediction information corresponding to the selected intra prediction mode, the intra prediction signal, and the evaluation value of the intra prediction are supplied to the encoding method determination unit 106 as the intra prediction candidates of the prediction block.

  The intra PCM encoding unit 104 arranges PCM signals to be encoded as PCM signals in units of encoding blocks. The code amount of the PCM signal generated by the intra PCM encoding unit 104 is uniquely determined by the size of the encoding block and the bit depth of the signal. Also, since there is no coding deterioration and coding distortion does not occur, the distortion amount is set to zero. These code amount and distortion amount are supplied to the encoding method determining unit 106 together with the PCM signal as evaluation values for intra PCM encoding.

  The inter prediction unit 105 is a unit corresponding to each division mode (PartMode) in a plurality of encoding block units, that is, a prediction block unit, and a plurality of inter prediction units 105 from a decoded image signal stored in the second decoded image memory 114. Each inter prediction according to prediction mode (L0 prediction, L1 prediction, both predictions) and a reference image is performed, and an inter prediction signal is obtained. At that time, a motion vector search is performed, and inter prediction is performed according to the searched motion vector. In the case of bi-prediction, inter-prediction of bi-prediction is performed by averaging or weighting and adding two inter-prediction signals for each pixel. A prediction residual signal is obtained by subtracting the inter prediction signal in units of prediction blocks for each pixel from the signal to be encoded supplied in units of prediction blocks. An evaluation value for evaluating the amount of code and the amount of distortion is calculated using the prediction residual signal, and the most suitable mode from the viewpoint of the amount of code and the amount of distortion among a plurality of inter prediction modes for each prediction block. And the inter prediction information corresponding to the selected inter prediction mode, the inter prediction signal, and the evaluation value of the inter prediction are supplied to the encoding method determination unit 106 as candidates for the inter prediction of the prediction block.

  The encoding method determination unit 106 selects an optimum code based on an intra prediction evaluation value corresponding to intra prediction information selected for each prediction block in a plurality of encoding block units and an inter prediction evaluation value corresponding to inter prediction information. A coding block division method, a prediction mode (PredMode), a division mode (PartMode), whether to encode as a PCM signal, a division method of a transform block, and a coding block division method and a PCM signal according to the determination Intra prediction information including information indicating whether or not to encode, or encoded information including inter prediction information is supplied to the second encoded bit string generation unit 116, and is stored in the encoded information storage memory 112. Intra-prediction or inter-predicted prediction signal according to the determination is performed on the residual signal generation unit 107 and the decoded image signal superimposition unit. 110. Further, the quantization parameter according to the determination and the method of dividing the transform block are supplied to the orthogonal transform / quantization unit 108, the inverse quantization / inverse orthogonal transform unit 109, and the third coded bit string generation unit 117, and the coding is performed. The information is stored in the information storage memory 112. In addition, when intra PCM encoding is selected, the second encoded bit string generation unit 116 is supplied.

  The residual signal generation unit 107 subtracts the intra-predicted or inter-predicted prediction signal for each pixel from the image signal to be encoded when the encoding method determining unit 106 determines to encode by intra-prediction or inter-prediction. The residual signal is generated and supplied to the orthogonal transform / quantization unit 108.

The orthogonal transform / quantization unit 108 performs quantization parameters on the residual signal supplied from the encoding method determination unit 106 when the encoding method determination unit 106 determines to perform encoding by intra prediction or inter prediction. Using the luminance signal quantization parameter QP _Y ′ derived based on the quantization parameter QP _Y supplied from the determination unit 102 and the color difference signal quantization parameters QP _Cb ′ and QP _Cr ′ of the color difference signals Cb and Cr Then, an orthogonal transform and quantization for transforming into a frequency domain such as DCT and DST are performed to generate an orthogonal transform / quantized residual signal, a third encoded bit string generation unit 117, and an inverse quantization / inverse orthogonal This is supplied to the conversion unit 109.

  The first encoded bit string generation unit 115 calculates the value of the syntax element related to the encoding information of the sequence, the picture, and the slice unit according to the semantic rule that defines the meaning of the syntax element and the derivation method. Entropy coding is performed on the value of the syntax element in accordance with syntax rules, such as variable length coding and arithmetic coding, to generate the first coded bit string, and the coded first coded bit string is coded. This is supplied to the bit string multiplexing unit 118. A syntax element used for deriving a minimum size of a coding block that transmits a syntax element cu_qp_delta, which will be described later, indicating information related to a difference between quantization parameters, that is, a minimum size of a quantization group block is also encoded.

  The second encoded bit string generation unit 116 adds, in addition to the encoding block division information and the encoding information for each encoding block, for each tree block in accordance with the semantic rules that define the meaning of the syntax element and the derivation method. Thus, the value of the syntax element related to the encoding information determined by the encoding method determination unit 106 is derived for each prediction block. Specifically, in addition to coding information in units of coding blocks such as a coding block division method, prediction mode (PredMode), and division mode (PartMode), values of syntax elements related to coding information in units of prediction blocks Is derived. When the prediction mode (PredMode) is the intra mode (MODE_INTRA) and the intra prediction encoding, the syntax element pcm_flag indicating whether or not the intra PCM encoding is set to 0, the intra prediction mode including the intra luminance prediction mode and the intra color difference prediction mode In the case of intra PCM coding in intra mode (MODE_INTRA), the syntax element pcm_flag indicating whether or not intra PCM coding is set to 1, and the prediction mode (PredMode) is inter mode (MODE_INTER). In the case of, a value of a syntax element relating to inter prediction information such as an inter prediction mode, information specifying a reference image, and a motion vector is derived. The derived syntax element values are entropy-encoded by variable-length encoding, arithmetic encoding, etc. according to syntax rules to generate a second encoded bit string, and the encoded second encoding The bit string is supplied to the encoded bit string multiplexing unit 118. Further, in the case of intra PCM encoding, quantization parameter information and a PCM signal are encoded. In the case of intra PCM encoding, detailed processing for encoding quantization parameter information will be described later.

  The third encoded bit string generation unit 117 derives and encodes transform block division information and quantization parameter information. An orthogonal transform and quantized residual signal is subjected to entropy coding by variable length coding, arithmetic coding, etc. in accordance with a prescribed syntax rule to generate a third coded bit string, and a third coded bit string Is supplied to the encoded bit string multiplexing unit 118. Detailed processing for deriving and encoding the quantization parameter information will be described later.

  The encoded bit string multiplexing unit 118 generates a bit stream by multiplexing the first encoded bit string, the second encoded bit string, and the third encoded bit string according to a specified syntax rule. Output a bitstream.

The inverse quantization / inverse orthogonal transform unit 109 uses the orthogonal transform / quantized residual signal supplied from the orthogonal transform / quantization unit 108 based on the quantization parameter QP _Y supplied from the quantization parameter determination unit 102. Using the quantization parameter QP _Y ′ of the luminance signal derived in this way and the quantization parameters QP _Cb ′ and QP _Cr ′ of the color difference signals of the color difference signals Cb and Cr, the residual signal by inverse quantization and inverse orthogonal transformation Is supplied to the decoded image signal superimposing unit 110. The decoded image signal superimposing unit 110 includes a prediction signal that has been intra-predicted or inter-predicted according to the determination by the encoding method determining unit 106 and a residual that has been inversely quantized and inversely orthogonal transformed by the inverse quantization / inverse orthogonal transform unit 109 The decoded image is generated by superimposing the signal and stored in the first decoded image memory 113.

  The deblocking filter unit 111 performs a filtering process for reducing block distortion or the like due to encoding on the decoded image stored in the first decoded image memory 113 in accordance with the encoded information stored in the encoded information storage memory 112. And stored in the second decoded image memory 114. Detailed processing of the deblocking filter unit 111 will be described later.

  FIG. 2 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment corresponding to the image encoding apparatus of FIG. The image decoding apparatus according to the embodiment includes an encoded bit string separating unit 201, a first encoded bit string decoding unit 202, a second encoded bit string decoding unit 203, a third encoded bit string decoding unit 204, and a quantization parameter derivation. Unit 205, intra prediction unit 206, intra PCM decoding unit 207, inter prediction unit 208, inverse quantization / inverse orthogonal transform unit 209, decoded image signal superimposing unit 210, deblocking filter unit 211, encoded information storage memory 212, 1 decoded image memory 213, second decoded image memory 214, and switches 215, 216, and 217.

  The bit stream supplied to the encoded bit string separation unit 201 is separated according to a rule of a prescribed syntax, and the first encoded bit string indicating the encoded information in sequence, picture, and slice units is the first encoded bit string decoding The second encoded bit sequence that is supplied to the unit 202 and includes encoding information in units of encoded blocks is supplied to the second encoded bit sequence decoding unit 203, and includes a residual signal that is subjected to orthogonal transform and quantization. Are supplied to the third encoded bit string decoding unit 204.

  The first encoded bit string decoding unit 202 entropy-decodes the supplied first encoded bit string in accordance with the syntax rule, and each value of the syntax element regarding the encoded information in units of sequences, pictures, and slices. Get. The encoding information for each sequence, picture, and slice is calculated from the value of the syntax element for the decoded sequence, picture, and coding information for each slice according to the semantic rules that define the meaning of the syntax element and the derivation method. To do. The first encoded bit string decoding unit 202 is an encoded bit string decoding unit corresponding to the first encoded bit string generation unit 115 on the encoding side, and the sequence encoded by the first encoded bit string generation unit 115, It has a function of returning to each encoded information from an encoded bit string including encoded information in units of pictures and slices. Coding information in units of sequences, pictures, and slices obtained by the first coded bit string decoding unit 202 is supplied to the coded information storage memory 212 and is used in all blocks although not shown. A syntax element used for deriving a minimum size of a coding block that transmits a syntax element cu_qp_delta, which will be described later, indicating information related to a difference between quantization parameters, that is, a minimum size of a quantization group block is also decoded.

  The second encoded bit string decoding unit 203 entropy-decodes the supplied second encoded bit string in accordance with the syntax rule, and for each tree block, the division information of the encoded block, the encoded block, and Each value of the syntax element related to the encoding information of the prediction block unit is obtained. In accordance with the semantic rules that define the meaning of the syntax element and the derivation method, the division information of the supplied encoded block is decoded, and the code is calculated from the value of the syntax element related to the encoded information in the encoded block unit and the predicted block unit. Encoding information for each block and prediction block is calculated. The second encoded bit string decoding unit 203 is an encoded information calculation unit corresponding to the second encoded bit string generation unit 116 on the encoding side, and is encoded by the second encoded bit string generation unit 116. It has the function to return to each encoding information from the 2nd encoding bit sequence containing the encoding information of a block and a prediction block unit. Specifically, the encoding block division method, prediction mode (PredMode), and division mode (PartMode) are decoded from each syntax element obtained by decoding the second encoded bit string according to a prescribed syntax rule. When the prediction mode (PredMode) to be performed is the intra mode (MODE_INTRA), the syntax element pcm_flag indicating whether it is intra PCM encoding is decoded. If pcm_flag is 0, intra prediction encoding is performed, and an intra prediction mode including an intra luminance prediction mode and an intra color difference prediction mode is obtained. If pcm_flag is 1, intra-PCM coding is performed, and quantization parameter information and a PCM signal are obtained. On the other hand, when the prediction mode (PredMode) is the inter mode (MODE_INTER), inter prediction information such as the inter prediction mode, information specifying the reference image, and a motion vector is obtained. When the prediction mode (PredMode) is the intra mode (MODE_INTRA) and pcm_flag is 0, the intra prediction mode 206 including the intra luminance prediction mode and the intra color difference prediction mode is supplied to the intra prediction unit 206 through the switch 215, and the prediction mode (PredMode ) Is in the inter mode (MODE_INTER), the inter prediction mode, the information specifying the reference image, and the inter prediction information such as the motion vector are supplied to the inter prediction unit 208 through the switch 215. In the case of intra PCM encoding, the PCM signal is supplied to the intra PCM decoding unit 207. Detailed processing for decoding and deriving quantization parameter information in the case of intra PCM encoding will be described later.

The third encoded bit string decoding unit 204 decodes the supplied encoded bit string to acquire transform block division information, quantization parameter information, and an orthogonal transform / quantized residual signal, and performs orthogonal transform / quantization. The converted residual signal is supplied to the inverse quantization / inverse orthogonal transform unit 209. Further, a syntax element cu_qp_delta, which will be described later, indicating information on the difference between the quantization parameters is supplied to the quantization parameter deriving unit 205. The quantization parameter deriving unit 205 obtains a quantization parameter QP _Y from a syntax element cu_qp_delta (to be described later) indicating information on the difference between the supplied quantization parameters and the encoded information obtained by the first encoded bit string decoding unit 202. Derived and supplied to the inverse quantization / inverse orthogonal transform unit 209 and stored in the encoded information storage memory 212. Detailed processing for decoding and deriving the quantization parameter information will be described later.

  The intra prediction unit 206 predicts a predicted image by intra prediction from the decoded peripheral blocks stored in the first decoded image memory 213 according to the intra prediction mode including the supplied intra luminance prediction mode and the intra color difference prediction mode. A signal is generated, and the predicted image signal is supplied to the decoded image signal superimposing unit 210 via the switch 216. Further, in the present embodiment, when the value of the intra color difference prediction mode is predicted from the value of the intra luminance prediction mode, the method for deriving the intra color difference prediction mode differs depending on the color difference format. In this case, intra prediction is performed using an intra prediction mode derived by a different method depending on the color difference format.

  The intra PCM decoding unit 207 performs a decoding process on the input signal to obtain a PCM signal in units of encoded blocks. The obtained encoded block unit PCM signal is stored in the first decoded image memory 213 through the switch 217.

  The inter prediction unit 208 performs motion compensation from the decoded reference picture stored in the second decoded image memory 214 by using the supplied inter prediction mode, information for specifying the reference picture, and inter prediction information such as a motion vector. A predicted image signal is generated by inter prediction using, and the predicted image signal is supplied to the decoded image signal superimposing unit 210 via the switch 216. In the case of bi-prediction, the two motion-compensated prediction image signals of L0 prediction and L1 prediction are adaptively multiplied and weighted to generate a final prediction image signal.

The inverse quantization / inverse orthogonal transform unit 209 includes a luminance parameter quantization parameter QP _Y ′ derived based on the quantization parameter QP _Y supplied from the quantization parameter derivation unit 205, and color differences between the color difference signals Cb and Cr. Using the signal quantization parameters QP _Cb ′ and QP _Cr ′, inverse orthogonal transform and inverse quantization are performed on the orthogonal transform / quantized residual signal decoded by the third coded bit string decoding unit 204. To obtain a residual signal subjected to inverse orthogonal transform and inverse quantization.

  The decoded image signal superimposing unit 210 includes a prediction image signal predicted by the intra prediction unit 206 or the inter prediction unit 208, and a residual signal that has been inversely orthogonal transformed / inversely quantized by the inverse quantization / inverse orthogonal transform unit 209. Is decoded, and the decoded image signal is decoded and stored in the first decoded image memory 213 via the switch 217.

  The deblocking filter unit 211 performs a filtering process for reducing block distortion or the like due to encoding on the decoded image stored in the first decoded image memory 213, and stores it in the second decoded image memory 214. The decoded image signals stored in the second decoded image memory 214 are output in the output order.

  Next, the quantization group block which is one of the points of the embodiment will be described in detail.

As described above, in the present embodiment, a quantization parameter is set for each coding block, and a syntax element cu_qp_delta, which will be described later, indicating information related to the difference between the quantization parameters is transmitted. However, the minimum size of a coding block (hereinafter referred to as MinCUDQPSize) that transmits a syntax element cu_qp_delta, which will be described later, indicating information related to the quantization parameter difference, that is, the minimum size of the coding block for setting the quantization parameter is set. For a coding block smaller than the set minimum size, the value of the quantization parameter is shared by a plurality of coding blocks grouped into the set minimum size MinCUDQPSize. A quantization group block, which is a unit for setting a quantization parameter, is defined, and a quantization parameter is set for each quantization group block. The quantization group block is composed of one encoded block larger than the set minimum size or a plurality of encoded blocks smaller than the set minimum size MinCUDQPSize. However, when the set minimum size MinCUDQPSize is smaller than the size of the target coding block, the size of the quantization group block corresponding to the target coding block is the same size as the coding block. When the set minimum size MinCUDQPSize is smaller than the size of the target coding block, the size of the quantization group block corresponding to the target coding block is the same size as the minimum size MinCUDQPSize. In the present embodiment, the minimum size of a coding block that transmits a later-described syntax element cu_qp_delta that indicates information related to a difference in quantization parameter is the same as the minimum size of the quantization group block. The minimum size of MinCUDQPSize. FIG. 5 is a diagram for explaining an example of the quantization group block defined in the present embodiment. In FIG. 5, the blocks A, B ₀ , B ₁ , B ₂ , C ₀ , C ₁ , C ₂ and D ₀ , D ₁ , D ₂ , D ₃ drawn with solid lines are coding blocks, The drawn block is a block having the same size as the minimum size MinCUDQPSize of the quantization group block. In the example shown in FIG. 5, the size of the tree block is set to 64 × 64 pixels for luminance signals and 32 × 32 pixels for color difference signals. The encoding block A is a single encoding block without dividing the tree block, and the size of the encoding block A is 64 × 64 pixels for the luminance signal and 32 × 32 pixels for the color difference signal. The encoding blocks B ₀ , B ₁ , B ₂ are encoding blocks formed by dividing the tree block into four, and the sizes of the encoding blocks B ₀ , B ₁ , B ₂ are 32 × 32 pixels as luminance signals, The color difference signal is 16 × 16 pixels. The coding blocks C ₀ , C ₁ , C ₂ are coding blocks obtained by further dividing the block obtained by dividing the tree block into four, and the sizes of the coding blocks C ₀ , C ₁ , C ₂ Is 16 × 16 pixels for the luminance signal and 8 × 8 pixels for the color difference signal. The coding blocks D ₀ , D ₁ , D ₂ , and D ₃ are coding blocks that are obtained by further dividing the block obtained by dividing the tree block into four parts and dividing the block into four hierarchically, and the coding block D _0. , D ₁ , D ₂ , and D ₃ have a luminance signal of 8 × 8 pixels and a color difference signal of 4 × 4 pixels. In the example shown in FIG. 5, the minimum size MinCUDQPSize of the quantization group block is set to 16 (when the color difference format is 4: 2: 0, the luminance signal is 16 × 16 pixels and the color difference signal is 8 × 8 pixels).

  In the present embodiment, the quantization parameters of the coding block and the transform block existing at the same position as the quantization group block are the same. The minimum quantization group block size MinCUDQPSize may be set to be the same as or smaller than the size of the tree block. Further, the size of the quantization group block may be the same as the size of the coding block, or may be larger than the size of the coding block.

When the set minimum size of the quantization group block is larger than the size of the corresponding encoding block, the quantization group block includes a plurality of encoding blocks, but the quantization parameters of these encoding blocks are the same. Value. In the example shown in FIG. 5, the quantization parameters of the four encoding blocks D ₀ , D ₁ , D ₂ , and D ₃ smaller than the quantization group block at the lower right are set to the same value.

When the minimum size of the quantization group block is smaller than the size of the corresponding coding block, the size of the quantization group block is the same as the size of the corresponding coding block. In the example shown in FIG. 5, the size of the quantization group block corresponding to the coding block A is 64 × 64 pixels which is the same as the size of the coding block A, and a common quantization parameter is used in the coding block A. . The size of the quantization group blocks corresponding to the coding block B ₀ are the same 32 × 32 pixels and the size of the encoded block B _0, using a common quantization parameter in encoding blocks B within _0.

  In the present embodiment, quantization is not performed in intra-PCM coding, but quantization parameters are used in the deblocking filter. Therefore, a quantization parameter is also set for the intra PCM coding block. Similarly to the quantization parameter of a coding block that is not intra-PCM coding, the quantization parameter of a quantization group block that includes an intra-PCM coding block is regarded as the quantization parameter of the intra-PCM coding block.

  Next, quantization parameter determination / encoding / decoding / derivation according to the first example of the embodiment will be described.

  First, the relationship between the size of the intra PCM block and the size of the quantization group block and the value of the quantization parameter in the present embodiment will be described.

In the present embodiment, in order to reduce the block distortion by setting the filtering strength of the deblocking filter more appropriately, the value of the quantization parameter QP _Y is encoded even when the encoding block is intra PCM encoding. / Decrypt. In this case, the value of the quantization parameter QP _Y to be encoded / decoded is stored in the encoding information storage memory 112 of the image encoding device or the encoding information storage memory 212 of the image decoding device.

When the target coding block is intra PCM coding and the size of the coding block is equal to or larger than the minimum size of the quantization group block, the quantum derived by the quantization parameter determination unit 102 of the image coding apparatus for each intra PCM block parameter QP _Y syntax elements cu_qp_delta with information about the difference between the quantization parameter and encoding, the quantization parameter determination unit 102 quantization parameter QP _Y derived by the image encoding apparatus of the image coding apparatus It is stored in the encoded information storage memory 112 or the encoded information storage memory 212 of the image decoding apparatus.

If the target coding block is intra PCM coding and the size of the coding block is less than the minimum size of the quantization group block, the quantization parameter is not coded in the corresponding quantization group block. However, the encoding process of the quantization parameter information of the quantization parameter QP _Y derived by the quantization parameter determination unit 102 of the image encoding device is omitted. When the coding block is intra PCM coding and the size of the coding block is less than the minimum size of the quantization group block, the intra PCM block does not need to encode the quantization parameter information, and the intra block within the same quantization group block is used. This is because, when a transform block included in a coded block that is not PCM coded is a non-zero coefficient, a syntax element cu_qp_delta indicating information on a difference between quantization parameters is coded.

The target coding block is not intra PCM coding, the transform block in the coding block includes a non-zero coefficient (non-zero coefficient), and the quantization parameter is encoded in the corresponding quantization group block If not, the syntax element cu_qp_delta indicating information on the difference between the quantization parameters is encoded, and the value of the quantization parameter QP _Y derived by the quantization parameter determination unit 102 of the image encoding device is the code of the image encoding device. Stored in the encoded information storage memory 112 or the encoded information storage memory 212 of the image decoding apparatus.

  Next, the quantization parameter determination and encoding method will be described. FIG. 6 is a flowchart for explaining the quantization parameter determination and encoding processing procedure. These processes are performed by the quantization parameter determination unit 102, the second encoded bit string generation unit 116, and the third encoded bit string generation unit 117 of the image encoding device. The processing from step S1102 to S1118 is performed for each slice in the picture (steps S1101 to S1119). Further, the processing from step S1103 to S1117 is performed for each coding tree block in the slice (steps S1102 to S1118).

First, the quantization parameter QP _Y quantization is determined for each quantization group block in the coding tree block (steps S1103 to S1105). These processes are performed by the quantization parameter determination unit 102 of the image encoding device. In the quantization parameter QP _Y code amount control, when decreasing the amount of generated code sets a greater value to the quantization parameter QP _Y, when increasing the amount of generated code, the smaller value to the quantization parameter QP _Y Set. In the case of adaptive quantization for adjusting subjective image quality, a quantization group block that tends to be noticeable in coding deterioration is set to a large value for the quantization parameter QP _Y, and a quantization group block that is noticeable in coding deterioration is quantized. A small value is set in the parameter QP _Y.

Subsequently, the quantization parameter QP _Y of the quantization group block is encoded by performing the processing from step S1107 to S1116 for each encoding block in the encoding tree block (steps S1106 to S1117). These processes are performed by the second encoded bit string generation unit 116 and the third encoded bit string generation unit 117 in the image encoding device. When the quantization parameter QP _Y is encoded, a syntax element cu_qp_delta indicating information on the difference between the quantization parameters is derived, and the derived syntax element cu_qp_delta is encoded.

First, when the target coding block is the head of the quantization group block in the coding order (YES in step S1107), the syntax element corresponding to the QP _Y of the quantization group block corresponding to the target coding block A variable IsCuQpDeltaCoded indicating whether or not cu_qp_delta has been encoded is set to 0 (step S1108), and a quantization parameter prediction value QP _PRED is derived (step S1109). The processing in step S1109 will be described in detail with reference to the flowchart in FIG.

Subsequently, the quantization parameter prediction value QP _PRED derived in step S1109 is used as the quantization group block QP _Y in the coding information storage memory 112, and the quantization parameter of the previous quantization group block in the coding / decoding order. Set to QP _PREV (step S1110). When setting the coding information storage memory 112, if the size of the coding block is larger than the size of the quantization group block, the value of the predicted value QP _PRED of the same quantization parameter is included in the coding block. Set to the quantization parameter QP _Y of all quantization group blocks.

  On the other hand, when the target encoding block is not the head of the quantization group block in the encoding order (NO in step S1107), the process proceeds to step S1111.

  Subsequently, in steps S1111 to S1114, processing related to intra PCM coding is performed. The prediction mode (PredMode) is the intra mode (MODE_INTRA), the partition mode (PartMode) is 2N × 2N partition (PART_2Nx2N), and the size of the coding block is set to be larger than the minimum intra PCM block size. If it is equal to or smaller than the size of the maximum intra PCM block (YES in step S1111), the process proceeds to step S1112; otherwise (NO in step S1111), intra PCM encoding is not performed, and steps S1112 to S1114 are skipped. The process proceeds to S1115.

  Subsequently, in step S1112, a syntax element pcm_flag indicating whether the encoded block is encoded as intra PCM encoding is encoded (step S1112). When encoding an encoded block as an intra PCM block, 1 is set to pcm_flag, and when not encoding an encoded block as an intra PCM block, 0 is set to pcm_flag. When the encoded block is encoded as an intra PCM block (YES in step S1113), the intra PCM block is encoded (step S1114), and when the encoded block is not encoded as an intra PCM block (NO in step S1113). ), The process proceeds to step S1115. The intra PCM block encoding process in step S1114 will be described in detail with reference to the flowchart of FIG.

  FIG. 7 is a flowchart for explaining an intra-PCM block encoding process procedure according to the first example of the present embodiment. These processes are performed by the intra PCM encoding unit 104 and the second encoded bit string generation unit 116 of the image encoding device.

  First, 1 is subtracted from the number NumPCMBlock in which intra PCM blocks having the same size as the encoding block to be encoded (encoded blocks encoded by intra PCM) are consecutive, and the same size as the encoding block to be encoded The value of the syntax element num_subsequent_pcm indicating the number that follows the encoded block encoded by the intra PCM is derived (step S1201), and the syntax element num_subsequent_pcm is entropy encoded (step S1202). Note that the value of the syntax element num_subsequent_pcm ranges from 0 to 3.

Subsequently, when the set minimum size of the quantization group block is equal to or smaller than the size of the encoded block (YES in step S1203), the processes of steps S1205 and S1206 are performed for each intra PCM block to be encoded (steps S1204 to S1204). Step S1207). Quantization parameter QP Y value of the quantization parameter QP _Y of set quantization group block at step S1104 the quantization group blocks of encoded information storage memory 112 of FIG. _6, and the quantum of the immediately preceding the encoding / decoding order The quantization parameter QP _PREV of the quantization group block is set (step S1205). The quantization parameter QP _Y of the quantization group block set in the encoding information storage memory 112 in step S1110 of FIG. 6 and the quantization parameter QP _PREV of the previous quantization group block in the encoding / decoding order are overwritten. . Further, the encoding parameter QP _Y is encoded (step S1206). For the encoding process of the encoding parameters QP _Y in step S1206, the encoding parameter QP _Y may be encoded with a fixed length. This is because the IPCM block encoding / decoding is used for the purpose of an emergency mode that guarantees a predetermined unit code amount. Note that the syntax element cu_qp_delta can be derived and the syntax element cu_qp_delta can be entropy-coded as in steps S1305 to S1306 in FIG.

  On the other hand, if the set minimum size of the quantization group block is larger than the size of the encoded block (NO in step S1203), the process of steps S1204 to S1207 is skipped and the process proceeds to step S1212.

  Subsequently, the encoded bit string is aligned with a byte boundary (step S1212).

  Subsequently, the processes of steps S1214 and S1215 are performed for each successive intra PCM block (steps S1213 to S1216). The luminance signal of the intra PCM block is PCM encoded with a fixed length (step S1214), and then the color difference signal of the intra PCM block is PCM encoded with a fixed length (step S1215). When the PCM encoding of all the consecutive NumPCMBlock intra PCM blocks is completed, the encoding process of the intra PCM block ends, and the process proceeds to step S1115 in FIG.

  Subsequently, when the encoded block is intra PCM encoding (NO in step S1115), the process of step S1116 is skipped and the process proceeds to step S1117. If the coding block is not intra PCM coding (YES in step S1115), the quantization parameter information and the transform block are coded (step S1116). The quantization parameter information and transform block encoding processing in step S1116 will be described in detail with reference to the flowchart of FIG.

FIG. 8 is a flowchart for explaining the quantization parameter information and the coding process procedure of the transform block in the coding block of the first example of the present embodiment. These processes are performed by the third encoded bit string generation unit 117 of the image encoding device. In the quantization parameter information and transform block coding processing, the syntax for deriving the quantization parameter QP _Y of the quantization group block by performing the processing from step S1302 to S1308 for each transform block in the coding block. The element cu_qp_delta is encoded (steps S1301 to S1309). First, the values of the variables cbf_luma, cbf_cb, and cbf_cr of the target conversion block are determined (step S1302). Here, the variable cbf_luma includes a non-zero coefficient (non-zero coefficient) when the target luminance conversion block includes a non-zero coefficient, and does not include a non-zero coefficient when the coefficient is encoded. It is a variable that becomes zero. The variable cbf_cb includes a non-zero coefficient (non-zero coefficient) when the transformation block of the target color difference Cb includes a coefficient, and 0 when a coefficient is not encoded and a coefficient is not encoded. It is a variable. The variable cbf_cr includes a non-zero coefficient (non-zero coefficient) when the target color difference Cr conversion block includes a non-zero coefficient, and does not include a non-zero coefficient and zero when the coefficient is not encoded. It is a variable. If any of the variables cbf_luma, cbf_cb, cbf_cr of the target transformation block is 1 (YES in step S1302), the processing from steps S1303 to S1308 is performed, and none of the variables cbf_luma, cbf_cb, cbf_cr of the transformation block is 1. If this is the case, that is, if all are 0 (NO in step S1302), the processing from step S1303 to S1308 is skipped, and the process proceeds to step S1309.

  In step S1303, the variable IsCuQpDeltaCoded is determined (step S1303). If the variable IsCuQpDeltaCoded is 0 (YES in step S1303), the quantization parameter encoding processing from steps S1304 to S1307 is performed. If the variable IsCuQpDeltaCoded is 1 (step S1303) The process from step S1304 to S1307 is skipped, and the process proceeds to step S1308.

Next, quantization group block quantization parameter QP Y encoded information value of the quantization parameter QP _Y of set quantization group block at step S1104 storage memory 112 of FIG. _6, and the encoding / decoding order in The quantization parameter QP _PREV of the immediately preceding quantization group block and the quantization parameter QP _PREV of the immediately preceding quantization group block in the encoding / decoding order are set (step S1304). When setting the coding information storage memory 112, if the size of the coding block is larger than the size of the quantization group block, the value of the same quantization parameter QP _Y is set to all of the quantum blocks included in the coding block. Set to the quantization parameter QP _Y of the quantization group block. The quantization parameter QP _Y of the quantization group block stored in the encoding information storage memory 112 in step S1110 of FIG. 6 and the quantization parameter QP _PREV of the previous quantization group block in the encoding / decoding order are overwritten. .

In step S1305, a syntax element cu_qp_delta is derived from the following equation (step S1305).
QP _DIFF = QP _Y -QP _PRED
cu_qp_delta = (QP _DIFF + 78 + QpBdOffset _Y + (QpBdOffset _Y / 2))% (52 + QpBdOffset _Y )-26-(QpBdOffset _Y / 2);
However, QP _DIFF is the difference value of the quantization parameter. The t variable QpBdOffset _Y is a variable set based on the bit depth of the video signal, and is 0 for 8 bits and 12 for 10 bits. '%' Is a remainder calculation operator (an operator for calculating the remainder of integer precision division).

  Subsequently, the syntax element cu_qp_delta is entropy encoded (step S1306), the variable IsCuQpDeltaCoded is set to 1 (step S1307), and the coefficient of the transform block is encoded (step S1308).

  When the processing of all transform blocks in the coding block is completed, the quantization parameter information and transform block coding processing in FIG. 8 are completed. Subsequently, the process proceeds to step S1117 in FIG. 6 to process the next encoded block.

  When processing of all the coding blocks in the coding tree block is completed, the process proceeds to step S1118, and processing of the next coding tree block is performed.

  When processing of all the coding tree blocks in the slice is completed, the process proceeds to step S1119, and processing of the next slice is performed.

  When the processing of all the slices in the picture is completed, the quantization parameter determination / encoding processing ends.

  Next, a method of quantization parameter decoding / derivation processing will be described. FIG. 9 is a flowchart for explaining the quantization parameter decoding / derivation processing procedure. These processes are performed by the quantization parameter deriving unit 205 of the image decoding apparatus. However, the entropy decoding process in step S1604 is performed by the third encoded bit string decoding unit 204 of the image decoding apparatus. The processing from step S1402 to S1415 is performed for each slice in the picture (steps S1401 to S1416). Further, the processing from step S1403 to S1414 is performed for each coding tree block in the slice (steps S1402 to S1415).

Subsequently, by performing the processing from step S1404 to S1413 for each coding block in the coding tree block, the quantization parameter QP _Y of the quantization group block is decoded and derived (steps S1403 to S1414). In the image decoding apparatus, these processes are performed by the second encoded bit string decoding unit 203, the third encoded bit string decoding unit 204, and the quantization parameter deriving unit 205. When deriving the quantization parameter QP _Y , the prediction value QP _PRED of the quantization parameter is derived, the syntax element cu_qp_delta is decoded, and the quantization parameter QP _Y is derived using the decoded syntax element cu_qp_delta. To do.

First, when the target coding block is the head of the quantization group block in decoding order (YES in step S1404), the thin block corresponding to the quantization parameter QP _Y of the quantization group block corresponding to the target coding block is determined. A variable IsCuQpDeltaCoded indicating whether the tax element cu_qp_delta has been decoded is set to 0 (step S1405), and a predicted value QP _PRED of the quantization parameter QP _Y is derived (step S1406). The process in step S1406 will be described in detail with reference to FIG.

Subsequently, the prediction value QP _PRED of the quantization parameter derived in step S1109 is used as the quantization parameter QP _Y of the quantization group block of the encoding information storage memory 212, and the previous quantization group block in the encoding / decoding order. The quantization parameter QP _PREV is set (step S1407). In the encoded information storage memory 212, if the size of the encoded block is larger than the size of the quantized group block, the predicted value QP _PRED of the same quantization parameter is set to all the quantized blocks included in the encoded block. Set to the quantization parameter QP _Y of the quantization group block.

  On the other hand, if the target coding block is not the head of the quantization group block in decoding order (NO in step S1404), the process proceeds to step S1408.

  Subsequently, in steps S1408 to S1411, processing related to intra PCM encoding is performed. The prediction mode (PredMode) is the intra mode (MODE_INTRA), the partition mode (PartMode) is 2N × 2N partition (PART_2Nx2N), and the size of the coding block is set to be larger than the minimum intra PCM block size. If it is equal to or smaller than the size of the maximum intra PCM block (YES in step S1408), the process proceeds to step S1409; otherwise (NO in step S1408), intra PCM encoding is not performed, and steps S1409 to S1411 are skipped. The process proceeds to S1412.

  Subsequently, in step S1409, a syntax element pcm_flag indicating whether the encoded block is encoded as intra PCM encoding is decoded (step S1409). When the encoded block is encoded as an intra PCM block, pcm_flag is set to 1, and when the encoded block is encoded as an intra PCM block, pcm_flag is set to 0. If the encoded block is encoded as an intra PCM block (YES in step S1410), the intra PCM block is decoded (step S411). If the encoded block is encoded as an intra PCM block (step S411) (NO in S1410), the process proceeds to step S1412. The decoding process of the intra PCM block in step S1411 will be described in detail with reference to the flowchart of FIG.

  FIG. 10 is a flowchart for explaining a decoding process procedure of an intra PCM block. These processes are performed by the second encoded bit string decoding unit 203 and the intra PCM decoding unit 207 of the image decoding apparatus.

  First, the syntax element num_subsequent_pcm indicating the number that follows the encoded block encoded by the intra-PCM having the same size as the encoding block to be decoded is entropy decoded (step S1501), and 1 is added to the syntax element num_subsequent_pcm. Thus, the number NumPCMBlock in which intra PCM blocks having the same size as the encoding block to be encoded (encoded blocks encoded by intra PCM) are consecutive is calculated (step S1502).

Subsequently, when the set minimum size of the quantization group block is equal to or smaller than the size of the coding block (YES in step S1503), the processing in steps S1505 and S1506 is performed for each intra PCM block to be decoded (step S1504 to step S1504). S1507). Each corresponding set encoding parameter QP _Y is decoded (step S1505), and each decoded encoding parameter QP _Y is converted into a quantization parameter QP _Y of a quantization group block in the encoding information storage memory 212, and The quantization parameter QP _PREV of the previous quantization group block in the encoding / decoding order is set (step S1506). The quantization parameter QP _Y of the quantization group block set in the encoding information storage memory 212 in step S1407 of FIG. 9 and the quantization parameter QP _PREV of the previous quantization group block in the encoding / decoding order are overwritten. . Regarding the decoding process of the encoding parameter QP _{Y in} step S1505, the quantization parameter QP _Y encoded with a fixed length may be decoded. This is because the IPCM block encoding / decoding is used for the purpose of an emergency mode that guarantees a predetermined unit code amount. Note that the entropy-encoded syntax element cu_qp_delta may be entropy decoded to derive the quantization parameter QP _Y as in steps S1604 to S1605 in FIG.

  On the other hand, if the set minimum size of the quantization group block is larger than the size of the coding block (NO in step S1503), the process of steps S1504 to S1507 is skipped and the process proceeds to step S1512.

  Subsequently, the encoded bit string is aligned on a byte boundary (step S1512).

  Subsequently, the processes of steps S1514 and S1515 are performed for each successive intra PCM block (steps S1513 to S1516). The luminance signal of the intra PCM block is PCM decoded with a fixed length (step S1514), and then the color difference signal of the intra PCM block is PCM decoded with a fixed length (step S1515). When the PCM decoding of all the consecutive NumPCMBlock intra PCM blocks is completed, the decoding process of the present intra PCM block ends, and the process proceeds to step S1412 in FIG.

  Subsequently, when the encoded block is intra PCM encoding (NO in step S1412), the process of step S1413 is skipped and the process proceeds to step S1414. When the encoded block is not intra PCM encoding (YES in step S1412), the quantization parameter information and the transform block are decoded (step S1413). The quantization parameter information and transform block decoding processing in step S413 will be described in detail with reference to the flowchart of FIG.

FIG. 11 is a flowchart for explaining the quantization parameter information and the decoding processing procedure of the transform block in the coding block. These processes are performed by the third encoded bit string decoding unit 204 of the image decoding apparatus. In the quantization parameter information and transform block coding process, the process from step S1602 to S1608 is performed for each transform block in the coding block, thereby decoding and deriving the quantization parameter QP _Y of the quantization group block ( Steps S1601 to S1609). First, the values of the variables cbf_luma, cbf_cb, and cbf_cr of the target conversion block are determined (step S1602). Here, the variable cbf_luma includes a non-zero coefficient (non-zero coefficient) when the target luminance conversion block includes a non-zero coefficient, and does not include a non-zero coefficient when the coefficient is encoded. It is a variable that becomes zero. The variable cbf_cb includes a non-zero coefficient (non-zero coefficient) when the transformation block of the target color difference Cb includes a coefficient, and 0 when a coefficient is not encoded and a coefficient is not encoded. It is a variable. The variable cbf_cr includes a non-zero coefficient (non-zero coefficient) when the target color difference Cr conversion block includes a non-zero coefficient, and does not include a non-zero coefficient and zero when the coefficient is not encoded. It is a variable. If any of the variables cbf_luma, cbf_cb, cbf_cr of the target conversion block is 1 (YES in step S1602), the processing from step S1603 to S1608 is performed, and none of the variables cbf_luma, cbf_cb, cbf_cr of the conversion block is 1. If this is the case, that is, if all are zero (NO in step S1602), the processing from step S1603 to S1608 is skipped and the process proceeds to step S1609.

  In step S1603, the variable IsCuQpDeltaCoded is determined (step S1603). If the variable IsCuQpDeltaCoded is 0 (YES in step S1603), the processes from step S1604 to S1607 are performed. If the variable IsCuQpDeltaCoded is 1 (NO in step S1603), The process from step S1604 to S1607 is skipped, and the process proceeds to step S1608.

  Subsequently, the syntax element cu_qp_delta is entropy decoded (step S1604).

Subsequently, the quantization parameter QP _Y of the quantization group block is derived by the following equation (step S1605).
QP _Y = (((QP _PRED + cu_qp_delta + 52 + 2 * QpBdOffset _Y )% (52 + QpBdOffset _Y ))-QpBdOffset _Y
However, the variable QpBdOffset _Y is a variable that is set based on the bit depth of the video signal.

The value of the quantization parameter QP _Y of the quantization group block derived in step S1605 is stored in the encoding information storage memory 212, and the quantization group block QP _Y of the quantization group block and the previous quantization group block in the encoding / decoding order The quantization parameter QP _PREV is set (step S1606). When setting the coding information storage memory 212, if the size of the coding block is larger than the size of the quantization group block, the value of the same quantization parameter QP _Y is set to all of the quantum blocks included in the coding block. Set to the quantization parameter QP _Y of the quantization group block. The quantization parameter QP _Y of the quantization group block stored in the encoding information storage memory 212 in step S1407 of FIG. 9 and the quantization parameter QP _PREV of the immediately preceding quantization group block in the encoding / decoding order are overwritten. .

  Subsequently, 1 is set to the variable IsCuQpDeltaCoded (step S1607), and the coefficient of the transform block is decoded (step S1608).

  When the processing of all transform blocks in the coding block is completed, the quantization parameter information and transform block decoding processing in FIG. 11 are completed. Subsequently, the process proceeds to step S1414 in FIG. 9 to process the next encoded block.

  When processing of all the coding blocks in the coding tree block is completed, the process proceeds to step S1415, and processing of the next coding tree block is performed.

  When processing of all the coding tree blocks in the slice is completed, the process proceeds to step S1416, and processing of the next slice is performed.

  When the processing of all slices in the picture is completed, the quantization parameter decoding / derivation processing ends.

  FIG. 12 is a flowchart for explaining the prediction process procedure of the prediction value of the quantization parameter in step S1109 in FIG. 6 and S1406 in FIG. These processes are performed by the third encoded bit string generation unit 117 of the image encoding device and the quantization parameter derivation unit 205 of the image decoding device. First, the prediction value of the quantization parameter is derived using the encoded / decoded quantization parameter stored in the encoding information storage memory 112 of the image encoding device or the encoding information storage memory 212 of the image decoding device. To do.

It is determined whether the quantization group block adjacent to the left can be used. If the quantization group block adjacent to the left can be used (YES in step S1701), the encoding information storage memory 112 or the image decoding of the image encoding device is used. The value of the quantization parameter QP _LEFT of the quantization group block adjacent to the left stored in the encoded information storage memory 212 of the device is set as the quantization parameter QP _A (step S1702), and when it cannot be used (in step S1701) NO), the value of the quantization parameter QP _PREV of the immediately preceding quantization group block in the encoding / decoding order stored in the encoding information storage memory 112 of the image encoding device or the encoding information storage memory 212 of the image decoding device Is set as the quantization parameter QP _A (step S1703).

It is determined whether or not the upper adjacent quantization group block can be used. If the upper adjacent quantization group block can be used (YES in step S1704), the encoding information storage memory 112 or the image decoding of the image encoding device The value of the quantization parameter QP _UPPER stored in the encoded information storage memory 212 of the apparatus and adjacent to the quantization group block is set in the quantization parameter QPB (step S1705), and when it cannot be used (NO in step S1704) ), The value of the quantization parameter QP _PREV of the immediately preceding quantization group block in the encoding / decoding order stored in the encoding information storage memory 112 of the image encoding device or the encoding information storage memory 212 of the image decoding device. The quantization parameter QPBQP _B is set (step S1706).

Subsequently, the average value of the quantization parameter QP _A and the quantization parameter QP _B is calculated by the following equation to obtain the quantization parameter predicted value QP _PRED (step S1707), and the quantization parameter derivation process is terminated.
QP _PRED = (QP _A + QP _B + 1) >> 1

  Next, encoding / decoding of the deblocking filter, which is one of the points of the embodiment, will be described in detail. In the present embodiment, in the luminance signal, the boundary between the transform block and the prediction block on the boundary when the picture is divided into 8 × 8 pixel blocks is derived, and deblocking filter processing is performed. FIG. 13 is a diagram illustrating an example of a pixel at a block boundary. When performing the vertical edge filtering process, the left block across the vertical boundary is block P and the right block is block Q as shown in FIG. Further, the vertical boundary pixels included in the block P are pixels p0, p1, p2, and p3 in order from the block boundary to the left, and the pixels included in the block Q are sequentially shifted from the block boundary to the right in the pixel q0, Let q1, q2, and q3. Only the top line is shown, but filtering is performed for each line across the vertical boundary. On the other hand, when horizontal edge filtering is performed, the upper block across the horizontal boundary is referred to as block P and the lower block as block Q as shown in FIG. Further, the pixels on the vertical boundary to be filtered included in the block P are pixels p0, p1, p2, and p3 in order from the block boundary, and the pixels included in the block Q are pixel q0 in order from the block boundary to the bottom. Let q1, q2, and q3. Only the leftmost line is shown, but filtering is performed for each line across the horizontal boundary.

  FIG. 14, FIG. 15, FIG. 18, and FIG. 19 are flowcharts for explaining processing procedures of the deblocking filter unit 111 of the image encoding device and the deblocking filter unit 211 of the image decoding device. In the image encoding device, after all the decoded image signals of the picture to be encoded are stored in the first decoded image memory 113, the deblocking filter unit 111 performs the deblocking filter process, and the second decoded image memory 114. In the image decoding apparatus, after all the decoded image signals of the decoding target picture are stored in the first decoded image memory 213, the deblocking filter unit 211 performs the deblocking filter process, and the second decoded image memory 214 stores the decoded image signal. Stored.

  First, the deblocking filter processing procedure of FIG. 14 will be described. Steps S2102 and S2103 are repeated for each encoded block in the decoded picture stored in the first decoded image memory 113 of the image encoding device or the first decoded image memory 213 of the image decoding device ( Steps S2101 to S2104). The horizontal filtering of the vertical edge is performed (step S2102), and the output pixels p ′ and q ′ are stored in the image memory (step S2103). At this time, it is stored in an image memory for temporarily storing intermediate data. Subsequently, the processes in steps S2106 and S2107 are repeated for each encoded block in the picture subjected to the deblocking filter process (steps S2105 to S2108). A filtering process in the vertical direction of the horizontal edge is performed (step S2106), and the output pixels p ′ and q ′ are stored in the image memory (step S2107). At this time, the image data is stored in the second decoded image memory 114 of the image encoding device or the second decoded image memory 214 of the image decoding device. The filtering process in the horizontal direction of the vertical edge in step S2102 and the filtering process in the vertical direction of the horizontal edge in step S2106 differ only in the direction of the process, and the processing procedure is common, and FIG. 15, FIG. 18, and FIG. 19 are used. Will be described in detail.

  Deblocking filter processing procedures for each coding block in FIG. 15 will be described. First, using the coding information stored in the coding information storage memory 112 or the coding information storage memory 212, the boundary of the transform block in the coding block is derived (step S2201). FIG. 16 is a diagram for explaining an example of the vertical boundary and the horizontal boundary of the transform block in the coding block. When performing vertical edge filtering processing, the boundaries of transform blocks in the coding block indicated by the thick line in FIG. 16A on the left side and the inside of the coding block are derived. On the other hand, when the horizontal edge filtering process is performed, the boundaries of the transform blocks in the coding block indicated by the thick line in FIG. 16B on the inside and upper sides of the coding block are derived.

  Subsequently, a prediction block boundary is derived (step S2202). FIG. 17 is a diagram for explaining an example of a vertical boundary and a horizontal boundary of a prediction block in a coding block. When performing the vertical edge filtering process, the boundaries of the prediction blocks in the coding block indicated by the thick lines in FIG. 17A on the left side and the inside of the coding block are derived. On the other hand, when the horizontal edge filtering process is performed, the boundary of the prediction block in the coding block indicated by the thick line in FIG. 17B on the inside and the upper side of the coding block is derived.

  Subsequently, the strength of each block boundary of the transform block and the prediction block prediction is derived (step S2203). Instead of uniformly performing the deblocking filter process with the same strength for each block boundary, it is determined how strong the deblocking filter process should be performed based on the condition of each block boundary.

  When the pixel p0 or q0 is included in the encoded block encoded in the intra prediction mode, the value of the variable bS indicating the strength of the block boundary is set to 2. 2 is a value indicating the strongest strength.

  Otherwise, if the pixel p0 or q0 is included in a transform block including a non-zero transform coefficient, the value of the variable bS indicating the block boundary strength is set to 1. 1 is a value indicating the most moderate intensity.

  Otherwise, if the prediction block including the pixel p0 is a reference picture different from the prediction block including the pixel q0 or the number of motion vectors of a different prediction block, the value of the variable bS indicating the block boundary strength is set to 1. .

  Otherwise, if the value of the motion vector of the prediction block including the pixel p0 and the value of the motion vector of the prediction block including the pixel p0 are not less than a predetermined value, the variable bS indicating the strength of the block boundary Set the value to 1.

  Otherwise, the value of the variable bS indicating the block boundary strength is set to zero. 0 is a value indicating the weakest intensity.

  Subsequently, the luminance signal is filtered (step S2204). The luminance signal filtering processing procedure in step S2204 will be described in detail with reference to FIG. Subsequently, the color difference signal is filtered (step S2205), and this deblocking filter processing procedure is terminated. For the filtering of the color difference signal, the same method as the filtering of the luminance signal is adopted, and the details are omitted.

  FIG. 18 is a flowchart showing the signal filtering processing procedure in step S2204. Luminance block edge filtering is performed for each 8 × 8 block in the encoded block (steps S2301 to S2303) (step S2302).

FIG. 19 is a flowchart showing the filtering process procedure of the block edge in step S2302. First obtains the quantization parameter QP _P blocks including the pixel p (step S3101). The quantization parameter QP _Q of the block including the pixel q is acquired (step S3102). Subsequently, an average value of the quantization parameter QP _P and the quantization parameter QP _Q is derived by the following equation, and is set as QP _AVE (step S3107).
QP _AVE = (QP _P + QP _Q + 1) >> 1

Subsequently, the variable β is derived (step S3108). The variable β is a variable used when determining the strength of the deblocking filter. As the average value QP _AVE of the quantization parameter increases, the value of the variable β increases. In the derivation of the variable β, the index indexB is derived by the following equation. When the index indexB is smaller than 0, the index β is clipped to 0, and when it is larger than 51, the variable β is referred to by referring to the table of FIG. To derive.
indexB = QP _AVE + (beta_offset_div2 << 1)
However, beta_offset_div2 is a syntax element encoded in sequence units, and has a value of −13 to 13. When the value of β is small, the filter is not easily applied as will be described later, and the weak filter is easily selected.

Subsequently, a variable t _c is derived (step S3109). The value of the variable t _c is a variable used when determining the strength of the deblocking filter, and the value of the variable t _c increases as the value of the quantization parameter average value QP _AVE or the variable bS indicating the boundary strength increases. In the derivation of the variable tc, the index indexTc is derived by the following equation. When the index indexTc is smaller than 0, the index t is clipped to 0, and when the index indexTc is larger than 53, the variable t _{c is referred to.} Is derived.
indexTc = QP _AVE + 2 * (bS-1) + (tc_offset_div2 << 1)
However, tc_offset_div2 is a syntax encoded in sequence units, and has a value of −13 to 13.

  Subsequently, an evaluation value d is derived from the relationship between pixel values in the vicinity of the boundary (step S3110). If the evaluation value d is greater than or equal to the variable β (NO in step S3111), the input pixel is used as an output pixel as it is (step S3116), and the luminance signal filtering process is terminated. If the value of the variable β is small, the filter is not easily applied.

  If the evaluation value d is smaller than the variable β (YES in step S3111), the type of deblocking filter is determined from the relationship between the pixel value near the boundary and the variable β (step S3112). In the present embodiment, two types of filters, a strong filter and a weak filter, are switched and used. If the value of the variable β is large, the strong filter is easily selected, and if the value of the variable β is small, the weak filter is easily selected.

  Subsequently, the signal is filtered for each line of the filtering block (steps S3113 to S3115). Filtering of pixels p0, p1, p2, p3, q0, q1, q2, and q3 sandwiching the block boundary of the signal for each line is performed (step S3114).

  FIG. 22 is a flowchart showing the filtering process procedure of the luminance signal for each line in step S3114. When applying a strong filter (YES in step S3201), a strong filter is applied (step S3202). Otherwise (NO in step S3201), a weak filter is applied (step S3203). Let p ′ and q ′ be the filtered pixels, respectively.

  Subsequently, the value of the flag pcm_loop_filter_disable_flag is determined (step S3204). pcm_loop_filter_disable_flag is a flag indicating whether or not to apply a deblocking filter or other loop filter to a pixel encoded by intra PCM, and is set and encoded in units of sequence, picture, or slice. If pcm_loop_filter_disable_flag is 1 (YES in step S3204), the process proceeds to step S3205. If pcm_loop_filter_disable_flag is 0 (NO in step S3204), the processing of steps S3205 to S3208 is skipped and the filtered pixels p ′ and q ′ Is the output pixel, and this filtering processing procedure is terminated.

  If the block including the pixel p is an intra PCM block (YES in step S3205), the pixel value of the input pixel p is set to the output pixel p ′ (step S3206). In other words, the output pixel is output as it is without filtering the pixel p of the intra PCM block. If the block including the pixel p is not an intra PCM block (YES in step S3205), the filtered pixel p ′ becomes an output pixel as it is, and the process proceeds to step S3207.

  When the block including the pixel q is an intra PCM block (YES in step S3207), the pixel value of the input pixel q is set to the output pixel q ′ (step S3208). That is, the pixel q of the intra PCM block is output as it is without being filtered. If the block including the pixel q is not an intra PCM block (YES in step S3207), the filtered pixel q ′ becomes an output pixel as it is, and this filtering processing procedure ends.

In the present embodiment, quantization parameter information is encoded even in an intra PCM block, and the quantization parameter is used, so that a filter having an appropriate strength can be applied. For example, in FIG. 13 (a) or (b), the bit depth of the luminance signal is 8 bits, the block P is intra PCM coding, the coding parameter QP _P is 20, and the block Q is coded in the intra prediction mode. When the value of the parameter QP _Q is 30 and the value of the syntax element beta_offset_div2 is 0, the value of the encoding parameter QP _AVE is 25 and the value of the variable β is 15. In this case, an appropriate filter is applied to the block Q, and the block distortion is removed. For intra PCM blocks with no coding degradation, by setting 1 to pcm_loop_filter_disable_flag, pixels that are not always filtered are output, so that coding degradation does not occur for intra PCM blocks.

  In this embodiment, since it is possible to switch whether or not to apply a deblocking filter to pixels encoded with intra PCM based on the flag pcm_loop_filter_disable_flag, the flag pcm_loop_filter_disable_flag should be set to 0 particularly at a high bit rate. Thus, the determination of whether the block is intra PCM can be omitted, and the amount of calculation can be reduced.

  In the present embodiment, the flag pcm_loop_filter_disable_flag is set, but the determination process of the flag pcm_loop_filter_disable_flag in step S3204 may be omitted. When the determination process of the flag pcm_loop_filter_disable_flag is omitted, the determination processes of steps S3205 and S3207 are always performed, and pixels that are not filtered are always output for intra PCM blocks that have no coding deterioration.

In the first example of the present embodiment, when the target coding block is intra PCM coding and the size of the coding block is less than the minimum size of the quantization group block, the corresponding quantization group block Encoding of the syntax element cu_qp_delta indicating the quantization parameter information of the quantization parameter QP _Y derived by the quantization parameter determination unit 102 of the image encoding device even if the quantization parameter is not encoded in FIG. Processing was omitted.

In the second example of the present embodiment, when the target coding block is intra PCM coding and the size of the coding block is less than the minimum size of the quantization group block, quantization is performed within the quantization group block. When the parameter is not encoded, the syntax element cu_qp_delta indicating the quantization parameter information is encoded. When the coding block is intra PCM coding and the size of the coding block is less than the minimum size of the quantization group block, and the quantization parameter is not coded in the corresponding quantization group block, the image code By encoding the quantization parameter information of the quantization parameter QP _Y derived by the quantization parameter determination unit 102 of the encoding device, all of the encoded blocks that are not intra-PCM encoded in the same quantization group block are included. Even when the transform block has a coefficient of 0, the syntax element cu_qp_delta is encoded. At this time, the value of the quantization parameter QP _Y derived by the quantization parameter determination unit 102 of the image encoding device is stored in the encoding information storage memory 112 of the image encoding device or the encoding information storage memory 212 of the image decoding device. Is done. A method of encoding quantization parameter information with an intra PCM block when the size of the encoding block is smaller than the size of the quantization group block will be described in detail with reference to FIGS.

  Also in the image coding apparatus according to the second example of the present embodiment, the quantization parameter is determined and coded according to the processing procedure shown in FIG. 6 as in the second example of the present embodiment. However, the difference is that in step S1114 of FIG. 6, the intra PCM block is encoded according to the processing procedure shown in FIG. FIG. 23 is a flowchart for explaining an intra-PCM block encoding process procedure according to the second example of the present embodiment. These processes are performed by the intra PCM encoding unit 104 and the second encoded bit string generation unit 116 of the image encoding device. 7 is different from FIG. 7 in that steps S1208 to S1211 are added.

  First, 1 is subtracted from the number NumPCMBlock in which intra PCM blocks having the same size as the encoding block to be encoded (encoded blocks encoded by intra PCM) are consecutive, and the same size as the encoding block to be encoded The value of the syntax element num_subsequent_pcm indicating the number that follows the encoded block encoded by the intra PCM is derived (step S1201), and the syntax element num_subsequent_pcm is entropy encoded (step S1202). Note that the value of the syntax element num_subsequent_pcm ranges from 0 to 3.

Subsequently, when the set minimum size of the quantization group block is equal to or smaller than the size of the encoded block (YES in step S1203), the processes of steps S1205 and S1206 are performed for each intra PCM block to be encoded (steps S1204 to S1204). Step S1207). Quantization parameter QP Y value of the quantization parameter QP _Y of set quantization group block at step S1104 the quantization group blocks of encoded information storage memory 112 of FIG. _6, and the quantum of the immediately preceding the encoding / decoding order The quantization parameter QP _PREV of the quantization group block is set (step S1205). The quantization parameter QP _Y of the quantization group block set in the encoding information storage memory 112 in step S1110 of FIG. 6 and the quantization parameter QP _PREV of the previous quantization group block in the encoding / decoding order are overwritten. . Further, the encoding parameter QP _Y is encoded (step S1206). For the encoding process of the encoding parameters QP _Y in step S1206, the encoding parameter QP _Y may be encoded with a fixed length. This is because the IPCM block encoding / decoding is used for the purpose of an emergency mode that guarantees a predetermined unit code amount. Note that the syntax element cu_qp_delta may be derived and the syntax element cu_qp_delta may be entropy-coded as in steps S1305 to S1306 in FIG.

On the other hand, if the set minimum size of the quantization group block is larger than the size of the encoded block (NO in step S1203), the intra PCM blocks to be encoded are assumed to be included in one quantization group block. The processes of S1209 and S1210 are performed. Quantization parameter QP Y value of the quantization parameter QP _Y of set quantization group block at step S1104 the quantization group blocks of encoded information storage memory 112 of FIG. _6, and the quantum of the immediately preceding the encoding / decoding order The quantization parameter QP _PREV of the quantization group block is set (step S1209). The quantization parameter QP _Y of the quantization group block set in the encoding information storage memory 112 in step S1110 of FIG. 6 and the quantization parameter QP _PREV of the previous quantization group block in the encoding / decoding order are overwritten. . Further, the encoding parameter QP _Y is encoded (step S1210). For the encoding process of the encoding parameters QP _Y in step S1210, the encoding parameter QP _Y may be encoded with a fixed length. This is because the IPCM block encoding / decoding is used for the purpose of an emergency mode that guarantees a predetermined unit code amount. Note that the syntax element cu_qp_delta may be derived and the syntax element cu_qp_delta may be entropy-coded as in steps S1305 to S1306 in FIG.

  Subsequently, the encoded bit string is aligned with a byte boundary (step S1212).

  Subsequently, the processes of steps S1214 and S1215 are performed for each successive intra PCM block (steps S1213 to S1216). The luminance signal of the intra PCM block is PCM encoded with a fixed length (step S1214), and then the color difference signal of the intra PCM block is PCM encoded with a fixed length (step S1215). When the PCM encoding of all the consecutive NumPCMBlock intra PCM blocks is completed, the encoding process of the intra PCM block ends, and the process proceeds to step S1115 in FIG.

  Also in the image decoding apparatus according to the second example of the present embodiment, the quantization parameter decoding / derivation process is performed according to the processing procedure shown in FIG. 9 as in the second example of the present embodiment. However, the difference is that the intra PCM block decoding process is performed in step S1411 of FIG. 9 according to the processing procedure shown in FIG. FIG. 24 is a flowchart for explaining the decoding process procedure of the intra PCM block according to the second example of the present embodiment. These processes are performed by the second encoded bit string decoding unit 203 and the intra PCM decoding unit 207 of the image decoding apparatus. 10 is different from FIG. 10 in that steps S1508 to S1511 are added.

  First, the syntax element num_subsequent_pcm indicating the number that follows the encoded block encoded by the intra-PCM having the same size as the encoding block to be decoded is entropy decoded (step S1501), and 1 is added to the syntax element num_subsequent_pcm. Thus, the number NumPCMBlock in which intra PCM blocks having the same size as the encoding block to be encoded (encoded blocks encoded by intra PCM) are consecutive is calculated (step S1502).

Subsequently, when the set minimum size of the quantization group block is equal to or smaller than the size of the coding block (YES in step S1503), the processing in steps S1505 and S1506 is performed for each intra PCM block to be decoded (step S1504 to step S1504). S1507). Each corresponding set encoding parameter QP _Y is decoded (step S1505), and each decoded encoding parameter QP _Y is converted into a quantization parameter QP _Y of a quantization group block in the encoding information storage memory 212, and The quantization parameter QP _PREV of the previous quantization group block in the encoding / decoding order is set (step S1506). The quantization parameter QP _Y of the quantization group block set in the encoding information storage memory 212 in step S1407 of FIG. 9 and the quantization parameter QP _PREV of the previous quantization group block in the encoding / decoding order are overwritten. . Regarding the decoding process of the encoding parameter QP _{Y in} step S1505, the quantization parameter QP _Y encoded with a fixed length may be decoded. This is because the IPCM block encoding / decoding is used for the purpose of an emergency mode that guarantees a predetermined unit code amount. Note that, as in steps S1604 to S1605 in FIG. 11, the entropy-encoded syntax element cu_qp_delta may be entropy decoded to derive the quantization parameter QP _Y.

On the other hand, if the set minimum size of the quantization group block is larger than the size of the encoding block (NO in step S1503), the intra-PCM blocks to be continuously encoded are included in one quantization group block. The processing of S1509 and S1510 is performed. The encoding parameter QP _Y is decoded (step S1509), the corresponding set encoding parameter QP _Y of the quantization group block is converted into the quantization parameter QP _Y of the quantization group block of the encoding information storage memory 212, and encoding. / Set to the quantization parameter QP _PREV of the immediately preceding quantization group block in decoding order (step S1510). The quantization parameter QP _Y of the quantization group block set in the encoding information storage memory 212 in step S1407 of FIG. 9 and the quantization parameter QP _PREV of the previous quantization group block in the encoding / decoding order are overwritten. . For the encoding process of the encoding parameters QP _Y in step S1509, it may decode the encoded quantized parameters QP _Y fixed length. This is because the IPCM block encoding / decoding is used for the purpose of an emergency mode that guarantees a predetermined unit code amount. Note that, as in steps S1604 to S1605 in FIG. 11, the entropy-encoded syntax element cu_qp_delta may be entropy decoded to derive the quantization parameter QP _Y.

  Subsequently, the encoded bit string is aligned on a byte boundary (step S1512).

  Subsequently, the processes of steps S1514 and S1515 are performed for each successive intra PCM block (steps S1513 to S1516). The luminance signal of the intra PCM block is PCM decoded with a fixed length (step S1514), and then the color difference signal of the intra PCM block is PCM decoded with a fixed length (step S1515). When the PCM decoding of all the consecutive NumPCMBlock intra PCM blocks is completed, the decoding process of the present intra PCM block ends, and the process proceeds to step S1412 in FIG.

  The encoded stream of the image output from the image encoding device according to the embodiment described above has a specific data format so that it can be decoded according to the encoding method used in the embodiment. The image decoding device corresponding to the image encoding device can decode the encoded stream of this specific data format.

  When a wired or wireless network is used to exchange an encoded stream between an image encoding device and an image decoding device, the encoded stream is converted into a data format suitable for the transmission form of the communication path and transmitted. May be. In that case, an image transmission apparatus that converts the encoded stream output from the image encoding apparatus into encoded data in a data format suitable for the transmission form of the communication path and transmits the encoded data to the network, and receives the encoded data from the network There is provided an image receiving apparatus that restores the encoded stream and supplies the encoded stream to the image decoding apparatus.

  An image transmission device includes a memory that buffers an encoded stream output from the image encoding device, a packet processing unit that packetizes the encoded stream, and a transmission unit that transmits packetized encoded data via a network. including. The image reception device receives a packetized encoded data via a network, a memory for buffering the received encoded data, packetizes the encoded data to generate an encoded stream, And a packet processing unit provided to the image decoding device.

  The above processing relating to encoding and decoding can be realized as a transmission, storage, and reception device using hardware, and is stored in a ROM (Read Only Memory), a flash memory, or the like. It can also be realized by firmware or software such as a computer. The firmware program and software program can be recorded on a computer-readable recording medium, provided from a server through a wired or wireless network, or provided as a data broadcast of terrestrial or satellite digital broadcasting Is also possible.

  The present invention has been described based on the embodiments. The embodiments are exemplifications, and it will be understood by those skilled in the art that various modifications can be made to combinations of the respective constituent elements and processing processes, and such modifications are within the scope of the present invention. .

  Reference Signs List 101 image memory, 102 quantization parameter determination unit, 103 intra prediction unit, 104 intra PCM encoding unit, 105 inter prediction unit, 106 encoding method determination unit, 107 residual signal generation unit, 108 orthogonal transform / quantization unit, 109 Inverse quantization / inverse orthogonal transformation unit, 110 Decoded image signal superimposing unit, 111 Deblocking filter unit, 112 Encoded information storage memory, 113 First decoded image memory, 114 Second decoded image memory, 115 First Encoded bit string generation unit, 116 Second encoded bit string generation unit, 117 Third encoded bit string generation unit, 118 Encoded bit string multiplexing unit, 119 switch, 201 Encoded bit string separation unit, 202 First encoding A bit string decoding unit; 203 a second encoded bit string decoding unit; 04 third encoded bit string decoding unit, 205 quantization parameter derivation unit, 206 intra prediction unit, 207 intra PCM decoding unit, 208 inter prediction unit, 209 inverse quantization / inverse orthogonal transform unit, 210 decoded image signal superimposing unit, 211 Deblocking filter unit, 212 Encoded information storage memory, 213 First decoded image memory, 214 Second decoded image memory, 215 switch, 216 switch

Claims (5)

An image decoding device for decoding an image signal in units of blocks,
A quantization parameter derivation unit that decodes quantization parameter information corresponding to a block to be inversely quantized and also decodes quantization parameter information corresponding to a block to be decoded without compression to derive a quantization parameter of the block When,
A first decoding unit that dequantizes and decodes an image signal of a block that has been compressed and encoded based on the quantization parameter;
A second decoding unit that decodes the image signal of the non-compressed encoded block;
A deblocking filter unit that determines a filtering strength using a quantization parameter set in each of the blocks on both sides of the block boundary and performs a filtering process on the decoded image signals of the blocks on both sides of the block boundary. An image decoding apparatus characterized by the above.   The quantization parameter deriving unit does not decode the quantization parameter in an encoded block encoded with non-compression that is less than the minimum size of the encoded block to which the quantization parameter information is to be sent. The image decoding device described.   The quantization parameter deriving unit treats the quantization parameters derived by decoding the quantization parameter information as one group in an encoding block less than the minimum size of the encoding block to which the quantization parameter information is to be sent. The image decoding apparatus according to claim 1, wherein the image decoding apparatus is set to quantization parameters of a plurality of coding blocks. An image decoding method for decoding an image signal in block units,
Quantization parameter derivation step for decoding quantization parameter information corresponding to a block to be dequantized and decoding quantization parameter information corresponding to a block to be decoded without compression to derive a quantization parameter for the block When,
A first decoding step of dequantizing and decoding an image signal of a block encoded by compression based on the quantization parameter;
A second decoding step for decoding the image signal of the non-compressed encoded block;
A deblocking filter step of determining a filtering strength using a quantization parameter set for each of the blocks on both sides of the block boundary and performing a filtering process on the decoded image signals of the blocks on both sides of the block boundary. An image decoding method characterized by the above. An image decoding program for decoding an image signal in block units,
Quantization parameter derivation step for decoding quantization parameter information corresponding to a block to be dequantized and decoding quantization parameter information corresponding to a block to be decoded without compression to derive a quantization parameter for the block When,
A first decoding step of dequantizing and decoding an image signal of a block encoded by compression based on the quantization parameter;
A second decoding step for decoding the image signal of the non-compressed encoded block;
A deblocking filter step for determining a filtering strength using a quantization parameter set for each of the blocks on both sides of the block boundary, and filtering the decoded image signals of the blocks on both sides of the block boundary; An image decoding program that is executed.

Images