JP7796810B2 - Decoding device and decoding method - Google Patents

Description

The present invention relates to an encoding device, a decoding device, and a program.

Video coding methods that perform coding processing on a block-by-block basis, such as the series of standards recommended by the MPEG (Moving Picture Experts Group) committee, are characterized by uneven quality between blocks because each block is coded under different conditions.

The advantage of this feature is that it allows for coding control that does not result in visual degradation of image quality, by controlling the coding to occur according to the signal characteristics of a partial region of the image, for example, by encoding regions with signal characteristics that are easily detected by humans into a high-quality signal, and regions that are difficult to detect into a low-quality signal. The disadvantage is that because coding is controlled on a block-by-block basis, differences in image quality become apparent at block boundaries, and block-like distortion may be detected.

To reduce this type of signal degradation, deblocking filtering is commonly used in recent encoding methods. Deblocking filters are filters designed to reduce signal gaps, and generally have the characteristics of a low-pass filter.

If a deblocking filter is applied in a fixed manner, the signal degradation caused by the filter processing may have a greater effect than the quality improvement at the boundaries between high-quality regions, and conversely, the filter may not be effective enough at the boundaries between regions that have been coded at excessively low quality.

In light of this situation, HEVC, a conventional video coding method, controls the filter strength of a deblocking filter based on a quantization parameter (specifically, the average value of the quantization parameters of blocks on either side of a boundary) (see, for example, Non-Patent Document 1). Here, a quantization parameter is a parameter for which one value is set for one block, and which specifies the coarseness (step size) of quantization for that block.

By the way, the committee's draft for VVC, a new video coding method, uses Luma Mapping and Chroma Scaling (LMCS) (see, for example, Non-Patent Document 2).

In LMCS, the value of the luminance signal of the original image is changed by a mapping process using a lookup table (conversion table) before performing processes such as prediction and conversion, depending on the signal characteristics of the input image, and prediction, conversion, and quantization processes are then applied to the mapped luminance signal. Furthermore, the value of the chrominance signal of the original image is changed by a scaling process depending on the value of the luminance signal of the original image, and prediction, conversion, and quantization processes are then applied to the scaled chrominance signal.

LMCS can improve coding efficiency by mapping to the luminance signal and scaling to the color difference signal, taking advantage of the bias in pixel values in the original image.

Recommendation ITU-T H. 265, (12/2016), “High efficiency video coding”, International Telecommunication Union JVET-O2002, “Algorithm description for Versatile Video Coding and Test Model 6 (VTM 6)”

The above-mentioned LMCS mapping and scaling processes are performed on the encoding device before the prediction, transformation, and quantization processes are performed on the original image. On the decoding device, the inverse mapping process corresponding to the mapping process and the inverse scaling process corresponding to the scaling process are performed after the inverse quantization process, inverse transformation process, and prediction process. In other words, the quantization error generated by the quantization process and inverse quantization process may increase or decrease due to the inverse mapping and inverse scaling processes.

However, while conventional video coding methods control the filter strength of a deblocking filter by taking into account the quantization parameter, which is a unique parameter set on a block-by-block basis, they do not take into account the inverse mapping and inverse scaling processes described above when controlling the filter strength.

As a result, there is no way to control the filter strength of the deblocking filter to address the block noise caused by the inverse mapping and inverse scaling processes, and the deblocking filter is not effective in reducing block noise.

The present invention aims to provide an encoding device, a decoding device, and a program that improve image quality and encoding efficiency by appropriately controlling a deblocking filter.

The encoding device according to the first aspect is an encoding device that encodes encoding target blocks obtained by dividing an image into block units, and includes: a conversion unit that generates a new encoding target block by performing a conversion process consisting of a mapping process for each pixel value of a luminance signal and a scaling process for each pixel value of a color difference signal based on a conversion table within the encoding target block; a transform/quantization unit that performs a conversion process and a quantization process on a prediction residual, which is the difference between the encoding target block newly generated by the conversion unit and a prediction block generated by predicting the encoding target block; an inverse quantization/inverse transform unit that restores the prediction residual by performing an inverse quantization process and an inverse transform process on the transform coefficients generated by the transform/quantization unit; and a composite block that generates a composite block by combining the restored prediction residual with the prediction block. the deblocking filter that performs filtering on the boundary between the restored target block and an adjacent block adjacent to the target block; and a filter control unit that controls the filter strength of the deblocking filter based on the quantization parameters used in the quantization and inverse quantization processes and the conversion table used in the conversion and inverse conversion processes, wherein the conversion table is a table set for one or more slices and is a table consisting of values indicating the number of pixel values after conversion to be assigned to each band obtained by dividing the range from the minimum to maximum possible values of the target pixel values into a specified number of bands.

Here, "slice" may be interpreted as "picture," "tile," "tile group," or "brick."

A decoding device according to a second aspect is a decoding device that decodes an encoded stream in units of blocks generated by dividing an image, and is configured to include an entropy decoding unit that decodes the encoded stream and outputs transform coefficients corresponding to a block to be decoded; an inverse quantization and inverse transform unit that performs inverse quantization and inverse transform processing on the transform coefficients output by the entropy decoding unit to restore a prediction residual; a synthesis unit that generates a composite block by synthesizing the restored prediction residual with a prediction block generated by predicting the block to be decoded; and a mapping process for each pixel value of a luminance signal and a scaling process for each pixel value of a color difference signal that are performed on the encoding side based on a conversion table for each pixel value in the generated composite block. The coding system comprises an inverse transformation unit that restores the block to be coded by performing an inverse transformation corresponding to the transformation performed in the dequantization process; a deblocking filter that performs a filtering process on the boundary between the restored block to be coded and an adjacent block adjacent to the block to be coded; and a filter control unit that controls the filter strength of the deblocking filter based on the quantization parameter used in the inverse quantization process and the transformation table used in the inverse transformation process, wherein the transformation table is a table set for one or more slices and is a table consisting of values indicating the number of pixel values after transformation to be assigned to each band obtained by dividing the range from the minimum to maximum possible values of pixel values to be transformed into a specified number of bands.

The program according to the third aspect causes a computer to function as the encoding device according to the first aspect.

The program according to the fourth aspect is intended to cause a computer to function as the decryption device according to the second aspect.

The present invention provides an encoding device, a decoding device, and a program that improve image quality and encoding efficiency by appropriately controlling a deblocking filter.

FIG. 1 is a diagram illustrating a configuration of an encoding device according to an embodiment. 10 is a graph showing the relationship between input pixel values and output pixel values in conversion processing according to the embodiment. FIG. 2 is a diagram illustrating a configuration of a conversion unit of the encoding device according to the embodiment. FIG. 2 is a diagram illustrating an example of a quantization matrix according to the embodiment. FIG. 2 is a diagram illustrating a configuration of an inverse conversion unit of the encoding device according to the embodiment. FIG. 10 is a diagram illustrating an example of the operation of a deblocking filter according to the embodiment. FIG. 2 is a diagram illustrating a configuration of a filter control unit of the encoding device according to the embodiment. FIG. 1 is a diagram illustrating a configuration of a decoding device according to an embodiment. FIG. 2 is a diagram illustrating a configuration of an inverse conversion unit of a decoding device according to an embodiment. FIG. 2 is a diagram illustrating a configuration of a filter control unit of a decoding device according to an embodiment. FIG. 10 is a diagram illustrating an example of an operation flow of a filter control unit according to the embodiment.

An encoding device and a decoding device according to an embodiment will be described with reference to the drawings. The encoding device and the decoding device according to an embodiment encode and decode moving images, respectively, as typified by MPEG. In the following description of the drawings, the same or similar parts are designated by the same or similar reference numerals.

<Configuration of the encoding device>
First, the configuration of an encoding device according to this embodiment will be described. Fig. 1 shows the configuration of an encoding device 1 according to this embodiment. The encoding device 1 is a device that performs encoding in units of blocks obtained by dividing an image in units of slices.

As shown in FIG. 1, the encoding device 1 includes a block division unit 100, a transformation unit 110, a subtraction unit 120, a transformation/quantization unit 130, an entropy encoding unit 140, an inverse quantization/inverse transformation unit 150, a synthesis unit 160, an inverse transformation unit 170, a deblocking filter 180, a filter control unit 181, a memory 185, and a prediction unit 190.

The block division unit 100 divides an input image, which is made up of frames (or pictures) that make up a moving image, into multiple image blocks and outputs the divided image blocks to the conversion unit 110. The size of the image blocks is, for example, 32x32 pixels, 16x16 pixels, 8x8 pixels, or 4x4 pixels. The shape of the image blocks is not limited to squares and may be rectangular (non-square). An image block is the unit used for coding by the encoding device 1 (block to be encoded) and the unit used for decoding by the decoding device (block to be decoded). Such an image block is sometimes called a CU (Coding Unit).

The block division unit 100 divides the luminance signal and the color difference signal into blocks. The following mainly describes the case where the shape of the block division is the same for the luminance signal and the color difference signal, but it is also possible to control the division independently for the luminance signal and the color difference signal. When no particular distinction is made between luminance blocks and color difference blocks, they are simply referred to as the blocks to be coded.

The conversion unit 110 generates a new converted block to be coded by converting each pixel value in the block to be coded output from the block division unit 100 based on a conversion table, and outputs the new converted block to be coded to the subtraction unit 120.

Figure 2 is a graph showing the relationship between input pixel values and output pixel values in the conversion process according to this embodiment. In Figure 2, the horizontal axis represents the input signal value, and the vertical axis represents the output signal value.

As shown in Figure 2, the conversion table is a table set for one or more slices, and is a coefficient table that represents the relationship between the input signal before conversion processing and the output signal after conversion processing. Specifically, the conversion table stores values that indicate the number of pixel values of the converted output signal to be assigned to each band obtained by dividing the input signal before conversion (pixel values to be converted) from its minimum to maximum possible values into a predefined number (N).

For example, the conversion table will be explained using an example where the number of bands N is 16 in the conversion process for a 10-bit image signal. The minimum possible value of the input signal before conversion, 0, to the maximum possible value, 1023, is assigned as the input signal corresponding to each equally divided band. For example, the first band corresponds to input pixel values 0 to 63. The second band corresponds to input pixel values 64 to 127. Input signals are assigned in a similar manner up to the 16th band.

Each band corresponds to the position of each coefficient in the conversion table. The coefficients stored in the conversion table represent the number of output pixel values assigned to each band. For example, if the conversion table lmcs _CW is lmcs _CW = {39, 40, 55, 70, 80, 90, 97, 97, 104, 83, 57, 55, 49, 44, 34, 30}, the output pixel values corresponding to the first band are 0 to 38, and the output pixel values corresponding to the second band are 39 to 78. The same allocation is performed for the third to sixteenth bands.

If the value corresponding to a certain band in the conversion table is large, the number of output pixel values assigned to that band will be large; conversely, if the value is small, the number of output pixel values assigned to that band will be small. For this reason, when quantization processing is performed on a signal region converted using a conversion table, and then the original signal region is restored by inverse conversion, the amount of quantization error will be different if the conversion table values used in the conversion processing are different, even if the same quantization processing is performed.

Specifically, when the value in the conversion table is 32, this means that 64 input pixel values of a certain band are converted to be represented by 32 output pixel values. When the encoding device 1 applies quantization processing to the output signal, resulting in quantization error, the decoding device 2 (see Figure 8) applies an inverse conversion process corresponding to the conversion processing, including the resulting quantization error. In this case, the conversion table value for the band is 32, and the inverse conversion process inversely converts 32 input pixel values for the inverse conversion process so that they are represented by 64 output pixel values. This means that the decoding device 2 doubles the quantization error that occurred in the signal corresponding to that band before outputting it.

In other words, even when the same quantization parameter is applied to an image, the quantization error will be large in areas where the conversion table values are small, and conversely, the quantization error will be small in areas where the conversion table values are large.

The conversion table may be set by the encoding device according to the frequency of occurrence of the luminance signal values of the original image of one or more slices, or may be selected by the encoding device from among multiple conversion tables predefined by the system, or a conversion table predefined by the system may be used. Note that the conversion table may store values indicating the number of pixel values of the pre-conversion input signal to be assigned to each band obtained by dividing the range from the minimum to maximum possible values of the converted output signal into a predefined number, or the values in the conversion table may be quantized and stored. The conversion table is not limited to the above examples, as long as it shows the relationship between the input signal and output signal before and after conversion.

Furthermore, if the encoding device 1 sets a conversion table according to the frequency of occurrence of luminance signal values or selects from multiple conversion tables, the conversion table information is transmitted to the decoding device 2 by some means. For example, the table value information may be entropy coded and streamed. Furthermore, the encoding device 1 and decoding device 2 may switch between pre-prepared conversion tables based on video format information (for example, parameters representing the relationship between optical and electrical signals in a video signal).

Figure 3 is a diagram showing the configuration of the conversion unit 110 according to this embodiment. As shown in Figure 3, the conversion unit 110 includes a parameter derivation unit 111, a mapping unit 112, and a scaling unit 113.

The parameter derivation unit 111 calculates, based on the conversion table, a mapping scale table, a mapping offset table, and an inverse mapping scale table, which are components of a lookup table used in the mapping process applied to the luminance signal of the block to be coded output from the block division unit 100, and outputs the mapping scale table and mapping offset table to the mapping unit 112, and outputs the mapping offset table and inverse mapping scale table to the inverse conversion unit 170.

Furthermore, the parameter derivation unit 111 uses the calculated inverse mapping scale table to calculate a scaling coefficient table to be used in the scaling process applied to the color difference signal of the block to be coded in the scaling unit 113, and outputs the calculated scaling coefficient table to the scaling unit 113.

The following describes in detail the operation of the parameter derivation unit 111. When the conversion table for the division number N is lmcs _CW , the parameter derivation unit 111 calculates the mapping offset table LmcsPivot, the mapping scale table ScaleCoeff, and the inverse mapping scale table InvScaleCoeff using the following equation (1).

LmcsPivot[0] = 0
for (i=0; i<=N-1; i++){
LmcsPivot[i+1] = LmcsPivot[i]+lmcs _CW [i]
ScaleCoeff[i] = (lmcs _CW [i]*(1<<11)+(1<<(Log2(64)-1)))>>(Log2(64))
if (lmcs _CW [ i ] == 0 )
InvScaleCoeff[i] = 0
else
InvScaleCoeff[i]=64*(1<<11)/lmcs _CW [i]
}
...(1)

where "x<<n" is a left shift operator that shifts x to the left by n places, "x>>n" is a right shift operator that shifts x to the right by n places, and "Log2(x)" is an operator that finds the logarithm of x to the base 2.

Furthermore, the parameter derivation unit 111 calculates the scaling coefficient table ChromaScaleCoeff using the following formula (2):

for (i=0;i<=15;i++){
if (lmcwCW[i]==0)
ChromaScaleCoeff[i]=(1<<11)
else
ChromaScaleCoeff[i]=InvScaleCoeff[i]
}
...(2)

The mapping unit 112 performs mapping processing on the luminance signal of the block to be coded output from the block division unit 100 using the mapping scale table and mapping offset table output from the parameter derivation unit 111, and outputs the generated luminance signal of the block to be coded after the mapping processing to the subtraction unit 120.

Specifically, when the luminance signal in the block to be coded is SamplesY[i][j], the mapping unit 112 uses the mapping scale table ScaleCoeff and the mapping offset table LmcsPivot to calculate the luminance signal MapSamples[i][j] in the block to be coded after the mapping process using the following equation (3):

idxY=SamplesY[i][j]>>Log2(64)
MapSamples[i][j]=LmcsPivot[idxY]+(ScaleCoeff[idxY]*(SamplesY[i][j]-idxY*64)+(1<<10))>>11
...(3)

where i = 0...Wy-1 and j = 0...Hy-1 are the coordinates of the luminance signal within the block to be coded, and Wy and Hy represent the width and height, respectively, of the luminance signal of the block to be coded.

The scaling unit 113 performs scaling processing on the color difference signals of the block to be coded output from the block division unit 100 using the scaling coefficient table output from the parameter derivation unit 111, and outputs the generated color difference signals of the block to be coded after the scaling processing to the subtraction unit 120.

Specifically, when the color difference signal in the block to be coded is SamplesC[i][j] and the average value of the decoded luminance signals adjacent to the block to be coded is Yi, the scaling unit 113 uses the scaling coefficient table ChromaScaleCoeff to calculate the color difference signal ScaleSamples[i][j] in the block to be coded after scaling processing using the following equation (4):

idxYInv = Yi>>Log2(64)
ScaleSamples[i][j]=sign(SampleC[i][j])*((abs(SamplesC[i][j])<<11)+(ChromaScaleCoeff[idxYInv]>>1))/ChromaScaleCoeff[idxYInv]
...(4)

where i = 0...Wc-1, j = 0...Hc-1 are the coordinates of the color difference signal within the block to be coded, and Wc and Hc represent the width and height, respectively, of the color difference signal within the block to be coded. Also, "sign(x)" is a sign function that returns a value of -1, 0, or 1 depending on the sign of the variable x, and "abs(x)" is an operator that returns the absolute value of the variable x.

In this way, the conversion unit 110 outputs to the subtraction unit 120 a new converted block to be coded, which is composed of the luminance signal of the block to be coded after mapping calculated by the mapping unit 112 and the chrominance signal of the block to be coded after scaling calculated by the scaling unit 113.

The subtraction unit 120 calculates a prediction residual that represents the difference (error) between the new, converted block to be coded output from the conversion unit 110 and the predicted block that the prediction unit 190 predicts from this new block to be coded. The subtraction unit 120 calculates the prediction residual by subtracting each pixel value of the predicted block from each pixel value of the block, and outputs the calculated prediction residual to the transformation/quantization unit 130.

The transform/quantization unit 130 performs transform processing and quantization processing on a block-by-block basis. The transform/quantization unit 130 includes a transform unit 131 and a quantization unit 132.

The transform unit 131 performs a transform process on the prediction residual output from the subtraction unit 120 to calculate transform coefficients for each frequency component, and outputs the calculated transform coefficients to the quantization unit 132. Transform process (conversion) refers to a process of converting a pixel-domain signal into a frequency-domain signal, such as a discrete cosine transform (DCT), a discrete sine transform (DST), a Karhunen-Loeve transform (KLT), or an integer version of any of these. The transform process may also include a transform skip, which adjusts the pixel-domain signal by scaling or the like without converting it into a frequency-domain signal.

The quantization unit 132 quantizes the transform coefficients output from the transform unit 131 using a quantization parameter and a quantization matrix, and outputs the quantized transform coefficients to the entropy coding unit 140 and the inverse quantization/inverse transform unit 150. The quantization unit 132 also outputs information related to the quantization process (specifically, information about the quantization parameter and quantization matrix used in the quantization process) to the entropy coding unit 140, the inverse quantization unit 151, and the filter control unit 181.

Here, the quantization parameter is a parameter for which one value is set for one block. Specifically, the quantization parameter is a parameter that is commonly applied to all transform coefficients within a block, and determines the coarseness (step size) of quantization.

A quantization matrix is a matrix consisting of values set for each component within a block. Specifically, a quantization matrix is a matrix consisting of values (weighting coefficients) set for each component of i x j elements according to the block size, and is used to adjust the quantization coarseness for each component ranging from low to high frequencies of the transform coefficients.

Figure 4 shows an example of a quantization matrix. Figure 4 shows an example where i x j = 4 x 4. In the example shown in Figure 4, the quantization matrix has larger values as the horizontal and vertical orders increase. In such a quantization matrix, transform coefficients associated with elements arranged further to the bottom right, i.e., transform coefficients in higher frequencies, are quantized with lower precision. Therefore, by taking advantage of the human visual characteristic that the lower the frequency, the more sensitive the spatial changes in shade and hue, it is possible to reduce the amount of information in higher frequencies through quantization without degrading subjective image quality.

The quantization matrix can be set for each combination of the prediction mode (intra prediction or inter prediction) of the block to be coded, the block size (e.g., 2x2, 4x4, 8x8, 16x16, 32x32, 64x64), and one luminance and two color difference signals (in the case of RGB signals, the R signal, B signal, and G signal).

The entropy coding unit 140 performs entropy coding on the transform coefficients output from the quantization unit 132, compresses the data, generates a coded stream (bit stream), and outputs the coded stream to the outside of the coding device 1. Entropy coding can use Huffman coding or CABAC (Context-based Adaptive Binary Arithmetic Coding), etc.

The entropy coding unit 140 also obtains information such as the size and shape of each block to be coded from the block division unit 100, obtains information about the quantization process from the quantization unit 132, and obtains information about prediction (e.g., prediction mode and motion vector information) from the prediction unit 190, and encodes this information.

The inverse quantization and inverse transform unit 150 performs inverse quantization and inverse transform processing on a block-by-block basis. The inverse quantization and inverse transform unit 150 includes an inverse quantization unit 151 and an inverse transform unit 152.

The inverse quantization unit 151 performs inverse quantization processing corresponding to the quantization processing performed by the quantization unit 132. Specifically, the inverse quantization unit 151 restores the transform coefficients by inverse quantizing the transform coefficients output from the quantization unit 132 using a quantization parameter and a quantization matrix, and outputs the restored transform coefficients to the inverse transform unit 152.

The inverse transform unit 152 performs inverse transform processing corresponding to the transform processing performed by the transform unit 131. For example, if the transform unit 131 performs DCT, the inverse transform unit 152 performs inverse DCT. The inverse transform unit 152 performs inverse transform processing on the transform coefficients output from the inverse quantization unit 151 to restore the prediction residual, and outputs the restored prediction residual, which is the restored prediction residual, to the synthesis unit 160.

The synthesis unit 160 synthesizes the reconstructed prediction residual output from the inverse transform unit 152 with the predicted block output from the prediction unit 190 on a pixel-by-pixel basis. The synthesis unit 160 adds each pixel value of the reconstructed prediction residual to each pixel value of the predicted block to calculate a synthesized block, and outputs the calculated synthesized block to the inverse transform unit 170.

The inverse transformation unit 170 performs an inverse transformation process corresponding to the transformation process performed by the transformation unit 110. Specifically, the inverse transformation unit 170 restores (decodes) the block to be coded by performing an inverse transformation process corresponding to the transformation process performed by the transformation unit 110 on each pixel value in the composite block output from the synthesis unit 160 based on a transformation table, and outputs the restored decoded image (restored block) in block units to the deblocking filter 180.

Figure 5 is a diagram showing the configuration of the inverse transformation unit 170 according to this embodiment. As shown in Figure 5, the inverse transformation unit 170 includes a de-mapping unit 171 and a de-scaling unit 172.

The inverse mapping unit 171 performs an inverse mapping process on the luminance signal of the composite block output from the synthesis unit 160 using the mapping scale table and mapping offset table output from the parameter derivation unit 111 to restore the luminance signal of the block to be coded, and outputs the restored luminance signal of the block to be coded to the deblocking filter 180.

Specifically, when the luminance signal in the composite block is MapSampleY'[i][j], the inverse mapping unit 171 uses the inverse mapping scale table InvScaleCoeff and the mapping offset table LmcsPivot to calculate the luminance signal RecSmaplesY[i][j] in the block to be coded that has been restored after the inverse mapping process, using the following equation (5):

for (idxYInv = 0; idxYInv<=15;idxYInv++){
if (MapSampleY'[i][j] < LmcsPivot[idxYInv+1])
break
}
idxYInv = Min(idxYInv, 15)
RecSamplesY[i][j]=idxYInv*64+(InvScaleCoeff[idxYInv]*(MapSampleY'[i][j]-LmcsPivot[idxYInv])+(1<<10))>>11
...(5)

where i = 0...Wy-1, j = 0...Hy-1 are the coordinates of the luminance signal within the block to be coded, and Wy and Hy are the width and height of the luminance signal of the block to be coded. Also, "Min(x, y)" is an operator that returns the smaller of x and y.

The inverse scaling unit 172 performs inverse scaling processing on the chrominance signal of the composite block output from the synthesis unit 160 using the scaling coefficient table output from the parameter derivation unit 111 to restore the chrominance signal of the block to be coded, and outputs the restored chrominance signal of the block to be coded to the deblocking filter 180.

Specifically, when the color difference signal within the composite block is ScaleSamplesC[i][j] and the average value of the decoded luminance signals adjacent to the composite block is Yi, the inverse scaling unit 172 uses the scaling coefficient table ChromaScaleCoeff to calculate the color difference signal RecSampleC[i][j] of the restored encoding target block after the inverse scaling process using the following equation (6):

idxYInv = Yi>>Log2(64)
RecSamplesC[i][j]=sign(ScaleSamplesC[i][j])*(abs(ScaleSamplesC[i][j])*ChromaScaleCoeff[idxYInv]+(1<<10))>>11
...(6)

where i = 0...Wc-1 and j = 0...Hc-1 are the coordinates of the color difference signal within the block to be coded, and Wc and Hc represent the width and height, respectively, of the color difference signal of the block to be coded.

In this way, the inverse transformation unit 170 outputs to the deblocking filter 180 the restored block to be coded, which is composed of the luminance signal of the block to be coded restored by the inverse mapping unit 171 and the chrominance signal of the block to be coded restored by the inverse scaling unit 172.

The deblocking filter 180 performs filtering on the block boundary between two blocks consisting of a reconstructed block and an adjacent block adjacent to the reconstructed block, and outputs the filtered reconstructed block to memory 185. The filtering is performed to reduce signal degradation caused by block-by-block processing, and smooths signal gaps at the block boundary between two adjacent blocks. The deblocking filter 180 is generally configured as a low-pass filter that smooths signal fluctuations.

Figure 6 is a diagram showing an example of the operation of the deblocking filter 180 according to this embodiment. In the example shown in Figure 6, the deblocking filter 180 performs filtering on block boundaries for each 8x8 pixel block. The deblocking filter 180 also performs filtering in units of 4 rows or 4 columns. Blocks P and Q shown in Figure 6 are one unit of filtering by the deblocking filter 180, and show an example where the block size is 4x4 pixels. Each of blocks P and Q may also be referred to as a sub-block. Details of the operation of the deblocking filter 180 will be described later.

The filter control unit 181 controls the deblocking filter 180. Specifically, the filter control unit 181 controls the boundary strength (Bs), which indicates whether or not to perform filtering on the block boundaries of the current block pair, and the filter strength of the deblocking filter 180. The boundary strength Bs is a parameter that determines whether or not to apply filtering and the type of filtering. Note that controlling whether or not to perform filtering can be considered as controlling whether or not to set the boundary strength Bs to 1 or greater, or to zero.

The filter control unit 181 controls the deblocking filter 180 based on the fluctuations in pixel values in the region near the boundary between the current block pair, the prediction mode, the quantization parameter, and the value of the motion vector used in motion compensation prediction (inter prediction).

As will be described in more detail below, the filter control unit 181 in this embodiment controls the filter strength of the deblocking filter 180 based on the quantization parameters used in the quantization process in the quantization unit 132 and the inverse transformation process in the inverse quantization unit 151, and the transformation table used in the transformation process in the transformation unit 110 and the inverse transformation process in the inverse transformation unit 170.

In this way, by controlling the filter strength of the deblocking filter taking into account not only the quantization parameter, which is a unique parameter set on a block-by-block basis, but also the conversion table, it becomes possible to control the filter strength of the deblocking filter 180 in response to the occurrence of block noise caused by the conversion process and inverse conversion process, thereby improving the effectiveness of the deblocking filter 180 in reducing block distortion.

Memory 185 accumulates the reconstructed blocks output from the deblocking filter 180 as decoded images on a frame-by-frame basis. Memory 185 outputs the stored decoded images to the prediction unit 190 (inter prediction unit 191).

The prediction unit 190 performs prediction processing on a block-by-block basis to generate a prediction block corresponding to the block to be coded, and outputs the generated prediction block to the subtraction unit 120 and the synthesis unit 160. The prediction unit 190 has an inter prediction unit 191, a prediction conversion unit 192, an intra prediction unit 193, and a switching unit 194.

The inter prediction unit 191 uses the decoded image stored in memory 185 as a reference image to calculate a motion vector using a technique such as block matching, and predicts the block to be coded to generate an inter prediction block. The inter prediction unit 191 selects the most appropriate inter prediction method from inter prediction using multiple reference images (typically bi-prediction) and inter prediction using a single reference image (unidirectional prediction), and performs inter prediction using the selected inter prediction method. The inter prediction unit 191 outputs information related to the inter prediction (motion vectors, etc.) to the entropy coding unit 140 and the filter control unit 181.

For the luminance signal, the inter prediction unit 191 outputs the inter prediction block to the prediction conversion unit 192. On the other hand, for the chrominance signal, the inter prediction unit 191 skips the prediction conversion unit 192 and outputs the inter prediction block to the switching unit 194.

The prediction conversion unit 192 generates a new inter-prediction block by performing the same conversion process as the conversion unit 110 (specifically, the mapping process of the mapping unit 112). The prediction conversion unit 192 performs mapping process based on a conversion table on the luminance signal in the inter-prediction block to generate a new inter-prediction block, and outputs the new inter-prediction block to the switching unit 194.

The intra prediction unit 193 selects the optimal intra prediction mode to apply to the current block from among multiple intra prediction modes, and predicts the current block using the selected intra prediction mode. Specifically, the intra prediction unit 193 generates an intra prediction block by referencing pixel values of synthesized blocks adjacent to the current block from the synthesized blocks output by the synthesis unit 160, and outputs the generated intra prediction block to the switching unit 194. The intra prediction unit 193 also outputs information related to the selected intra prediction mode to the entropy coding unit 140 and the filter control unit 181.

The switching unit 194 switches between the inter-predicted block output from the prediction conversion unit 192 (or the inter-prediction unit 191) and the intra-predicted block output from the intra-prediction unit 193, and outputs either of the predicted blocks to the subtraction unit 120 and the synthesis unit 160.

As such, the encoding device 1 according to this embodiment includes a conversion unit 110 that performs a conversion process on a current block to be encoded to generate a new current block to be encoded; a transform/quantization unit 130 that performs a conversion process and a quantization process on a prediction residual that represents the difference between the new current block to be encoded and a prediction block generated by predicting the new current block to be encoded; an inverse quantization/inverse transform unit 150 that performs an inverse quantization process and an inverse transform process on the transform coefficients generated by the transform/quantization unit to restore the prediction residual; a synthesis unit 160 that synthesizes the restored prediction residual and the prediction block to generate a composite block; an inverse transformation unit that performs an inverse transformation process on the composite block to restore the current block to be encoded; a deblocking filter 180 that performs a filtering process on the boundary between the restored current block to be encoded and an adjacent block adjacent to the current block to be encoded; and a filter control unit 181 that controls the filter strength of the deblocking filter 180 based on the quantization parameters used in the quantization process and inverse quantization process and the transformation table used in the transformation process and inverse transformation process.

Next, the configuration of the filter control unit 181 according to this embodiment will be described. The filter control unit 181 according to this embodiment controls the filter strength of the deblocking filter 180 based on the quantization parameters used in the quantization and inverse quantization processes for the block to be coded, the quantization parameters used in the quantization and inverse quantization processes for the adjacent blocks, the conversion tables used in the conversion and inverse conversion processes for the block to be coded and the adjacent blocks, and the pixel values of the block to be coded and the adjacent blocks. In the following, it is assumed that the block to be coded or its sub-block is block P, and the adjacent block or its sub-block is block Q.

Figure 7 is a diagram showing the configuration of the filter control unit 181 according to this embodiment. As shown in Figure 7, the filter control unit 181 according to this embodiment includes a boundary strength determination unit 181a, a representative value derivation unit 181b, an offset calculation unit 181c, a threshold derivation unit 181d, and a filter strength control unit 181e.

The boundary strength determination unit 181a determines the boundary strength Bs, for example, based on Table 1 below, and outputs the determined value of boundary strength Bs to the threshold derivation unit 181d and the filter strength control unit 181e. In this embodiment, the value of boundary strength Bs is set to 0, 1, or 2. Note that the boundary strengths for the luminance signal and color difference signal blocks may be calculated separately, or the combination of the boundary strengths for the luminance signal and color difference signal blocks may be determined as a single boundary strength.

As shown in Table 1, the boundary strength determination unit 181a sets the value of Bs to 2 when intra prediction is applied to at least one of blocks P and Q.

On the other hand, the boundary strength determination unit 181a sets the Bs value to 1 if inter prediction is applied to both blocks P and Q and at least one of the following conditions (a) to (d) is met, and sets the Bs value to 0 otherwise.

(a) At least one of blocks P and Q contains significant transform coefficients (i.e., non-zero transform coefficients).

(b) The number of motion vectors or reference images of blocks P and Q are different.

(c) The absolute value of the difference between the motion vectors of blocks P and Q is greater than or equal to a threshold value (e.g., 1 pixel).

If the determined value of boundary strength Bs is 0, the boundary strength determination unit 181a controls the deblocking filter 180 so that filtering is not performed on the boundaries of blocks P and Q.

The representative value derivation unit 181b derives a representative value y _pq of the luminance signal based on the luminance signals of the blocks P and Q, and outputs the derived representative value y _pq to the offset calculation unit 181c.

The representative value y _pq derived by the representative value deriving unit 181b is either 1) or 2) below.

1) The representative value y _pq is the average value of the luminance signals of the block P and the block Q.

Specifically, the representative derivation unit 181b determines the average value of the block P and the block Q as the representative value y _pq as shown in the following equation (7).

y _pq =(ΣYp[i][j]+ΣYq[s][t])/(Wp*Hp+Wq*Hq)
... (7)

where Yp and Yq represent the luminance signals contained in block P and block Q, respectively; i = 0...Hp-1, j = 0...Wp-1 are coordinates within block P; s = 0...Hq-1, t = 0...Wq-1 are coordinates within block Q; Wp and Wq represent the widths of block P and block Q, respectively; and Hp and Hq represent the heights of block P and block Q, respectively.

2) The representative value y _pq is the average value of the luminance signals of each of the blocks P and Q near the block boundary.

Specifically, the representative derivation unit 181b sets the representative value _ypq to the average value of the block boundary between block P and block Q. For example, when block P and block Q are horizontally adjacent to each other (i.e., when performing deblocking processing on the vertical block boundary between block P and block Q), the representative value ypq is set to the average value of luminance pixel values located at the block boundary between block P and block Q, as shown in the following equation (8).

y _pq =(Yp[0][Wp-1]+Yp[Hp-1][Wp-1]+Yq[0][0]+Yq[Hq-1][0])>>2
...(8)

Similarly, when blocks P and Q are vertically adjacent (i.e., when performing deblocking processing on the horizontal block boundary between blocks P and Q), the average value of the luminance pixel values located at the block boundary of blocks P and Q is set as the representative value y _pq , as shown in the following equation (9).

y _pq =(Yp[Hp-1][0]+Yp[Hp-1][Wp-1]+Yq[0][0]+Yq[0][Wq-1])>>2
...(9)

The offset calculation unit 181c calculates an offset value based on the representative value y _pq derived by the representative value derivation unit 181b and the conversion table used in the conversion process and inverse conversion process for the block P and the block Q.

Specifically, the offset calculation unit 181c uses the representative value y _pq to calculate an offset value QpOffset for controlling a deblocking filter to be applied to the boundary between block P and block Q, and outputs the calculated offset value QpOffset to the threshold derivation unit 181d. For example, the offset calculation unit 181c calculates the offset value QpOffset using the following equation (10).

QpOffset=log2(lmcs _CW [y _pq >>6]/64)×6
...(10)

In addition, when the offset calculation unit 181c is implemented in hardware, equations (4) and (5) may be implemented using a lookup table instead of division or log2 function processing.

The threshold derivation unit 181d derives thresholds β and tC for controlling the filter strength of the deblocking filter 180 based on the quantization parameter _QpP used in the quantization process and inverse quantization process for the block P, the quantization parameter _QpQ used in the quantization process and inverse quantization process for the block Q, and the offset value QpOffset calculated by the offset calculation unit 181c, and outputs the derived thresholds β and _tC to _the filter strength control unit 181e.

First, the threshold derivation unit 181d calculates the variable qP, for example, using the following equation (11):

qP=((Qp _Q +Qp _P +1)>>1+QpOffset)
...(11)

Formula (11) is basically a calculation formula for finding the average of the quantization parameter _QpP of block P and the quantization parameter _QpQ of block Q, but this calculation formula introduces an offset value QpOffset based on a conversion table. Note that formula (11) is merely an example, and other calculation formulas may be used as long as they reflect the offset value QpOffset.

Second, the threshold derivation unit 181d calculates Q for deriving the threshold β, for example, using the following equation (12):

Q=Clip3(0,63,qP+(slice_beta_offset_div2<<1))
...(12)

where "Clip3(x, y, z)" is a clip operator that returns x if z is less than x, returns y if z is greater than y, and returns z otherwise. "slice_beta_offset_div2" is one of the parameters signaled to the decoding device 2.

Furthermore, the threshold value deriving unit 181d calculates Q for deriving the threshold value _tC , for example, using the following equation (13).

Q=Clip3(0,65,qP+2*(bS?1)+(slice_tc_offset_div2<<1))
...(13)

where "bS" is the Bs value output by the boundary strength determination unit 181a. "slice_tc_offset_div2" is one of the parameters signaled to the decoding device 2.

Third, the threshold value deriving unit 181d derives threshold values β and _tC from the calculated Q according to Table 2 below.

The filter strength control unit 181e controls the filter strength of the deblocking filter 180 based on the value of the boundary strength Bs output by the boundary strength determination unit 181a and the threshold β and _tC output by the threshold derivation unit 181d.

The representative value derivation unit 181b described above sets the average value of the luminance pixel values located at the block boundary of block P and block Q as the representative value _ypq . However, it is also possible to calculate the representative value yp of block _P and the representative value _yq of block Q by averaging a part or the whole of each block, and derive offset values for block P and block Q using the representative values.

Specifically, the representative values _yp and _yq derived by the representative value deriving unit 181b are either 1) or 2) below.

1) The average values of block P and block Q are set as representative values _yp and _yq , respectively.

y _p =(ΣYp[i][j])/(Wp*Hp)
y _q =(ΣYq[s][t])/(Wq*Hq)
...(14)

2) The average values of the luminance pixel values of each block located on the block boundary between block P and block Q are set as representative values _yp and _yq .

y _p =(Yp[0][Wp-1]+Yp[Hp-1][Wp-1]) >>1
y _q =(Yq[0][0]+Yq[Hq-1][0]) >>1
...(15)

The offset calculation unit 181c calculates offset values QppOffset and QpqOffset to be applied to the block P and the block Q, respectively, using the _yp and _yq according to the following equations.

QppOffset=log2(lmcs _CW [y _p >>6]/64)×6
QpqOffset=log2(lmcs _CW [y _q >>6]/64)×6
...(16)

The threshold derivation unit 181d derives thresholds β and tC for controlling the filter strength of the deblocking filter 180 based on the quantization parameters _QpP used in the quantization processing and inverse quantization processing for the block P, the quantization parameters _QpQ used in the quantization processing and inverse quantization processing for the block Q, and the offset values QppOffset and QpqOffset calculated by the offset calculation unit 181c, and outputs the derived thresholds β and _tC to _the filter strength control unit 181e.

The threshold derivation unit 181d calculates the variable qP, for example, using the following equation (17):

qP=(Qp _Q +Qp _P +QppOffset +QpqOffset +1)>>1
...(17)

Using the variable qP calculated as described above, Q for deriving the threshold value β is calculated using equations (12) and (13), and the threshold value β and _tC are derived using the calculated Q and Table 2.

When the value of boundary strength Bs is 1 or 2, the filter strength control unit 181e may control the deblocking filter 180 to perform filtering only when the following equation (18) is satisfied (see Figure 6).

Furthermore, when performing filter processing, the filter strength control unit 181e may apply a strong filter if all of the following conditional expressions (19) to (24) are satisfied, and a weak filter in other cases (see Figure 6).

In this way, the filter control unit 181 according to this embodiment controls the filter strength of the deblocking filter 180 taking into consideration not only the quantization parameter, which is a unique parameter set for each block, but also the conversion table. This makes it possible to control the filter strength of the deblocking filter 180 in response to the occurrence of block noise caused by the conversion process and the inverse conversion process, thereby improving the effectiveness of the deblocking filter 180 in reducing block distortion.

For example, the filter control unit 181 identifies the bands corresponding to the block to be coded and the adjacent blocks from among the multiple bands in the conversion table, and controls the deblocking filter 180 so that the filter strength of the deblocking filter 180 increases as the value indicating the number of converted pixel values assigned to the identified band decreases.

This makes it easier to apply a strong filter to block boundaries in areas with small conversion table values, taking into account the fact that even when the same quantization parameters are applied to an image, the quantization error will be large in areas with small conversion table values, and conversely, the quantization error will be small in areas with large conversion table values.

<Configuration of Decoding Device>
Next, the configuration of the decoding device according to this embodiment will be described, focusing mainly on the differences from the configuration of the encoding device described above.

Figure 8 shows the configuration of a decoding device 2 according to this embodiment. The decoding device 2 is a device that decodes a current block from an encoded stream.

As shown in FIG. 8, the decoding device 2 has an entropy decoding unit 200, an inverse quantization and inverse transform unit 210, a synthesis unit 220, an inverse transformation unit 230, a deblocking filter 240, a filter control unit 241, a memory 250, and a prediction unit 260.

The entropy decoding unit 200 decodes the encoded stream generated by the encoding device 1 and decodes various signaling information. Specifically, the entropy decoding unit 200 obtains information related to the quantization process applied to the block to be decoded and outputs the obtained information to the inverse quantization unit 211 and the filter control unit 241. The entropy decoding unit 200 also obtains information related to the prediction applied to the block to be decoded (e.g., prediction type information, motion vector information) and outputs the obtained information to the prediction unit 260 and the filter control unit 241. When a conversion table is transmitted from the encoding device 1, the entropy decoding unit 200 outputs the conversion table information to the inverse conversion unit 230 and the filter control unit 241.

In addition, the entropy decoding unit 200 decodes the encoded stream, obtains quantized transform coefficients, and outputs the obtained transform coefficients to the inverse quantization/inverse transform unit 210 (inverse quantization unit 211).

The inverse quantization and inverse transform unit 210 performs inverse quantization and inverse transform processing on a block-by-block basis. The inverse quantization and inverse transform unit 210 includes an inverse quantization unit 211 and an inverse transform unit 212.

The inverse quantization unit 211 performs inverse quantization processing corresponding to the quantization processing performed by the quantization unit 132 of the encoding device 1. The inverse quantization unit 211 restores the transform coefficients of the block to be decoded by inverse quantizing the quantized transform coefficients output from the entropy decoding unit 200 using a quantization parameter and a quantization matrix, and outputs the restored transform coefficients to the inverse transform unit 212.

The inverse transform unit 212 performs inverse transform processing corresponding to the transform processing performed by the transform unit 131 of the encoding device 1. The inverse transform unit 212 performs inverse transform processing on the transform coefficients output from the inverse quantization unit 211 to restore the prediction residual, and outputs the restored prediction residual (restored prediction residual) to the synthesis unit 220.

The synthesis unit 220 generates a synthesis block by synthesizing the prediction residual output from the inverse transform unit 212 and the prediction block output from the prediction unit 260 on a pixel-by-pixel basis, and outputs the generated synthesis block to the inverse transform unit 230.

The inverse transformation unit 230 performs operations similar to those of the inverse transformation unit 170 of the encoding device 1. Specifically, the inverse transformation unit 230 performs inverse transformation processing on the composite block output from the synthesis unit 220 based on a transformation table, thereby restoring (decoding) the block to be decoded, and outputs the reconstructed block to the deblocking filter 240.

Figure 9 is a diagram showing the configuration of the inverse transformation unit 230 according to this embodiment. As shown in Figure 9, the inverse transformation unit 230 includes a demapping unit 231 and an inverse scaling unit 232.

The inverse mapping unit 231 performs an inverse mapping process on the luminance signal of the composite block output from the synthesis unit 220 using the mapping scale table and mapping offset table to restore the luminance signal of the block to be decoded, and outputs the restored luminance signal of the block to be decoded to the deblocking filter 180. For example, the inverse mapping unit 231 calculates the restored luminance signal of the block to be decoded after the inverse mapping process using the above equation (5).

The inverse scaling unit 232 performs inverse scaling processing on the chrominance signal of the composite block output from the composition unit 220 using a scaling coefficient table to restore the chrominance signal of the block to be decoded, and outputs the restored chrominance signal of the block to be decoded to the deblocking filter 180. For example, the inverse scaling unit 232 calculates the restored chrominance signal of the block to be decoded after the inverse scaling processing using equation (6).

The deblocking filter 240 operates in the same manner as the deblocking filter 180 of the encoding device 1. The deblocking filter 240 performs filtering on the boundary of a block pair (blocks P and Q) consisting of a reconstructed block output from the inverse transformation unit 230 and a block adjacent to the reconstructed block, and outputs the reconstructed block after filtering to the memory 250.

The filter control unit 241 performs operations similar to those of the filter control unit 181 of the encoding device 1, based on information output from the entropy decoding unit 200. The filter control unit 241 selects the boundary strength Bs using, for example, the method shown in Table 1, and controls the filter strength of the deblocking filter 180 using methods such as those shown in equations (7) to (28).

The memory 250 stores the reconstructed blocks output from the deblocking filter 240 as decoded images on a frame-by-frame basis. The memory 250 outputs the decoded images on a frame-by-frame basis to an external device outside the decoding device 2.

The prediction unit 260 performs prediction on a block-by-block basis. The prediction unit 260 includes an inter prediction unit 261, a prediction conversion unit 262, an intra prediction unit 264, and a switching unit 265.

The inter prediction unit 261 predicts the current block to be decoded by inter prediction, using the decoded image stored in the memory 250 as a reference image. The inter prediction unit 261 generates an inter prediction block by performing inter prediction using the motion vector information output from the entropy decoding unit 200, and outputs the generated inter prediction block to the prediction conversion unit 262.

The prediction conversion unit 262 operates in the same manner as the prediction conversion unit 192 of the encoding device 1. The prediction conversion unit 262 generates a new inter prediction block by performing a conversion process on the inter prediction block output from the inter prediction unit 261 based on a conversion table, and outputs the generated new inter prediction block to the switching unit 265.

Based on the information output from the entropy decoding unit 200, the intra prediction unit 264 generates an intra prediction block by referencing the pixel values of the synthesis block output from the synthesis unit 220 that is adjacent to the block to be coded, and outputs the generated intra prediction block to the switching unit 265.

The switching unit 265 switches between the inter-predicted block output from the prediction conversion unit 262 (or the inter-prediction unit 261) and the intra-predicted block output from the intra-prediction unit 264, and outputs one of the predicted blocks to the synthesis unit 220.

As such, the decoding device 2 according to this embodiment includes an entropy decoding unit 200 that decodes the coded stream and outputs transform coefficients corresponding to the block to be decoded; an inverse quantization and inverse transform unit 210 that performs inverse quantization and inverse transform processing on the transform coefficients output by the entropy decoding unit 200 to restore prediction residuals; a synthesis unit 220 that generates a composite block by synthesizing the restored prediction residuals with a prediction block that predicts the block to be decoded; an inverse transformation unit 230 that restores the block to be decoded by performing inverse transformation processing on the generated composite block based on a transformation table; a deblocking filter 240 that performs filtering on the boundaries between the restored block to be decoded and adjacent blocks adjacent to the block to be decoded; and a filter control unit 241 that controls the filter strength of the deblocking filter 240 based on the quantization parameter used in the inverse quantization processing and the transformation table used in the inverse transformation processing.

Next, the configuration of the filter control unit 241 according to this embodiment will be described. Figure 10 is a diagram showing the configuration of the filter control unit 241 according to this embodiment. As shown in Figure 10, the filter control unit 241 according to this embodiment has a boundary strength determination unit 241a, a representative value derivation unit 241b, an offset calculation unit 241c, a threshold derivation unit 241d, and a filter strength control unit 241e.

The boundary strength determination unit 241a, representative value derivation unit 241b, offset calculation unit 241c, threshold derivation unit 241d, and filter strength control unit 241e perform operations similar to those of the boundary strength determination unit 181a, representative value derivation unit 181b, offset calculation unit 181c, threshold derivation unit 181d, and filter strength control unit 181e of the encoding device 1, respectively.

<Example of operation of the filter control unit>
Next, an example of the operation of the filter control unit 181 and the filter control unit 241 according to this embodiment will be described. Since the filter control unit 181 and the filter control unit 241 perform the same operation, the filter control unit 241 will be used as an example for the description. Fig. 11 is a diagram showing an example of the operation flow of the filter control unit 241 according to this embodiment.

As shown in FIG. 11, in step S1, the boundary strength determination unit 241a determines the boundary strength Bs using the method shown in Table 1 and outputs the determined value of boundary strength Bs to the threshold derivation unit 241d and the filter strength control unit 241e. If the determined value of boundary strength Bs is 0, the processing from step S2 onwards is not performed, and filtering is not performed on the boundaries of blocks P and Q.

In step S2, the representative value derivation unit 241b derives a representative value _ypq of the luminance signals of each of the blocks P and Q, and outputs the derived representative value _ypq to the offset calculation unit 241c. The representative value _ypq is derived using the above-mentioned method 1) (equation (7)) or the above-mentioned method 2) (equations (8) and (9)).

In step S3, based on the representative value y _pq derived by the representative value derivation unit 241b, the offset calculation unit 241c calculates an offset value which is a parameter that controls deblocking filter processing for the boundary between block P and block Q. Specifically, the offset calculation unit 241c calculates the offset value QpOffset according to the above equation (10) using the representative value y _pq and the conversion table used in the inverse conversion processing applied to block P and block Q, and outputs the calculated offset value QpOffset to the threshold derivation unit 241d.

In step S4, the threshold derivation unit 241d calculates the variable qP using the above equation (11) based on the quantization parameter _QpP used in the inverse quantization process for the block P, the quantization parameter _QpQ used in the inverse quantization process for the block Q, and the offset value QpOffset calculated by the offset calculation unit 241c.

In step S5, the threshold derivation unit 241d calculates the variable Q using the above equations (12) and (13) based on the variable qP, and derives the thresholds β and _tC from the calculated variable Q using Table 2.

In step S6, the filter strength control unit 241e controls the filter strength of the deblocking filter 240 based on the value of the boundary strength Bs output by the boundary strength determination unit 241a and the threshold β and _tC output by the threshold derivation unit 241d.

In this way, the filter control unit 241 according to this embodiment controls the filter strength of the deblocking filter 240 taking into consideration not only the quantization parameter, which is a unique parameter set on a block-by-block basis, but also the conversion table used in the inverse conversion process. This makes it possible to control the filter strength of the deblocking filter 240 in response to the occurrence of block noise caused by the conversion process and the inverse conversion process, thereby improving the effectiveness of the deblocking filter 240 in reducing block distortion.

For example, the filter control unit 241 identifies a band corresponding to the block to be decoded and an adjacent block from among multiple bands in the conversion table, and controls the deblocking filter 240 so that the filter strength of the deblocking filter 180 increases as the value indicating the number of converted pixel values assigned to the identified band decreases.

This makes it easier to apply a strong filter to block boundaries in areas with small conversion table values, taking into account the fact that even when the same quantization parameters are applied to an image, the quantization error will be large in areas with small conversion table values, and conversely, the quantization error will be small in areas with large conversion table values.

<Other embodiments>
In the above-described embodiment, an example has been described in which the representative value derivation units 181b and 241b derive the representative value y _pq based on the luminance signals of the blocks P and Q. However, the representative value derivation units 181b and 241b may derive a representative value y _pq(P) of the luminance signal of the block P and a representative value y _pq(Q) of the luminance signal of the block Q.

In this case, the offset calculation units 181c and 241c calculate the offset value QpOffsetP of block P using the representative value _ypq(P) , and calculate the offset value _QpOffsetQ of block Q using the representative value _ypq(Q) , for example, according to the following equation (25 ₎ .

QpOffset _P =log2(lmcs _CW [y _pq(P) >>6]/64)×6
QpOffset _Q =log2(lmcs _CW [y _pq(Q) >>6]/64)×6
...(25)

The threshold derivation units 181d and 214d then calculate the variable qP using, for example, the following equation (26). Subsequent operations are the same as in the above-described embodiment.

qP=((Qp _Q + QpOffset _Q + QpOffset _P +Qp _P +1)>>1)
...(26)

A program may be provided that causes a computer to execute each process performed by the encoding device 1. A program may be provided that causes a computer to execute each process performed by the decoding device 2. The program may be recorded on a computer-readable medium. Using a computer-readable medium makes it possible to install the program on a computer. Here, the computer-readable medium on which the program is recorded may be a non-transitory recording medium. The non-transitory recording medium is not particularly limited, but may be, for example, a CD-ROM, DVD-ROM, or other recording medium.

The circuits that perform each process performed by the encoding device 1 may be integrated, and the encoding device 1 may be configured as a semiconductor integrated circuit (chip set, SoC). The circuits that perform each process performed by the decoding device 2 may be integrated, and the decoding device 2 may be configured as a semiconductor integrated circuit (chip set, SoC).

The above describes the embodiments in detail with reference to the drawings, but the specific configuration is not limited to that described above, and various design changes can be made without departing from the spirit of the invention.

1: Encoding device 2: Decoding device 100: Block division unit 110: Conversion unit 111: Parameter derivation unit 112: Mapping unit 113: Scaling unit 120: Subtraction unit 130: Transform/quantization unit 131: Conversion unit 132: Quantization unit 140: Entropy coding unit 150: Inverse quantization/inverse transform unit 151: Inverse quantization unit 152: Inverse transform unit 160: Synthesis unit 162: Prediction conversion unit 170: Inverse transform unit 171: Inverse mapping unit 172: Inverse scaling unit 180: Deblocking filter 181: Filter control unit 181a: Boundary strength determination unit 181b: Representative value derivation unit 181c: Offset calculation unit 181d: Threshold derivation unit 181e: Filter strength control unit 185: Memory 190: Prediction unit 191 : Inter prediction unit 192 : Prediction conversion unit 193 : Intra prediction unit 194 : Switching unit 200 : Entropy decoding unit 210 : Inverse transform unit 211 : Inverse quantization unit 212 : Inverse transform unit 220 : Synthesis unit 230 : Inverse transform unit 231 : Inverse mapping unit 232 : Inverse scaling unit 240 : Deblocking filter 241 : Filter control unit 241 a : Boundary strength determination unit 241 b : Representative value derivation unit 241 c : Offset calculation unit 241 d : Threshold derivation unit 241 e : Filter strength control unit 250 : Memory 260 : Prediction unit 261 : Inter prediction unit 262 : Prediction conversion unit 264 : Intra prediction unit 265 : Switching unit

Claims (2)

A decoding device that decodes an encoded stream in units of blocks generated by dividing an image,
an entropy decoding unit that decodes the coded stream to output transform coefficients corresponding to a block to be decoded;
an inverse quantization and inverse transform unit that performs inverse quantization and inverse transform on the transform coefficients output by the entropy decoding unit to restore prediction residuals;
a synthesis unit that synthesizes the reconstructed prediction residual and a prediction block generated by predicting the current block to be decoded to generate a synthesis block;
a mapping unit that performs a mapping process on each pixel value in the generated composite block based on a conversion table;
a deblocking filter that performs a filtering process on a boundary between the restored decoding target block and an adjacent block adjacent to the decoding target block;
a filter control unit that controls a filter strength of the deblocking filter based on a quantization parameter used in the inverse quantization process and a value related to a pixel value after the mapping process,
The conversion table is a table for converting a pixel value to be converted into a different pixel value. A decoding method for decoding an encoded stream in units of blocks generated by dividing an image, comprising:
decoding the coded stream to output transform coefficients corresponding to a block to be decoded;
performing inverse quantization and inverse transform processing on the output transform coefficients to restore a prediction residual;
generating a composite block by combining the reconstructed prediction residual with a prediction block generated by predicting the current block;
performing a mapping process on each pixel value in the generated composite block based on a conversion table;
a step of performing a filtering process on a boundary between the restored decoding target block and an adjacent block adjacent to the decoding target block;
controlling a filter strength of a deblocking filter used in the filtering process based on a quantization parameter used in the inverse quantization process and a value related to a pixel value after the mapping process;
The conversion table is a table for converting a pixel value to be converted into a different pixel value.