Coding device, decoding device, coding method and decoding method

Abstract

The encoding device (1) includes: a transform unit (13) which performs an orthogonal transform process on a residual image representing a difference between an input image and a predicted image predicting the input image and calculates an orthogonal transform coefficient; a quantization unit (14) which quantizes the orthogonal transform coefficient based on a quantization parameter and generates a quantized coefficient; an entropy encoding unit (24) which encodes the quantized coefficient and generates encoded data; an image decoding unit (10) which restores the orthogonal transform coefficient from the quantized coefficient based on the quantization parameter, adds the predicted image to the residual image restored by inverse orthogonal transforming the orthogonal transform coefficient to generate a reconstructed image; and a deblocking filter unit (18) which performs a filtering process on the reconstructed image, wherein the deblocking filter unit (18) controls the filtering strength according to the brightness signal level of the reconstructed image and the quantization parameter.

Description

[0001] This application is a divisional application of invention patent application No. 201780073676.5, filed on November 27, 2017. Technical Field

[0002] The present disclosure relates to an encoding device, a decoding device, an encoding method, and a decoding method that perform a deblocking filtering process. Background Art

[0003] In mainstream video coding methods represented by MPEG, video units such as frames (or pictures) are divided into small block-shaped areas from the perspective of easy real-time processing, and encoding (compression) is performed using techniques such as transformation and prediction in units of blocks. In this coding method, due to the difference in coding control between adjacent blocks, poor quality at the block boundaries is perceived as distortion. In recent years, in the coding methods called H.264 / AVC and H.265 / HEVC, in order to reduce this coding distortion, a process called deblocking filtering is used.

[0004] The signal degradation caused by the quantization of the orthogonal transformation coefficients at the boundary of two adjacent blocks causes a steep signal fluctuation in the adjacent area that should have been smooth, resulting in block distortion. The deblocking filter used to reduce this distortion is usually designed as a low-pass filter that smoothes the signal fluctuation. Since the amount of signal degradation fluctuates depending on the quantization roughness, the filtering strength is controlled by a quantization parameter that specifies the quantization roughness (for example, see Non-Patent Document 1).

[0005] Prior art literature

[0006] Non-patent literature

[0007] Non-patent document 1: Recommendation ITU-TH.265, (04 / 2013), “High efficiency video coding”, International Telecommunication Union, April 2013 Summary of the invention

[0008] An encoding device according to one embodiment is an encoding device for encoding an input image, characterized in that it includes: a transform unit that performs an orthogonal transform process on a residual image representing a difference between an input image and a predicted image of the input image and calculates an orthogonal transform coefficient; a quantization unit that quantizes the orthogonal transform coefficient based on a quantization parameter and generates a quantized coefficient; an entropy encoding unit that encodes the quantized coefficient and generates encoded data; an image decoding unit that restores the orthogonal transform coefficient from the quantized coefficient based on the quantization parameter, adds the predicted image to the residual image restored by inverse orthogonal transforming the orthogonal transform coefficient to generate a reconstructed image; and a deblocking filter unit that performs a filtering process on the reconstructed image. It is characterized in that the deblocking filter unit controls the filtering strength according to the brightness signal level of the reconstructed image and the quantization parameter.

[0009] A decoding device according to one embodiment is a decoding device for decoding coded data of an input image, wherein the device comprises: an entropy decoding unit that decodes the coded data and obtains quantized coefficients obtained by quantizing orthogonal transform coefficients; an image decoding unit that restores orthogonal transform coefficients from the quantized coefficients based on quantization parameters, adds a predicted image to a residual image restored by inverse orthogonal transforming the orthogonal transform coefficients to generate a reconstructed image; and a deblocking filter unit that performs filtering processing on the reconstructed image. The deblocking filter unit controls the filtering strength according to the brightness signal level of the reconstructed image and the quantization parameter. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 is a diagram showing the correspondence between the signal level and the brightness level.

[0011] Figure 2 : is a block diagram showing a configuration example of an encoding device according to an embodiment of the present disclosure.

[0012] Figure 3 FIG. 1 is a diagram showing a block boundary on which a deblocking filtering process is performed.

[0013] Figure 4 : is a block diagram showing a configuration example of a decoding device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0014] As a feature of new video media, HDR (high dynamic range) signals that expand the range of expression of black and white that cannot be expressed using video signals are standardized. In this HDR signal, compared with the conventional SDR (Standard Dynamic Range) signal, recording is performed from a lower light intensity to a higher light intensity (i.e., from a dark position to a bright position) within a limited bit depth, so a signal suppression process called γ correction that is more extreme than the conventional method is applied. As HDR methods, there are currently the HLG (Hybrid-Log Gamma) method specified in the ARIB STD-B67 standard and the PQ (Perceptual Quantize) method specified in the SMPTE ST.2084, which are internationally standardized in the ITU-R as Recommendation ITU-R BT.2100. It is worth noting that in the future, other methods besides these methods may also be specified.

[0015] Figure 1 In the figure, for SDR signals and HDR signals in the HLG and PQ modes, the corresponding relationship between the brightness signal level of the image and the display brightness level displayed on the display device (inverse gamma correction) is shown. Note that since the HLG mode is a relative system in which the peak brightness of the display device is the maximum value, the peak brightness is represented as 1000 cd / m 2 Similarly, since the conventional SDR signal is also a relative system that sets the peak brightness of the display device to the maximum value, taking the current commercially available display as an example, the peak brightness is expressed as 250cd / m 2 .

[0016] In HDR signals, the display brightness level (brightness) changes more significantly relative to the change in the brightness signal level than in conventional SDR signals. Therefore, in areas with high signal levels, block noise caused by quantization distortion is exaggerated compared to conventional signal degradation. In particular, in the PQ mode, which has the highest degree of signal suppression, the display brightness level changes significantly relative to the brightness signal level, and the effect of block distortion on signal degradation is also significant. In order to reduce this phenomenon, measures are usually taken to reduce the quantization parameter according to the signal level.

[0017] In the conventional deblocking filter, the threshold for switching the filter strength for achieving inter-block smoothing is determined in advance in units of frames according to the quantization parameter, and the control corresponding to the brightness signal level is not implemented. Therefore, in the HDR signal, the nonlinearity between the difference in brightness signal level and the difference in display brightness level is strong, and it is difficult for the conventional deblocking filter to fully reduce the block distortion. In addition, since the conventional SDR signal also has nonlinearity caused by gamma correction, there is a problem that the effect of deblocking filtering is reduced in areas with higher brightness levels.

[0018] Furthermore, although the average error of the block can be reduced by excessively reducing the quantization parameter, the effect of the deblocking filter is also reduced due to the reduction of the quantization parameter, and the difference in the brightness signal level at the boundary is not sufficiently improved. In addition, if the quantization parameter is reduced, the amount of information increases, so there is a problem that the compression effect is also reduced.

[0019] In view of this situation, an object of the present disclosure is to provide an encoding device, a decoding device, an encoding method, and a decoding method that can reduce block distortion even for a high-brightness portion of an SDR signal and an HDR signal without losing a compression effect.

[0020] According to one embodiment, an encoding device is an encoding device for encoding an input image, wherein the encoding device includes: a transform unit that performs an orthogonal transform process on a residual image representing a difference between an input image and a predicted image of the input image and calculates an orthogonal transform coefficient; a quantization unit that quantizes the orthogonal transform coefficient based on a quantization parameter and generates a quantized coefficient; an entropy encoding unit that encodes the quantized coefficient and generates encoded data; an image decoding unit that restores the orthogonal transform coefficient from the quantized coefficient based on the quantization parameter, adds the predicted image to the residual image restored by inverse orthogonal transforming the orthogonal transform coefficient to generate a reconstructed image; and a deblocking filter unit that performs a filtering process on the reconstructed image. The deblocking filter unit controls the filtering strength according to the brightness signal level of the reconstructed image and the quantization parameter.

[0021] A decoding device according to an embodiment is a decoding device for decoding coded data of an input image, wherein the decoding device includes: an entropy decoding unit that decodes the coded data and obtains quantized coefficients obtained by quantizing orthogonal transform coefficients; an image decoding unit that restores orthogonal transform coefficients from the quantized coefficients based on quantization parameters, adds a predicted image to a residual image restored by inverse orthogonal transforming the orthogonal transform coefficients to generate a reconstructed image; and a deblocking filtering unit that performs filtering processing on the reconstructed image. The deblocking filtering unit controls the filtering strength according to the brightness signal level of the reconstructed image and the quantization parameter.

[0022] According to such an encoding device and decoding device, the filtering strength can be changed according to the brightness signal level, and block distortion can be reduced not only for SDR signals but also for HDR signals without losing the compression effect.

[0023] Hereinafter, one embodiment will be described in detail with reference to the accompanying drawings.

[0024] (Encoding device)

[0025] A coding device according to an embodiment will be described below. Figure 2 A configuration example of an encoding device according to an embodiment is shown. Figure 2 The encoding device 1 shown includes a block division unit 11, a subtraction operation unit 12, a transform unit 13, a quantization unit 14, an inverse quantization unit 15, an inverse transform unit 16, an addition operation unit 17, a deblocking filter unit 18, a sample adaptive offset unit (Sample Adaptive Offset Unit) 19, a storage unit 20, an intra prediction unit (intraprediction Unit) 21, a motion compensation prediction unit 22, a switching unit 23, and an entropy encoding unit 24. Note that, instead of the sample adaptive offset unit 19 or in addition to the sample adaptive offset unit 19, a processing unit that performs other post-filtering processing may also be provided.

[0026] The block division unit 11 divides the encoding target frame as the input image into a plurality of blocks and outputs the block images to the subtraction operation unit 12. The size of the block may be variable, for example, 32×32 pixels, 16×16 pixels, 8×8 pixels, or 4×4 pixels.

[0027] The subtraction unit 12 subtracts each pixel value of the predicted image of the input image from each pixel value of the block image input from the block division unit 11, generates a residual image representing the difference between the block image and the predicted image, and outputs it to the transform unit 13. The predicted image is input from the intra prediction unit 21 or the motion compensation prediction unit 22 described later via the switching unit 23.

[0028] The transform unit 13 performs an orthogonal transform process on the residual image input from the subtraction operation unit 12 and calculates an orthogonal transform coefficient, and outputs the orthogonal transform coefficient for each block to the quantization unit 14 .

[0029] The quantization unit 14 quantizes the orthogonal transform coefficient of each block input from the transform unit 13 based on the quantization parameter (qP) that specifies the quantization coarseness to generate a quantized coefficient, and outputs the quantized coefficient to the inverse quantization unit 15 and the entropy encoding unit 24. More specifically, the quantization unit 14 generates the quantized coefficient by dividing the orthogonal transform coefficient of each block input from the transform unit 13 by a quantization step derived from the quantization parameter. For example, the value of the quantization parameter is from 0 to 51, and when the quantization parameter increases by 6, the quantization step correspondingly doubles (that is, the quantization parameter is proportional to the logarithm of the quantization step).

[0030] The inverse quantization unit 15 restores the orthogonal transform coefficient from the quantization coefficient input from the quantization unit 14 based on the quantization parameter, and outputs to the inverse transform unit 16. More specifically, the inverse quantization unit 15 restores the orthogonal transform coefficient of each block by multiplying the quantization coefficient input from the quantization unit 14 by the quantization step size derived from the quantization parameter.

[0031] The inverse transform unit 16 performs inverse orthogonal transform on the orthogonal transform coefficient input from the inverse quantization unit 15 to generate a residual image, and outputs the generated residual image to the addition operation unit 17. For example, when the transform unit 13 performs discrete cosine transform, the inverse transform unit 16 performs inverse discrete cosine transform.

[0032] The addition unit 17 adds the residual image input from the inverse transform unit 16 and each pixel value of the predicted image input from the switching unit 23 to generate a reconstructed image, and outputs the reconstructed image to the deblocking filtering unit 18 .

[0033] The deblocking filter unit 18 performs filtering processing on the reconstructed image input from the addition operation unit 17, and outputs the filtered reconstructed image to the sample adaptive offset unit 19. The deblocking filter unit 18 controls the filtering intensity according to the luminance signal level (pixel value of the luminance component) and the quantization parameter of the reconstructed image. The details of this processing will be described later.

[0034] The sample adaptive offset unit 19 classifies the image input from the deblocking filter unit 18 in units of pixels, adds an offset corresponding to the classification to each pixel value, and outputs the result as a decoded image to the storage unit 20. In addition, the sample adaptive offset unit 19 outputs information of the sample adaptive offset to the entropy coding unit 24.

[0035] The intra prediction unit 21 refers to the decoded image stored in the storage unit 20, performs intra prediction to generate an intra prediction image, and outputs the intra prediction image to the switching unit 23. Furthermore, the intra prediction unit 21 outputs the selected intra prediction mode to the entropy encoding unit 24.

[0036] The motion compensation prediction unit 22 refers to the decoded image stored in the storage unit 20, generates a motion vector by a method such as block matching, and outputs information of the motion vector to the entropy encoding unit 24. In addition, the motion compensation prediction unit 22 generates a motion compensation prediction image based on the motion vector, and outputs the motion compensation prediction image to the switching unit 23.

[0037] The switching unit 23 switches the intra-frame prediction image input from the intra-frame prediction unit 21 and the motion compensation prediction image input from the motion compensation prediction unit 22, and outputs the prediction image of the decoded image (intra-frame prediction image or motion compensation prediction image) to the subtraction unit 12 and the addition unit 17.

[0038] The inverse quantization unit 15, the inverse transform unit 16, the addition unit 17, the intra prediction unit 21, the motion compensation prediction unit 22, and the switching unit 23 constitute the image decoding unit 10. As described above, the image decoding unit 10 generates a reconstructed image by restoring an orthogonal transform coefficient from a quantization coefficient based on a quantization parameter and adding a predicted image to a residual image restored by inversely orthogonally transforming the orthogonal transform coefficient.

[0039] The entropy coding unit 24 performs entropy coding on the quantization coefficient input from the quantization unit 14, the intra prediction mode input from the intra prediction unit 21, the information of the predicted motion vector input from the motion compensation prediction unit 22, the information about filtering input from the deblocking filtering unit 18, and the information of the sample adaptive offset input from the sample adaptive offset unit 19, and performs data compression to generate coded data, and outputs the coded data to the outside of the encoding device 1. Any entropy coding method such as zero-order exponential Golomb coding and CABAC (Context-based Adaptive Binary Arithmetic Coding) can be used for entropy coding.

[0040] (Deblocking Filter Unit)

[0041] The following describes the details of the deblocking filter unit 18. In the present embodiment, the block size processed by the deblocking filter unit 18 is set to 8×8 pixels, for example. First, the deblocking filter unit 18 obtains a boundary strength Bs (Boundary Strength) value indicating the strength of the smoothing process for each block. The Bs value is any one of 0, 1, or 2.

[0042] Figure 3 The block boundaries for deblocking filtering are shown. Figure 3, an example of deblocking filtering processing in accordance with the H.265 / HEVC method is described. If block P or Q is a block to be intra-predicted, the Bs value is set to 2. If blocks P and Q are blocks to be inter-predicted and at least one of the following conditions is satisfied, the Bs value is set to 1, and for other cases, the Bs value is set to 0.

[0043] Block P or Q contains significant (non-zero) orthogonal transform coefficients and is the boundary of a transform unit TU (Transform Unit).

[0044] The number or reference images of motion vectors of blocks P and Q are different.

[0045] The absolute value of the difference between the motion vectors of blocks P and Q is 4 pixels or more.

[0046] When the Bs value is 0, the deblocking filter unit 18 does not perform filtering processing. Figure 3 When the Bs value is 1 or 2, filtering is performed only when the following equation (1) is satisfied.

[0047] [Formula 1]

[0048]

[0049] Furthermore, when performing the filtering process, the deblocking filtering unit 18 applies strong filtering when all of the following conditional expressions (2) to (7) are satisfied, and applies weak filtering otherwise.

[0050] [Formula 2]

[0051] 2(|p20-2p10+p00|+|q20-2q10+q00|)<β / 4 (2)

[0052] 2(|p23-2p13+p03|+|q23-2q13+q03|)<β / 4 (3)

[0053] |p30-p00|+|q00-q30|<β / 8 (4)

[0054] |p33-p03|+|q03-q33|<β / 8 (5)

[0055] |p00-q00|<(5t C +1) / 2 (6)

[0056] |p03-q03|<(5t C +1) / 2 (7)

[0057] Threshold β and tC The value of is based on the average value Q of the quantization parameters of the adjacent blocks P and Q av The deblocking filter unit 18 has a Q av With threshold β and t C An example of the reference table is shown in Table 1. The threshold β and t when the bit depth of the image is 8 bits C They are represented as threshold β′ and t C '. Attention, Q av , β′, t C ' may be a value obtained by adding compensation in units of frames or slices.

[0058] Table 1

[0059] <![CDATA[Q av ]]> 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 β′ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 7 8 <![CDATA[t C ′]]> 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 <![CDATA[Q av ]]> 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 β′ 9 10 11 12 13 14 15 16 17 18 20 22 24 26 28 30 32 34 36 <![CDATA[t C ′]]> 1 1 1 1 1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 <![CDATA[Q av ]]> 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 β′ 38 40 42 44 46 48 50 52 54 56 58 60 62 64 - - <![CDATA[t C ′]]> 5 5 6 6 7 8 9 10 11 13 14 16 18 20 22 24

[0060] Assuming the bit depth is B, the threshold β and t C They are represented by equations (8) and (9) respectively. Here, <<1 represents a 1-bit arithmetic left shift operation.

[0061] [Formula 3]

[0062] β=β′*(1<<(B-8)) (8)

[0063] t C =t C ′*(1<<(B-8)) (9)

[0064] In the present disclosure, the deblocking filtering unit 18 shifts the reference table by the newly added luminance signal level (pixel value of the luminance component) L to set the threshold values ​​β and t C The value of the luminance signal level L is set to, for example, (p00+p03+q00+q03) / 4 or (p00+p01+p02+p03+q00+q01+q02+q03) / 8. Note that the method of finding the luminance signal level L is an example, and the present disclosure is not limited thereto.

[0065] The deblocking filter unit 18 obtains a format that specifies the dynamic range of the brightness value of the input image. For the video format, for example, in the H.265 / HEVC method, the format is transmitted through a high-level syntax called a sequence parameter set, and the type of signal can usually be identified by a decoding device. In this embodiment, the format is, for example, three types: SDR, HLG, and PQ. Considering that the difference in brightness signal level has a greater impact on the display brightness than the SDR method in HDR methods such as the PQ method or the HLG method, the deblocking filter unit 18 controls the filtering strength according to the brightness signal level L. Specifically, the β′ and t shown in Table 1 are set.C ′ is set to Q av is the index of β′[Q av ] and t C ′[Q av ], the threshold β′[Q av -a] and t C ′[Q av -b]. For example, the shift amounts a and b are determined by the reference of equation (10). Equation (10) shows an example of the shift amounts a and b when the luminance signal level L is normalized to 0 to 1.0. In addition, the deblocking filter unit 18 may determine the shift amounts a and b according to the required conditions of each application, or may simply set a=b. In addition, a=0 or b=0 may be set, and only the shift thresholds β′ and t C 'one of the.

[0066] In the case of SDR signals,

[0067] if(L<0.75)

[0068] a=0;

[0069] else if(L>=0.75)

[0070] a=2;

[0071] if(L<0.75)

[0072] b = 0;

[0073] else if(L>=0.75)

[0074] b = 1;

[0075] In the case of HLG signal,

[0076] if(L<0.5)

[0077] a=0;

[0078] else if(L>=0.5&&L<0.75)

[0079] a=2;

[0080] else if(L>=0.75)

[0081] a=5;

[0082] if(L<0.5)

[0083] b = 0;

[0084] else if(L>=0.5&&L<0.7)

[0085] b = 3;

[0086] else if (L >= 0.7)

[0087] b = 6;

[0088] In the case of the PQ signal,

[0089] if (L < 0.3)

[0090] a = -1;

[0091] else if (L >= 0.3 && L < 0.4)

[0092] a = 0;

[0093] else if (L >= 0.4 && L < 0.5)

[0094] a = 1;

[0095] else if (L >= 0.5 && L < 0.7)

[0096] a = 5;

[0097] else if (L >= 0.7 && L < 0.85)

[0098] a = 10;

[0099] else if (L >= 0.85 && L < 0.95)

[0100] a = 12;

[0101] else if (L >= 0.95)

[0102] a = 15;

[0103] if (L < 0.3)

[0104] b = -1;

[0105] else if (L >= 0.3 && L < 0.5)

[0106] b = 0;

[0107] else if (L >= 0.5 && L < 0.7)

[0108] b = 1;

[0109] else if (L >= 0.7 && L < 0.95)

[0110] b = 5;

[0111] else if (L >= 0.95)

[0112] b = 7; (10)

[0114] For example, when the shift amount a=b=5, the deblocking filtering unit 18 refers to the threshold values ​​β′ and t C ' is only shifted to the left by 5. The results are shown in Table 2. Furthermore, if the shift amount a is negative, it is shifted to the right. Note that, as shown in the table, when the threshold β' and t C ′ becomes 0, the subsequent threshold is also set to the lower limit 0. If the threshold β′ and t C ' reaches the upper limit values ​​64 and 24, then compensation is performed with the upper limit values ​​thereafter.

[0115] Table 2

[0116] <![CDATA[Q av ]]> 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 β′ 0 0 0 0 0 0 0 0 0 0 0 6 7 8 9 10 11 12 13 <![CDATA[t C ′]]> 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 <![CDATA[Q av ]]> 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 β′ 14 15 16 17 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 <![CDATA[t C ′]]> 1 1 1 2 2 2 2 3 3 3 3 4 4 4 6 6 6 6 7 <![CDATA[Q av ]]> 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 β′ 48 50 52 54 56 58 60 62 64 64 64 64 64 64 - - <![CDATA[t C ′]]> 8 9 10 11 13 14 16 18 20 22 24 24 24 24 24 24

[0117] The deblocking filter unit 18 uses the threshold β′ and t C ' is shifted according to the shift amount a specified corresponding to the brightness signal level L, so that the av The threshold for switching the filter strength. When the threshold β′ and t C ′ shifts to the right, the threshold β and t C The value of becomes larger, so it is easy to satisfy the above conditional expressions (2) to (7), and it is easy to apply strong filtering. Therefore, in the encoding device 1 according to the present disclosure, when the luminance signal level L is large, it is easy to apply strong filtering in the deblocking filter unit 18, and as a result, the occurrence of block distortion can be reduced without losing the compression effect.

[0118] As described above, in the encoding device 1, the deblocking filter unit 18 controls the filter strength according to the brightness signal level of the reconstructed image and the quantization parameter. In one embodiment, the deblocking filter unit 18 maintains the filter strength setting information (reference table), which indicates the first value (Q av ) and the second value (β, t C ). The deblocking filter unit 18 corrects the filter strength setting information according to the brightness signal level of the reconstructed image, and derives the second value by applying the first value to the corrected filter strength setting information. The first value is a value determined by the average value of the quantization parameters of two adjacent blocks in the reconstructed image. The second value is a value that functions as a threshold for switching the filter strength. The deblocking filter unit 18 uses the second value to switch the filter strength.

[0119] In one embodiment, the deblocking filter unit 18 controls the filter strength according to a format that specifies the dynamic range of the brightness value of the input image. The deblocking filter unit 18 holds filter strength setting information (reference table) indicating the first value (Q calculated from the quantization parameter) and correction information (see equation (10)). av ) and the second value (β, t C ), the correction information defining a correction method of the filter strength setting information for each format. The deblocking filtering unit 18 determines a correction method corresponding to the format applied to the input image based on the correction information, corrects the filter strength setting information according to the luminance signal level of the reconstructed image using the determined correction method, and derives the second value by applying the first value to the corrected filter strength setting information.

[0120] Note that a computer can be preferably used as the above-mentioned encoding device 1. Such a computer stores a program in a storage unit of the computer, and the program is used to describe the processing content for realizing each function of the encoding device 1. The CPU of the computer can read and execute the program to realize the function. Furthermore, the program can be recorded on a computer-readable recording medium.

[0121] (Decoding device)

[0122] Next, a decoding device according to an embodiment will be described. Figure 4 is a block diagram showing a configuration example of a decoding device according to an embodiment. Figure 4 The decoding device 2 shown includes an entropy decoding unit 31, an inverse quantization unit 32, an inverse transform unit 33, an addition operation unit 34, a deblocking filter unit 35, a sample adaptive offset unit 36, a storage unit 37, an intra-frame prediction unit 38, a motion compensation prediction unit 39, and a switching unit 40. The decoding device 2 decodes the encoded data of the input image encoded by the encoding device 1.

[0123] The entropy decoding unit 31 decodes the encoded data output from the encoding device 1, and obtains the quantization coefficient, the intra prediction mode, the motion prediction information, the information about the filter, and the information of the sample adaptive offset. Then, the entropy decoding unit 31 outputs the quantization coefficient to the inverse quantization unit 32, outputs the intra prediction mode to the intra prediction unit 38, outputs the motion prediction information to the motion compensation prediction unit 39, outputs the information about the filter to the deblocking filtering unit 35, and outputs the information of the sample adaptive offset to the sample adaptive offset unit 36.

[0124] The inverse quantization unit 32 inputs the quantization coefficient and the quantization parameter from the entropy decoding unit 31 , multiplies the quantization coefficient by the quantization step size derived from the quantization parameter, restores the orthogonal transformation coefficient for each block, and outputs the orthogonal transformation coefficient to the inverse transformation unit 33 .

[0125] The inverse transform unit 33 inversely transforms the orthogonal transform coefficient input from the inverse quantization unit 32 to generate a residual image, and outputs the residual image to the addition operation unit 34 .

[0126] The addition unit 34 adds the residual image input from the inverse transform unit 33 and each pixel value of the predicted image input from the switching unit 40 to generate a reconstructed image, and outputs the reconstructed image to the deblocking filtering unit 35 .

[0127] The deblocking filter unit 35 performs filtering processing on the reconstructed image input from the addition operation unit 34, and outputs the filtered reconstructed image to the sample adaptive offset unit 36. The deblocking filter unit 35 controls the filtering intensity according to the luminance signal level (pixel value of the luminance component) of the reconstructed image and the quantization parameter.

[0128] The sample adaptive offset unit 36 ​​adds compensation to the image input from the deblocking filtering unit 35 according to the sample adaptive offset information input from the entropy decoding unit 31 , and outputs the result to the storage unit 37 as a decoded image.

[0129] The storage unit 37 stores an image of one frame, and outputs the stored image to the outside of the decoding device 2 .

[0130] The intra prediction unit 38 refers to the decoded image stored in the storage unit 37 , performs prediction processing according to the intra prediction mode input from the entropy decoding unit 31 and generates an intra prediction image, and outputs the intra prediction image to the switching unit 40 .

[0131] The motion compensation prediction unit 39 refers to the decoded image stored in the storage unit 37 , performs prediction processing according to the information of the motion vector input from the entropy decoding unit 31 , and generates a motion compensation prediction image, and outputs the motion compensation prediction image to the switching unit 40 .

[0132] The switching unit 40 switches between the intra prediction image input from the intra prediction unit 38 and the motion compensation prediction image input from the motion compensation prediction unit 39 , and outputs the intra prediction image or the motion compensation prediction image to the addition unit 34 .

[0133] The image decoding unit 30 is composed of the inverse quantization unit 32, the inverse transform unit 33, the addition unit 34, the intra prediction unit 38, the motion compensation prediction unit 39, and the switching unit 40. As described above, the image decoding unit 30 generates a reconstructed image by restoring the orthogonal transform coefficients from the quantization coefficients based on the quantization parameters and adding the predicted image of the decoded image to the residual image restored by inverse orthogonal transforming the orthogonal transform coefficients.

[0134] The processing of the deblocking filtering unit 35 is the same as that of the deblocking filtering unit 18. That is, the deblocking filtering unit 35 performs the same processing as that of the deblocking filtering unit 18 by making the threshold values ​​β′ and t C ' is shifted according to the shift amount a specified corresponding to the brightness signal level L, so that the av The threshold for switching the filter strength. When the threshold β′ and t C ′ shifts to the right, the threshold β and t C The value of becomes larger, making it easy to satisfy the above conditional expressions (2) to (7), and it is easy to apply strong filtering. Therefore, in the decoding device 2 according to the present disclosure, when the luminance signal level L is large, it is easy to apply strong filtering in the deblocking filter unit 35, and as a result, the occurrence of block distortion can be reduced without losing the compression effect.

[0135] As described above, in the decoding device 2, the deblocking filter unit 35 controls the filter strength according to the brightness signal level and the quantization parameter of the reconstructed image. In one embodiment, the deblocking filter unit 35 maintains the filter strength setting information (reference table), which indicates the first value (Q av ) and the second value (β, t C ). The deblocking filter unit 35 corrects the filter strength setting information according to the brightness signal level of the reconstructed image, and derives the second value by applying the first value to the corrected filter strength setting information. The first value is a value determined by the average value of the quantization parameters of two adjacent blocks in the reconstructed image. The second value is a value that functions as a threshold for switching the filter strength. The deblocking filter unit 35 uses the second value to switch the filter strength.

[0136] In one embodiment, the deblocking filter unit 35 controls the filter strength according to a format that specifies the dynamic range of the brightness value of the input image. The deblocking filter unit 35 maintains filter strength setting information (reference table) and correction information (see formula (10)), the filter strength setting information indicating the correspondence between the first value calculated from the quantization parameter and the second value used to determine the filter strength, and the correction information defines a correction method of the filter strength setting information for each format. The deblocking filter unit 35 determines a correction method corresponding to the format applied to the input image based on the correction information, corrects the filter strength setting information according to the brightness signal level of the reconstructed image using the determined correction method, and derives the second value by applying the first value to the corrected filter strength setting information.

[0137] Note that a computer can be preferably used as the above-mentioned decoding device 2. Such a computer stores a program in a storage unit of the computer, and the program is used to describe the processing content for realizing each function of the decoding device 2. The CPU of the computer can read and execute the program to realize the function. Furthermore, the program can be recorded on a computer-readable recording medium.

[0138] Although the above embodiments have been described as representative examples, it is obvious to those skilled in the art that various modifications and substitutions can be made within the spirit and scope of the present disclosure. Therefore, the present disclosure should not be interpreted as being limited by the above embodiments, and various deformations and modifications can be made without departing from the claims. For example, multiple configuration blocks described in the configuration diagram of the embodiment can be combined into one, or a configuration block can be divided.

[0139] The present disclosure claims priority from Japanese Patent Application No. 2016-230524 (filed on November 28, 2016) and Japanese Patent Application No. 2017-23345 (filed on February 10, 2017), the entire contents of which are hereby incorporated by reference.

[0140] Explanation of symbols

[0141] 1: Encoding device

[0142] 2: Decoding device

[0143] 10: Image decoding unit

[0144] 11: Block segmentation unit

[0145] 12: Subtraction unit

[0146] 13: Transformation unit

[0147] 14: Quantization unit

[0148] 15: Inverse quantization unit

[0149] 16: Inverse transformation unit

[0150] 17: Addition unit

[0151] 18: Deblocking filter unit

[0152] 19: Sample adaptive offset unit

[0153] 20: Storage unit

[0154] 21: Intra prediction unit

[0155] 22: Motion Compensated Prediction Unit

[0156] 23: Switching unit

[0157] 24: Entropy coding unit

[0158] 30: Image decoding unit

[0159] 31: Entropy decoding unit

[0160] 32: Inverse quantization unit

[0161] 33: Inverse transformation unit

[0162] 34: Addition unit

[0163] 35: Deblocking filter unit

[0164] 36: Sample adaptive offset unit

[0165] 37: Storage unit

[0166] 38: Intra prediction unit

[0167] 39: Motion Compensated Prediction Unit

[0168] 40: Switching unit.

Claims

1. A coding device for coding an input image, wherein: include: a transform unit that performs a transform process on a residual image representing a difference between an input image and a predicted image of the input image and calculates a transform coefficient; a quantization unit, which quantizes the transform coefficients based on a quantization parameter and generates quantized coefficients; an entropy coding unit, which encodes the quantized coefficients and generates coded data; an image decoding unit, which restores transform coefficients from the quantized coefficients based on the quantization parameters, and adds the predicted image to a residual image restored by inversely transforming the transform coefficients to generate a reconstructed image; as well as a deblocking filtering unit, which performs filtering processing on the reconstructed image, The deblocking filtering unit calculates a luminance signal level of the reconstructed image, and determines a parameter for controlling the filtering process according to the luminance signal level and the quantization parameter, The deblocking filtering unit: calculating an average value of a pixel value of a luminance component included in one of two adjacent blocks in the reconstructed image and a pixel value of a luminance component included in the other block as the luminance signal level; as well as A luminance signal level range corresponding to the luminance signal level is determined by comparing the luminance signal level to one or more thresholds, and the parameter is determined using an adjustment value associated with the determined luminance signal level range.

2. A decoding device for decoding coded data of an input image, wherein: include: An entropy decoding unit, which decodes the encoded data and obtains quantized coefficients obtained by quantizing the transform coefficients; an image decoding unit, which restores transform coefficients from the quantized coefficients based on the quantization parameters, and adds the predicted image to the residual image restored by inversely transforming the transform coefficients to generate a reconstructed image; as well as a deblocking filtering unit, which performs filtering processing on the reconstructed image, The deblocking filtering unit calculates a luminance signal level of the reconstructed image, and determines a parameter for controlling the filtering process according to the luminance signal level and the quantization parameter, The deblocking filtering unit: calculating an average value of a pixel value of a luminance component included in one of two adjacent blocks in the reconstructed image and a pixel value of a luminance component included in the other block as the luminance signal level; as well as A luminance signal level range corresponding to the luminance signal level is determined by comparing the luminance signal level to one or more thresholds, and the parameter is determined using an adjustment value associated with the determined luminance signal level range.

3. A method for encoding an input image, wherein: include: A step of transforming a residual image and calculating transform coefficients, wherein the residual image represents a difference between an input image and a predicted image of the input image; quantizing the transform coefficients based on quantization parameters and generating quantized coefficients; A step of encoding the quantized coefficients and generating encoded data; A step of restoring transform coefficients from the quantized coefficients based on the quantization parameters; The step of adding the predicted image to a residual image restored by inversely transforming the transform coefficients to generate a reconstructed image; and a step of filtering the reconstructed image, The step of performing the filtering process includes calculating a luminance signal level of the reconstructed image, and determining a parameter for controlling the filtering process according to the luminance signal level and the quantization parameter, The step of determining the parameters comprises: calculating an average value of a pixel value of a luminance component included in one of two adjacent blocks in the reconstructed image and a pixel value of a luminance component included in the other block as the luminance signal level; and The luminance signal level is compared to one or more thresholds to determine a luminance signal level range corresponding to the luminance signal level, and the parameter is determined using an adjustment value associated with the determined luminance signal level range.

4. A method for decoding coded data of an input image, wherein: include: A step of obtaining quantized coefficients obtained by quantizing the transform coefficients; A step of recovering transform coefficients from said quantized coefficients based on quantization parameters; The step of adding the predicted image to the residual image restored by inverse transforming the transform coefficients to generate a reconstructed image; as well as a step of filtering the reconstructed image, The step of performing the filtering process includes calculating a luminance signal level of the reconstructed image, and determining a parameter for controlling the filtering process according to the luminance signal level and the quantization parameter, The step of determining the parameters comprises: calculating an average value of a pixel value of a luminance component included in one of two adjacent blocks in the reconstructed image and a pixel value of a luminance component included in the other block as the luminance signal level; as well as The luminance signal level is compared to one or more thresholds to determine a luminance signal level range corresponding to the luminance signal level, and the parameter is determined using an adjustment value associated with the determined luminance signal level range.

5. A decoding device, wherein: include: an acquisition unit, which acquires a bit stream including a quantization coefficient and a sequence parameter set as a high-level syntax, wherein the quantization coefficient is obtained by quantizing a transform coefficient of a block obtained by dividing the image, The decoding device uses the sequence parameter set acquired by the acquisition unit to set one or more thresholds used in a deblocking filter process for the block; The decoding device generates a reconstructed image using the quantization coefficients acquired by the acquisition unit, The decoding device compares the brightness signal level in the reconstructed image with the one or more thresholds to determine the brightness signal level range to which the brightness signal level belongs, and uses the adjustment value associated with the determined brightness signal level range to determine the parameters for controlling the deblocking filtering process.

Images