JP7279128B2 - Prediction device and program - Google Patents

Description

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

In moving picture (video) coding systems represented by H.265/HEVC (High Efficiency Video Coding), there are two methods: inter prediction using temporal correlation between frames and intra prediction using spatial correlation within frames. After performing prediction while switching the type of prediction and generating a residual signal, it is configured to perform orthogonal transform processing, loop filter processing, and entropy coding processing and output the obtained stream.

In intra prediction in HEVC, a total of 35 types of intra prediction modes, including planar prediction, DC prediction, and directional prediction, are prepared. According to the intra prediction mode determined by the encoder, intra prediction is performed using adjacent decoded reference pixels. configured to do so. Hereinafter, unless otherwise specified, the term "reference pixel" indicates a decoded reference pixel.

Here, in intra prediction in HEVC, a specified value (10 It is configured to create reference pixels to be used when generating a predicted image by filling in "512" in the case of a bit moving image).

In addition, in conventional HEVC, encoding processing is performed in raster scan order from the upper left, so there are cases where reference pixels have not been decoded. In such a case, a prediction image is generated using a value obtained by extrapolating the nearest decoded reference pixel to the 0th order.

In particular, in intra prediction in conventional HEVC, there are many cases where reference pixels located on the lower left side and upper right side of the CU have not been decoded due to encoding processing in raster scan order. However, there is a problem that the prediction accuracy is lowered and the coding efficiency is reduced when the direction prediction is performed from the direction in which there is a .

In order to solve such a problem, in intra prediction, raster scan order (for example, Z type) is used as the encoding processing order for a plurality of transform blocks existing in a CU (hereinafter referred to as "TU: Transform Unit"). In addition, there is known a technique for improving the prediction accuracy by giving a degree of freedom to the order of encoding such as U type and X type (see Non-Patent Document 1).

In addition, the intra prediction used in HEVC is prediction using spatially adjacent upper or left reference pixels. Pixel accuracy tends to be low.

Therefore, when the size of the CU/TU is small, the accuracy of intra prediction tends to be high. Decrease in overall prediction accuracy. In addition, as the size of the CU/TU increases, the possibility of propagation of coding deterioration such as block distortion contained in reference pixels to the predicted image increases.

Therefore, conventional HEVC is configured to apply a smoothing filter to reference pixels according to the size of the CU/TU and the direction of the intra-prediction mode to generate a predicted image.

In general, prediction pixels at positions close to the reference pixels in the prediction image tend to be close to the true values of the original image to be predicted, and the prediction accuracy tends to be high. Pixels tend to deviate from the true values of the original image to be predicted, and the prediction accuracy tends to be low. tends to be higher.

Therefore, in Non-Patent Document 2, in order to increase the prediction accuracy, a predicted image (non-smoothed predicted image) generated by reference pixels before smoothing and a reference pixel after smoothing are generated. By generating a predicted image (smoothed predicted image) and combining the non-smoothed predicted image and the smoothed predicted image using a predetermined weighting factor according to the direction of the intra prediction mode and the size of CU/TU, We propose a method for generating highly accurate predicted images.

An example of such a weighting factor is shown below in FIGS. 12 and 13. FIG.

FIGS. 12(a) and 12(b) show an example of distribution of weighting factors when the intra prediction mode in HEVC is 2, and FIGS. 13(a) and 13(b) show intra prediction in HEVC. An example of the weighting factor distribution when the mode is 18 is shown.

Also, FIGS. 12(a) and 13(a) show examples of weighting coefficients in a predicted image (that is, a non-smoothed predicted image) using pre-smoothing reference pixels, and FIGS. 12(b) and 13( b) shows an example of a weighting factor in a predicted image using smoothed reference pixels (that is, a smoothed predicted image).

In FIGS. 12 and 13, the weighting factor for the black portion is "0" and the weighting factor for the white portion is "1".

As shown in FIGS. 12 and 13, the weight of the non-smoothed predicted image is higher in the predicted pixels located closer to the reference pixels, and the weight of the smoothed predicted images is higher in the predicted pixels located farther from the reference pixels. get higher

In the method of Non-Patent Document 2, the distribution of weighting factors is defined in advance according to the intra prediction mode and the size of the CU/TU, and the weighting factor is switched according to the size of the CU/TU to be encoded and the intra prediction mode. prescribed to apply.

In the figures of this specification, the arrow indicating the direction of intra prediction mode (prediction direction) is
Similar to the description in the HEVC standard, it is assumed that the intra-prediction target pixel is directed to the reference pixel (the same applies hereinafter).

Mochizuki et al., "Applied intra-prediction method based on mean value coordinates", Information Processing Society of Japan research report, vol, 2012-AVM-77, No.12 A. Said, X. Zhao, J. Chen, M. Karczewicz, W.-J. Chien, F. Zhou, "Position dependent intra prediction combination", MPEG doc. m37502 and ITU-T SG16 Doc. COM16-C1 016, October 2015

As described above, the method of Non-Patent Document 2 does not consider the position of the reference pixel used for intra prediction, and when the position of the reference pixel used for intra prediction includes the right side and the bottom side (that is, the right side and the bottom side) Even in the case where a prediction image is generated using adjacent reference pixels on the side), the weighting factor is determined according to the intra prediction mode and the CU/TU size.

That is, in the method of Non-Patent Document 2, the non-smoothed predicted image has higher prediction accuracy at positions close to the reference pixels, and the smoothed predicted image has higher prediction accuracy at positions far from the reference pixels. In spite of this, the weighting factor is determined regardless of the position of the reference pixel used for intra prediction, so there is a problem that the prediction accuracy is lowered and the encoding performance is lowered.

Therefore, the present invention has been made to solve the above-described problems, and it is possible to prevent deterioration in prediction accuracy by appropriately synthesizing a non-smoothed predicted image and a smoothed predicted image when generating a predicted image. An object of the present invention is to provide an encoding device, a decoding device, and a program that can

A first feature of the present invention is an encoding device that is configured to divide and encode an original image in units of frames that constitute a moving image into encoding target blocks, and uses an intra prediction mode to encode An intra prediction unit configured to generate a predicted image, wherein the intra prediction unit generates a non-smoothed predicted image using the intra prediction mode without performing smoothing processing on reference pixels. and a smoothed intra prediction unit configured to generate a smoothed predicted image using the intra prediction mode after performing smoothing processing on the reference pixels a weighting factor determination unit configured to determine weighting factors according to the intra prediction mode and the position of the reference pixel; and using the weighting factors, the non-smoothed predicted image and the smoothed predicted image. and a predicted image generator configured to generate the predicted image by synthesizing the .

A second feature of the present invention is a decoding device configured to divide an original image in units of frames that constitute a moving image into encoding target blocks and decode the predicted image using an intra prediction mode. wherein the intra prediction unit generates a non-smooth predicted image using the intra prediction mode without performing smoothing processing on reference pixels and a smooth intra prediction unit configured to generate a smooth predicted image using the intra prediction mode after performing smoothing processing on the reference pixels. a weighting factor determination unit configured to determine a weighting factor according to the intra prediction mode and the position of the reference pixel; and using the weighting factor, synthesize the non-smoothed predicted image and the smoothed predicted image. and a predicted image generation unit configured to generate the predicted image as described above.

A gist of a third aspect of the present invention is a program for causing a computer to function as the encoding device according to the first aspect.

A gist of a fourth aspect of the present invention is a program for causing a computer to function as the decoding device according to the second aspect.

According to the present invention, it is possible to provide an encoding device, a decoding device, and a program that can prevent a decrease in prediction accuracy by appropriately synthesizing a non-smoothed predicted image and a smoothed predicted image when generating a predicted image. can.

FIG. 1 is a diagram showing an example of functional blocks of an encoding device 1 according to the first embodiment. FIG. 2 is a diagram showing an example of functional blocks of the intra prediction unit 14a of the encoding device 1 according to the first embodiment. FIG. 3 is a diagram showing an example of directions of intra prediction modes used in the first embodiment. FIG. 4 is a diagram illustrating an example of a predicted image generation method according to the first embodiment. FIG. 5 is a diagram illustrating an example of a predicted image generation method according to the first embodiment. FIG. 6 is a flow chart showing an example of the operation of the encoding device 1 according to the first embodiment. FIG. 7 is a flow chart showing an example of the operation of the encoding device 1 according to the first embodiment. FIG. 8 is a flow chart showing an example of the operation of the encoding device 1 according to the first embodiment. FIG. 9 is a flow chart showing an example of the operation of the encoding device 1 according to the first embodiment. FIG. 10 is a diagram showing an example of functional blocks of the decoding device 3 according to the first embodiment. FIG. 11 is a diagram showing an example of functional blocks of the intra prediction unit 33a of the decoding device 3 according to the first embodiment. FIG. 12 is a diagram for explaining the conventional technology. FIG. 13 is a diagram for explaining the conventional technology.

(First embodiment)
An encoding device 1 and a decoding device 3 according to the first embodiment of the present invention will be described below with reference to FIGS. 1 to 11. FIG.

Here, the encoding device 1 and the decoding device 3 according to this embodiment are configured to support intra prediction in a video encoding method such as HEVC. Note that the encoding device 1 and the decoding device 3 according to the present embodiment are configured to be compatible with any video encoding method as long as it is a video encoding method that performs intra prediction.

The encoding device 1 according to the present embodiment is configured to divide an original image of each frame constituting a moving image into CUs and encode the CUs. Also, the encoding device 1 according to the present embodiment may be configured to be able to divide a CU into a plurality of TUs. Hereinafter, in the present embodiment, a case in which a CU is divided into a plurality of TUs will be described as an example. It is also applicable when including the lower side and the right side.

Note that, in the present embodiment, a specified value (“512” for a 10-bit moving image) is set for a CU to be encoded that does not have an adjacent decoded reference pixel, such as the CU located at the upper leftmost position in a frame. is configured to generate reference pixels used when generating a predicted image by filling in, all pixels adjacent to the left side of the CU to be coded can be used as reference pixels.

As shown in FIG. 1, the encoding device 1 according to the present embodiment includes an intra prediction mode determination unit 11, a TU partition determination unit 12, a coding order control unit 13, a sequential local decoded image generation unit 14, It comprises a memory 15 and an entropy coding unit 16 .

The intra-prediction mode determination unit 11 is configured to determine the optimum intra-prediction mode to apply to the CU.

The TU partition decision unit 12 is configured to decide whether to partition a CU into a plurality of TUs. In this embodiment, as a method of dividing a CU into a plurality of TUs, a case of 4 divisions is described as an example. , is not limited to such cases.

The coding order control unit 13 is configured to determine the coding order of the TUs within the CU based on the intra prediction mode (for example, the direction of the intra prediction mode).

For example, when the TU division determination unit 12 determines to divide a CU into a plurality of TUs, the coding order control unit 13 determines that the direction of the intra prediction mode determined by the intra prediction mode determination unit 11 is from upper right to When the direction is toward the lower left (intra prediction modes 2 to 9 shown in FIG. 3) (that is, when directional prediction is performed from the lower left to the upper right), the coding order of the TUs in the CU is the conventional raster Not in the scan order, but in the encoding order of the lower left TU in the CU → the lower right TU in the CU → the upper left TU in the CU → the upper right TU in the CU, or the lower left TU in the CU → the inside of the CU It may be configured to employ a predetermined encoding order among the encoding order of the upper left TU→the lower right TU within the CU→the upper right TU within the CU.

The sequential locally decoded image generation unit 14 is configured to generate a locally decoded image (a decoded image for each TU) based on the coding order determined by the coding order control unit 13 and the method of dividing the CU into TUs. ing.

Specifically, when the TU division determination unit 12 determines to divide the CU into a plurality of TUs, the sequential locally decoded image generation unit 14 follows the coding order determined by the coding order control unit 13. , is configured to sequentially generate a locally decoded image.

As shown in FIG. 1, the sequential local decoded image generation unit 14 includes an intra prediction unit 14a, a residual signal generation unit 14b, an orthogonal transform/quantization unit 14c, and an inverse quantization/inverse orthogonal transformation unit 14d. , and a locally decoded image generator 14e.

The intra prediction unit 14 a is configured to generate a predicted image using the intra prediction mode determined by the intra prediction mode determination unit 11 . That is, the intra prediction unit 14a is configured to determine the positions of reference pixels used when generating a predicted image according to the intra prediction mode, and generate the predicted image using the reference pixels.

Furthermore, the intra prediction unit 14a may be configured to generate a predicted image in the coding order determined by the coding order control unit 13. FIG.

As shown in FIG. 2, the intra prediction unit 14a includes a non-smooth intra prediction unit 14a1, a smooth intra prediction unit 14a2, a weighting factor determination unit 14a3, and a predicted image generation unit 14a4.

The non-smoothed intra prediction unit 14a1 is configured to generate a non-smoothed predicted image using an intra prediction mode without smoothing the reference pixels.

The smooth intra prediction unit 14a2 is configured to generate a smooth predicted image using an intra prediction mode after performing smoothing processing on reference pixels.

The weighting factor determination unit 14a3 is configured to determine a weighting factor according to the intra prediction mode and the position of the reference pixel.

FIG. 3 shows an example of intra prediction modes used in this embodiment. As shown in FIG. 3, in the present embodiment, intra prediction modes 2 to 9 are classified into category A, intra prediction modes 10 to 26 are classified into category B, and intra prediction modes 27 to 34 are classified into category C shall be classified as

Here, when at least one of the lower side and the right side is included as the position of the reference pixel, the weighting factor determination unit 14a3 converts the weighting factor predetermined according to the direction of the intra prediction mode to the weighting factor. It may be configured to be determined as a coefficient.

Specifically, when the lower side and the right side are not included as the position of the reference pixel, the weighting factor determination unit 14a3 determines to use the weighting factor defined in advance according to the direction of the intra prediction mode. may be configured.

That is, when the intra prediction mode belongs to category B, the weighting factor determination unit 14a3 may be configured to use a weighting factor defined in advance according to the direction of the intra prediction mode.

For example, when the intra prediction mode is 18, the weighting factor determination unit 14a3 may be configured to use the weighting factor distributions shown in FIGS. Here, if the distribution of the weighting factors applied to the non-smoothed predicted image is "α", the distribution of the weighting factors applied to the smoothed predicted image is "1-α".

Alternatively, when the intra prediction mode belongs to category A and the position of the reference pixel does not include the lower side (the predicted image is calculated using only the adjacent reference pixels on the left side). generated), it may be configured to determine to use a predetermined weighting factor according to the direction of the intra prediction mode.

For example, the weighting factor determination unit 14a3 is configured to use the weighting factor distribution shown in FIGS. may have been Here, if the distribution of the weighting factors applied to the non-smoothed predicted image is "α", the distribution of the weighting factors applied to the smoothed predicted image is "1-α".

Alternatively, when the intra prediction mode belongs to category C and the right side is not included as the position of the reference pixel, the weight factor determination unit 14a3 determines the weight predetermined according to the direction of the intra prediction mode. It may be configured to determine using coefficients.

On the other hand, when the lower side is included as the position of the reference pixel (when generating the predicted image using the adjacent reference pixels on the lower side), the weighting factor determination unit 14a3 reverses the direction of the intra prediction mode in the vertical direction. It may be configured to determine a weighting factor obtained by vertically inverting a weighting factor defined in advance according to the direction of the movement.

For example, if the intra prediction mode belongs to category A and the positions of the reference pixels include the left side and the bottom side (the adjacent reference pixels on the left side and the bottom side are used, the weighting factor determination unit 14a3 when the prediction image is generated by using the intra-prediction mode), a weighting factor obtained by vertically inverting a weighting factor that is predetermined according to the direction in which the direction of the intra prediction mode is vertically inverted is determined as the weighting factor. may

Here, "reversing the direction of the intra prediction mode in the vertical direction" means converting between the directions of intra prediction modes 2 to 9 and the directions of intra prediction modes 18 to 11, respectively That is, it means that the direction of each intra prediction mode is converted to the direction of the intra prediction mode having a linearly symmetrical positional relationship with respect to the direction of the intra prediction mode 10 .

FIGS. 4(a) and 4(b) show an example of the distribution of weighting factors when the intra prediction mode is 2 and the reference pixel positions include the left side and the lower side, and FIG. a) and FIG. 5(b) show an example of weighting factor distribution when the intra prediction mode is 18. FIG.

4(a) and 5(a) show an example of weighting coefficients in a predicted image (that is, a non-smoothed predicted image) using reference pixels before smoothing, and FIGS. 4(b) and 5( b) shows an example of a weighting factor in a predicted image using smoothed reference pixels (that is, a smoothed predicted image).

In addition, in FIGS. 4 and 5, the weighting coefficient of the black portion is "0" and the weighting coefficient of the white portion is "1".

Here, "inverting the weighting factors in the vertical direction" means converting the weighting factors corresponding to each reference pixel by inverting the distribution of the weighting factors shown in FIGS. 4 and 5 in the vertical direction V. do.

For example, if the weighting factor distribution shown in FIG. 4A is inverted in the vertical direction V, the weighting factor distribution shown in FIG. 5B is obtained, and the weighting factor distribution shown in FIG. Then, the weighting factor distribution shown in FIG. 5(a) is obtained.

That is, when the distribution of the weighting factors is vertically inverted when the intra prediction mode is 2 and the reference pixel positions include the left side and the lower side, the weighting factors when the intra prediction mode is 18 are distribution.

In other words, when the intra prediction mode is 2 and the positions of the reference pixels include the left side and the lower side, the weighting factor determination unit 14a3 uses the weighting factor distributions shown in FIGS. Instead, it is configured to use a weighting factor distribution obtained by vertically inverting the weighting factor distribution when the intra prediction mode is 18 shown in FIG. 5 (the weighting factor distribution shown in FIG. 4).

In addition, "inverting the weighting factor in the vertical direction" means converting the distribution of the weighting factor between α and α' as shown in (Equation 1) and (Equation 2) below. .

Here, "α11" to "αmn", which are the elements of α and α', indicate weighting factors corresponding to respective reference pixels.

Furthermore, when the right side is included as the position of the reference pixel (when generating the predicted image using the adjacent reference pixels on the right side), the weighting factor determination unit 14a3 horizontally reverses the direction of the intra prediction mode. It may be configured such that a weighting factor that is horizontally inverted from a weighting factor that is predetermined according to the direction is determined as the weighting factor.

For example, when the intra prediction mode belongs to category C and the upper and right sides are included as the reference pixel positions (using the upper and right adjacent reference pixels, the weighting factor determination unit 14a3 when generating a prediction image), a weighting factor that is horizontally inverted from a weighting factor that is predetermined according to the direction in which the direction of the intra prediction mode is horizontally inverted is determined as the weighting factor. good too.

Here, "reversing the direction of the intra prediction mode horizontally" means converting between the direction of the intra prediction modes 18 to 25 and the direction of the intra prediction modes 34 to 27 in the example of FIG. , that is, to convert the direction of each intra prediction mode to the direction of the intra prediction mode having a line-symmetrical positional relationship with respect to the direction of the intra prediction mode 26 .

Further, "reversing the weighting coefficients in the horizontal direction" means that the weighting coefficients corresponding to each reference pixel are converted by reversing the distribution of the weighting coefficients shown in FIGS. 4 and 5 in the vertical direction H. .

Alternatively, "horizontally inverting the weighting factors" means transforming the distribution of the weighting factors between α and α'' as shown in the following (Equation 3) and (Equation 4). do.

Here, “α11” to “αmn”, which are elements of α and α″, indicate weighting factors corresponding to respective reference pixels.

The predicted image generator 14a4 is configured to generate a predicted image by combining the non-smoothed predicted image and the smoothed predicted image using the weighting factor determined by the weighting factor determining part 14a3.

The residual signal generation unit 14b is configured to generate a residual signal from the difference between the predicted image generated by the intra prediction unit 14a and the original image.

The orthogonal transformation/quantization unit 14c is configured to perform orthogonal transformation processing and quantization processing on the residual signal generated by the residual signal generation unit 14b, and generate quantized transform coefficients. .

The inverse quantization unit/inverse orthogonal transform unit 14d performs inverse quantization processing and inverse orthogonal transform processing again on the quantized transform coefficients generated by the orthogonal transform/quantization unit 14c to generate a residual signal. configured to generate

The local decoded image generation unit 14e generates a local decoded image by adding the predicted image generated by the intra prediction unit 14a to the residual signal generated by the inverse quantization unit/inverse orthogonal transformation unit 14d. It is configured.

The memory 15 is configured to retain locally decoded images generated by the sequential local decoded image generator 14 so as to be usable as reference images.

The entropy coding unit 16 is configured to perform entropy coding processing on the flag information including the intra prediction mode determined by the intra prediction mode determination unit 11 and the quantized transform coefficients, and output a stream. there is

6 to 9 show flowcharts for explaining an example of the operation of generating a predicted image in the encoding device 1 according to this embodiment.

As shown in FIG. 6, in step S101, when the CU is divided into a plurality of TUs, the operation proceeds to step S102. On the other hand, if the CU is not split into a plurality of TUs in step S101, the operation proceeds to step S104.

In step S102, when the intra prediction mode belongs to category A, the operation proceeds to FIG. 7, and when the intra prediction mode does not belong to category A, the operation proceeds to step S103.

In step S103, if the intra prediction mode belongs to category B, the operation proceeds to FIG. 8, and if the intra prediction mode does not belong to category B If yes), the operation proceeds to FIG.

In step S104, the encoding device 1 generates a predicted image using the left and upper reference pixels.

As shown in FIG. 7, in step S201, if the left side and the lower side are included as the reference pixel positions, the operation proceeds to step S202, and if the lower side is not included as the reference pixel positions, , the operation proceeds to step S204.

In step S<b>202 , the encoding device 1 calculates a weighting factor defined in advance according to the direction obtained by vertically reversing the direction of the intra prediction mode.

In step S203, the encoding device 1 generates a predicted image using weighting factors obtained by vertically inverting the calculated weighting factors.

In step S204, the encoding device 1 calculates a weighting factor defined in advance according to the direction of the intra prediction mode.

In step S205, the encoding device 1 generates a predicted image using the calculated weighting factors.

As shown in FIG. 8, in step S301, the encoding device 1 calculates a weighting factor defined in advance according to the direction of the intra prediction mode.

In step S302, the encoding device 1 generates a predicted image using the calculated weighting factors.

As shown in FIG. 9, in step S401, if the upper and right sides are included as the reference pixel positions, the operation proceeds to step S402, and if the right side is not included as the reference pixel positions, , the operation proceeds to step S404.

In step S402, the encoding device 1 calculates a weighting factor defined in advance according to the direction obtained by horizontally reversing the direction of the intra prediction mode.

In step S403, the encoding device 1 generates a predicted image using weighting factors obtained by horizontally inverting the calculated weighting factors.

In step S404, the encoding device 1 calculates a weighting factor defined in advance according to the direction of the intra prediction mode.

In step S405, the encoding device 1 generates a predicted image using the calculated weighting factors.

According to the encoding device 1 according to the present embodiment, the weighting factor can be determined by considering not only the intra prediction mode and the CU/TU size but also the position of the reference pixel. A decrease in prediction accuracy can be prevented by appropriately synthesizing the non-smoothed predicted image and the smoothed predicted image.

For example, as shown in FIG. 4, when a predicted image is generated using intra prediction mode 2, the prediction accuracy of the non-smoothed predicted image is considered to be high for the predicted pixels located on the left side and the lower side near the reference pixel. On the other hand, it is considered that the prediction accuracy of the smoothed prediction image is high for the prediction pixel positioned on the upper right far from the reference pixel.

Therefore, according to the encoding apparatus 1 according to the present embodiment, when the positions of the reference pixels include the left side and the right side, the weights shown in FIG. 4 are used instead of the distribution of the weight coefficients shown in FIG. Prediction accuracy can be improved by using the distribution of coefficients.

Further, the decoding device 3 according to the present embodiment is configured to divide an original image of each frame constituting a moving image into CUs and decode the CUs. Also, the decoding device 3 according to this embodiment is configured to be able to divide a CU into a plurality of TUs, like the encoding device 1 according to this embodiment.

As shown in FIG. 10 , the decoding device 3 according to this embodiment includes an entropy decoding unit 31 , a decoding order control unit 32 , a sequential decoded image generation unit 33 and a memory 34 .

The entropy decoding unit 31 is configured to decode transform coefficients, flag information, and the like from the stream output from the encoding device 1 by performing entropy decoding processing on the stream output from the encoding device 1. ing. Here, the transform coefficients are quantized transform coefficients obtained by the encoding apparatus 1 as signals encoded by dividing the original image in units of frames into CUs.

The decoding order control unit 32 is configured to determine the decoding order of TUs within the CU based on the intra prediction mode.

Specifically, the decoding order control unit 32 uses a flag indicating whether or not the TU division output by the entropy decoding unit 31 has been performed (whether or not the CU is divided into a plurality of TUs) and the intra prediction mode. The direction is configured to determine the decoding order of the TUs within the CU.

For example, similar to the encoding order control unit 13, the decoding order control unit 32 may be configured such that when a CU is divided into a plurality of TUs and the direction of the intra prediction mode is from the upper right to the lower left (that is, the lower left direction prediction is performed toward the upper right), the decoding order of the lower left TU in the CU → the lower right TU in the CU → the upper left TU in the CU → the upper right TU in the CU, or The decoding process is performed in a predetermined decoding order among the decoding order of the lower left TU in the CU→the upper left TU in the CU→the lower right TU in the CU→the upper right TU in the CU. may

The sequentially decoded image generator 33 is configured to generate a decoded image for each TU based on the decoding order determined by the decoding order controller 32 and the method of dividing the CU into TUs.

Specifically, when the CU is divided into a plurality of TUs, the sequentially decoded image generation unit 33
By sequentially performing intra prediction, inverse quantization processing, and inverse orthogonal transformation processing on the quantized transform coefficients output by the entropy decoding unit 31 according to the decoding order determined by the decoding order control unit 32, configured to generate a decoded image;

As shown in FIG. 10, the sequentially decoded image generation unit 33 includes an intra prediction unit 33a, an inverse quantization/inverse transform unit 33b, and a decoded image generation unit 33c.

The intra prediction unit 33 a may be configured to generate a prediction image using the intra prediction mode output by the entropy decoding unit 31 according to the decoding order determined by the decoding order control unit 32 .

As shown in FIG. 11, the intra prediction unit 33a includes a non-smooth intra prediction unit 33a1, a smooth intra prediction unit 33a2, a weighting factor determination unit 33a3, and a predicted image generation unit 33a4.

The non-smoothed intra prediction unit 33a1 is configured to generate a non-smoothed predicted image using the intra prediction mode without smoothing the reference pixels, similarly to the non-smoothed intra prediction unit 14a1.

The smoothed intra prediction unit 33a2, like the smoothed intra prediction unit 14a2, is configured to perform smoothing processing on reference pixels and then generate a smoothed predicted image using the intra prediction mode.

The weighting factor determination unit 33a3 is configured to determine a weighting factor according to the intra prediction mode and the position of the reference pixel, similarly to the weighting factor determination unit 14a3.

Like the predicted image generator 14a4, the predicted image generator 33a4 uses the weighting factors determined by the weighting factor determining unit 14a3 to combine the non-smoothed predicted image and the smoothed predicted image to generate a predicted image. It is configured.

The inverse quantization/inverse transform unit 33b performs inverse quantization processing and inverse transform processing (for example, inverse orthogonal transform processing) on the quantized transform coefficients output from the entropy decoding unit 31, thereby generating a residual configured to generate a signal;

The decoded image generator 33c is configured to generate a decoded image by adding the predicted image generated by the intra prediction unit 33a and the residual signal generated by the inverse quantization/inverse transform unit 33b.

The memory 34 is configured to hold the decoded images generated by the sequential decoded image generation unit 33 so that they can be used as reference images for intra prediction and inter prediction.

Note that the operation of generating a predicted image in the decoding device 3 according to this embodiment is the same as the operation of generating a predicted image in the encoding device 1 according to this embodiment shown in FIGS. omitted.

According to the decoding device 3 according to the present embodiment, it is possible to determine the weighting factor in consideration of not only the intra prediction mode and the size of the CU/TU but also the position of the reference pixel. A decrease in prediction accuracy can be prevented by appropriately synthesizing the smoothed predicted image and the smoothed predicted image.

(Other embodiments)
As noted above, the present invention has been described through the above-described embodiments, but the statements and drawings forming part of the disclosure in such embodiments should not be understood to limit the present invention. From such disclosure, various alternative embodiments, implementations and operational techniques will become apparent to those skilled in the art.

Moreover, although not particularly mentioned in the above-described embodiment, a program may be provided that causes a computer to execute each process performed by the above-described encoding device 1 and decoding device 3 . Also, such a program may be recorded on a computer-readable medium. Such programs can be installed on a computer using a computer readable medium. 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 recording medium such as a CD-ROM or a DVD-ROM.

Alternatively, a chip configured by a memory storing a program for realizing at least part of the functions in the encoding device 1 and the decoding device 3 described above and a processor executing the program stored in the memory may be provided. good.

1 Encoding device 11 Intra prediction mode determination unit 12 TU division determination unit 13 Encoding order control unit 14 Sequential local decoded image generation unit 14a Intra prediction unit 14a1 Non-smooth intra prediction unit 14a2 Smooth intra prediction Unit 14a3 Weighting factor determination unit 14a4 Predicted image generation unit 14b Residual signal generation unit 14c Orthogonal transformation/quantization unit 14d Inverse quantization/inverse orthogonal transformation unit 14e Local decoded image generation unit 15 Memory 16 Entropy encoding unit 3 Decoding device 31 Entropy decoding unit 32 Decoding order control unit 33 Sequentially decoded image generation unit 33a Intra prediction unit 33a1 Non-smooth intra prediction unit 33a2 Smooth intra prediction unit 33a3 Weighting factor determination Part 33a4...Predicted image generation part 33b...Inverse quantization/inverse transformation part 33c...Decoded image generation part 34...Memory

Claims (2)

A prediction device that predicts a target block of an original image by intra prediction,
A first generation unit that generates a predicted image by intra prediction from reference pixels that have undergone smoothing processing using an intra prediction mode;
a second generating unit that generates an image for synthesis from reference pixels without using the smoothing process based on the intra prediction mode;
a weighted synthesis unit that performs weighted synthesis processing of the predicted image and the synthesis image ,
The weighted combining unit
When the intra prediction mode is smaller than the horizontal mode, and when the position of the reference pixel is only a position included in a pixel line adjacent to the left side of the target block is satisfied. , a process for determining a predetermined weighting factor is performed, and if the condition is not satisfied, a process different from the process is performed,
When the intra prediction mode is larger than the vertical mode, and when the position of the reference pixel is only the position included in the pixel line adjacent to the upper side of the target block is satisfied. performs a process of determining a predetermined weighting factor, and performs a process different from the process if the condition is not satisfied. A program for causing a computer to function as the prediction device according to claim 1.

Images