JP7460384B2 - Prediction device, encoding device, decoding device, and program - Google Patents

Description

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

In a video encoding system, the encoding device divides the original image into blocks and performs inter prediction using temporal correlation between frames and intra prediction using spatial correlation within frames. The block to be encoded is predicted while switching, and the prediction residual representing the difference between the predicted block and the block to be encoded is subjected to transformation processing, quantization processing, and entropy encoding processing, and a code that is a bit stream is generated. Output converted data.

Versatile Video Coding (VVC), a next-generation coding method, employs triangular partitioning prediction (TPM: Triangle Partitioning Mode) as one of the inter-prediction modes. TPM divides the coding target block diagonally, assigning motion vectors to each area, and performs motion compensation prediction.

In TPM, for predicted images near region boundaries, a mixing process (called blending) is performed in which predicted images for each region are mixed using weighted synthesis (weighted average) according to their position, thereby suppressing discontinuity caused by motion prediction compensation for each region. In sequences that contain many flat regions and edges, such as screen content, blending can cause edge blurring. For this reason, Non-Patent Document 1 introduces a non-blending mode in TPM prediction that synthesizes predicted images without blending.

JVET-O1172, “Non-CE4/8: On disabling blending process in TPM”

When introducing multiple blending modes (multiple combining methods) in TPM as in Non-Patent Document 1, it is necessary to signal a flag indicating which mode is to be used from the encoding side to the decoding side. Therefore, there is a problem in that the amount of flags to be signaled increases, leading to a decrease in encoding efficiency.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a prediction device, an encoding device, a decoding device, and a program that can suppress an increase in the amount of flags to be signaled even when a plurality of combining methods are introduced into segmented prediction. do.

A prediction device according to a first aspect is a prediction device that performs prediction in units of blocks obtained by dividing an image, and includes a dividing unit that divides a prediction target block along a straight line and outputs at least two prediction regions; For each prediction region, a generation unit that generates a region prediction image using a motion vector, a calculation unit that calculates similarity information indicating a degree of similarity between two region prediction images generated by the generation unit, and the calculation unit. a determining unit that determines a combining method to be used for combining the two region predicted images from among a plurality of methods based on the similarity information calculated by the determining unit; The gist of the present invention is to include a combining unit that combines two region predicted images and outputs a predicted block corresponding to the prediction target block.

The encoding device according to the second aspect is characterized in that it includes the prediction device according to the first aspect.

The decoding device according to the third aspect is characterized in that it includes the prediction device according to the first aspect.

The gist of the program according to the fourth aspect is to cause a computer to function as the prediction device according to the first aspect.

According to the present invention, it is possible to provide a prediction device, an encoding device, a decoding device, and a program that can suppress an increase in the amount of flags to be signaled even when a plurality of combining methods are introduced into divided prediction.

FIG. 1 is a diagram illustrating a configuration of an encoding device according to an embodiment. FIG. 2 is a diagram illustrating a configuration of a triangulation prediction unit of an encoding device according to an embodiment. 11A and 11B are diagrams illustrating an operation of a division unit according to the embodiment. 6A and 6B are diagrams illustrating operations of a calculation unit and a determination unit according to the embodiment. 6A and 6B are diagrams illustrating an operation of a synthesis unit according to the embodiment. FIG. 1 is a diagram showing the configuration of a decoding device according to an embodiment. FIG. 2 is a diagram showing a configuration of a triangulation prediction unit of a decoding device according to an embodiment. FIG. 11 is a diagram showing an operation flow of a triangulation prediction 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 the embodiment encode and decode moving images represented by MPEG, respectively. In the description of the drawings below, the same or similar parts are designated by the same or similar symbols.

<Configuration of encoding device>
First, the configuration of the encoding device according to this embodiment will be explained. FIG. 1 is a diagram showing 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.

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

The block division unit 100 divides an input image of a frame (or picture) constituting a moving image into a plurality of image blocks, and outputs the image blocks obtained by division to the subtraction unit 110. The size of the image block is, for example, 32 x 32 pixels, 16 x 16 pixels, 8 x 8 pixels, or 4 x 4 pixels. The shape of the image block is not limited to a square, but may be a rectangle (non-square). The image block is the unit (block to be coded) for coding by the coding device 1, and is the unit (block to be coded) for decoding by the decoding device. Such an image block is sometimes called a CU (Coding Unit).

The block division unit 100 performs block division on the luminance signal and the color difference signal. In the following, the case where the shape of the block division is the same for the luminance signal and the color difference signal will be mainly described, but it may be possible to control the division independently for the luminance signal and the color difference signal. When there is no particular distinction between the luminance block and the color difference block, they are simply referred to as the block to be coded.

The subtraction unit 110 calculates a prediction residual that represents the difference (error) between the block to be coded output by the block division unit 100 and a predicted block obtained by predicting the block to be coded by the prediction unit 170. The subtraction unit 110 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 120.

The conversion/quantization unit 120 performs conversion processing and quantization processing on a block-by-block basis. The conversion/quantization unit 120 includes a conversion unit 121 and a quantization unit 122.

The transform unit 121 performs a transform process on the prediction residual output by the subtraction unit 110 to calculate a transform coefficient for each frequency component, and outputs the calculated transform coefficient to the quantization unit 122. The transform process (transformation) refers to a process of converting a pixel domain signal into a frequency domain signal, such as discrete cosine transform (DCT), discrete sine transform (DST), Karhunen-Loeb transform (KLT), and transforms that convert these to integers.

The quantization unit 122 quantizes the transform coefficients output by the transform unit 121 using a quantization parameter (Qp) and a quantization matrix, and outputs the quantized transform coefficients to the entropy coding unit 130 and the inverse quantization and inverse transform unit 140. Note that the quantization parameter (Qp) is a parameter that is commonly applied to each transform coefficient in a block and determines the coarseness of quantization. The quantization matrix is a matrix whose elements are the quantization values used when quantizing each transform coefficient.

The entropy encoding unit 130 performs entropy encoding on the transform coefficients output by the quantization unit 122, performs data compression to generate an encoded stream (bit stream), and transmits the encoded stream to the encoding device 1. Output to outside. For entropy encoding, a Huffman code, CABAC (Context-based Adaptive Binary Arithmetic Coding), or the like can be used. Note that the entropy encoding unit 130 acquires information such as the size and shape of each block to be encoded from the block division unit 100, and acquires information regarding prediction (for example, information on prediction mode and motion vector) from the prediction unit 170. It also encodes this information.

The inverse quantization and inverse transform unit 140 performs inverse quantization and inverse transform processing on a block-by-block basis. The inverse quantization and inverse transform unit 140 includes an inverse quantization unit 141 and an inverse transform unit 142.

The dequantization unit 141 performs dequantization processing corresponding to the quantization processing performed by the quantization unit 122. Specifically, the dequantization unit 141 restores the transform coefficients by dequantizing the transform coefficients output by the quantization unit 122 using a quantization parameter (Qp) and a quantization matrix. The transform coefficients are output to the inverse transform section 142.

The inverse transform unit 142 performs an inverse transform process corresponding to the transform process performed by the transform unit 121. For example, when the transformer 121 performs DCT, the inverse transformer 142 performs inverse DCT. The inverse transform unit 142 performs inverse transform processing on the transform coefficients output by the inverse quantization unit 141 to restore the prediction residual, and outputs the restored prediction residual, which is the restored prediction residual, to the synthesis unit 150. .

The synthesis unit 150 synthesizes the reconstructed prediction residual output by the inverse transform unit 142 with the predicted block output by the prediction unit 170 on a pixel-by-pixel basis. The synthesis unit 150 adds each pixel value of the reconstructed prediction residual to each pixel value of the predicted block to reconstruct (decode) the block to be coded, and outputs the reconstructed decoded image (reconstructed block) on a block-by-block basis to the memory 160.

The memory 160 accumulates the reconstructed blocks output by the synthesis unit 150 as decoded images on a frame-by-frame basis. The memory 160 outputs the stored decoded images to the prediction unit 170. Note that a loop filter may be interposed between the synthesis unit 150 and the memory 160.

The prediction unit 170 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 110 and the synthesis unit 150. The block to be coded on which prediction processing is performed is called the block to be predicted.

The prediction unit 170 has an intra prediction unit 171, an inter prediction unit 172, and a switching unit 173. In this embodiment, the inter prediction unit 172 corresponds to a prediction device that performs prediction processing on a block-by-block basis.

The intra prediction unit 171 selects an optimal intra prediction mode to be applied to the prediction target block from among multiple intra prediction modes, and predicts the prediction target block using the selected intra prediction mode. The intra prediction unit 171 generates an intra prediction block by referring to decoded pixel values adjacent to the prediction target block in the decoded image stored in the memory 160, and outputs the generated intra prediction block to the switching unit 173. The intra prediction unit 171 also outputs information related to the selected intra prediction mode to the entropy coding unit 130.

The inter prediction unit 172 uses the decoded image stored in the memory 160 as a reference image, calculates a motion vector by a method such as block matching, predicts the prediction target block to generate an inter prediction block, and uses the generated inter prediction block. The predicted block is output to the switching unit 173. The inter prediction unit 172 selects the optimal inter prediction method from among inter prediction using multiple reference images (typically, bi-prediction) and inter prediction using one reference image (unidirectional prediction), Perform inter prediction using the selected inter prediction method. The inter prediction unit 172 outputs information related to inter prediction (motion vector information, etc.) to the entropy encoding unit 130.

In this embodiment, the inter prediction unit 172 includes a triangulation prediction unit 172a that performs triangulation prediction (TPM) (see FIG. 2). The triangulation prediction unit 172a assigns a motion vector to each region obtained by dividing the current block to be encoded diagonally, and performs motion compensation prediction. When applying triangulation prediction to the prediction target block, the inter prediction unit 172 outputs a triangulation prediction application flag indicating that triangulation prediction is applied to the prediction target block to the entropy encoding unit 130, and performs entropy encoding. The triangulation prediction application flag is transmitted from the unit 130 to the decoding side. Details of the triangulation prediction unit 172a will be described later.

The switching unit 173 switches between the inter prediction block output by the inter prediction unit 172 and the intra prediction block output by the intra prediction unit 171, and outputs one of the prediction blocks to the subtraction unit 110 and the synthesis unit 150.

Next, the triangulation prediction unit 172a according to this embodiment will be explained. FIG. 2 is a diagram showing the configuration of the triangulation prediction unit 172a according to this embodiment.

As shown in FIG. 2, the triangulation prediction unit 172a includes a division unit 1721, a generation unit 1722, a calculation unit 1723, a determination unit 1724, and a synthesis unit 1725.

The dividing unit 1721 divides the prediction target block diagonally, and outputs two triangular areas to the generating unit 1722. The triangular region is an example of a prediction region obtained by dividing the prediction target block by straight lines. There are two division methods: the method shown in FIG. 3(a) and the method shown in FIG. 3(b). In the method shown in FIG. 3A, the dividing unit 1721 divides the block to be encoded along a dividing line passing through the upper left apex position and the lower right apex position. In the method shown in FIG. 3B, the dividing unit 1721 divides the block to be encoded along a dividing line passing through the upper right zenith position and the lower left zenith position. The dividing unit 1721 outputs a dividing direction flag indicating whether the direction of division is the diagonal direction shown in FIG. 3(a) or the diagonal direction shown in FIG. 3(b) to the entropy encoding unit 130, and generates an entropy code. The division direction flag is transmitted from the encoding unit 130 to the decoding side.

The generation unit 1722 generates a region predicted image using the motion vector for each of the two triangular regions, and outputs the two region prediction images to the calculation unit 1723 and the synthesis unit 1725. Specifically, the generation unit 1722 assigns a motion vector to each region into which the prediction target block is divided, and generates a predicted image for each region using the assigned motion vector. Here, the generation unit 1722 arranges a plurality of motion vector reference sources in order of priority, and selects one motion vector for each triangular region from among the fixed number of high-ranking motion vectors. The generation unit 1722 outputs motion vector information indicating the selected motion vector to the entropy encoding unit 130, and transmits the motion vector information from the entropy encoding unit 130 to the decoding side.

The calculation unit 1723 calculates similarity information indicating the similarity between the two region predicted images generated by the generation unit 1722, and outputs the calculated similarity information to the determination unit 1724. For example, as shown in FIG. 4A, the calculation unit 1723 calculates the similarity between the boundary region R1 between the region predicted image 1 and the region predicted image 2 and the boundary region R2 between the region predicted image 1 and the region predicted image 2. Similarity information indicating the degree of similarity is calculated. The similarity information may be any information as long as it indicates the degree of similarity; for example, the sum of absolute values of differences (SAD) can be used as the similarity information.

The calculation unit 1723 may calculate the SAD between the pixel values (plurality) in the boundary region R1 and the pixel values (plurality) in the corresponding boundary region R2 as similarity information. A smaller SAD indicates a higher degree of similarity, and a larger SAD indicates a lower degree of similarity.

Calculating the similarity information may be calculating the similarity between two reference images corresponding to the two region predicted images. The degree of similarity between reference images may be calculated as the SAD between pixel values in one reference image and corresponding pixel values in the other reference image.

Based on the similarity information calculated by the calculation unit 1723, the determination unit 1724 determines a combination method to be used for combining the two region predicted images from among a plurality of methods, and sends information indicating the determined combination method to the combination unit 1725. Output to. For example, the plurality of methods includes a first method (hereinafter referred to as a "Blending method") that blends a boundary region between two regionally predicted images by weighted synthesis using a weighting coefficient for each pixel position. The determining unit 1724 determines, based on the similarity information calculated by the calculating unit 1723, whether to use the blending method as the combining method used to combine the two region predicted images. According to the blending method, discontinuity caused by motion prediction compensation for each region can be suppressed.

In this embodiment, the plurality of methods further includes a second method (hereinafter referred to as a "Non-blending method") in which boundary regions are not mixed by weighted synthesis. Based on the similarity information calculated by the calculating unit 1723, the determining unit 1724 determines which of the blending method and the non-blending method to use as the combining method for combining the two region predicted images. According to the non-blending method, blurring of edges caused by the blending method can be avoided for sequences that include many flat areas and edges.

When switching between the blending method and the non-blending method is enabled, it may be necessary to transmit a flag indicating which of the blending method and the non-blending method is applied to the decoding side. However, in the background area of an image (for example, a smooth area such as the sky), the characteristics of the predicted image generated will not change whether the blending method or the non-blending method is used, and it is difficult to select which method. However, the encoding efficiency is not affected.

In this embodiment, the determining unit 1724 determines either the blending method or the non-blending method based on the similarity information calculated by the calculating unit 1723. Such similarity information can be calculated on the decoding side as well, so it can be calculated on both the encoding and decoding sides without transmitting a flag indicating whether to apply the blending method or the non-blending method to the decoding side. The common method can be determined implicitly. This eliminates the need for such flag transmission (signaling) to the decoding side, and can suppress an increase in the amount of flags to be signaled.

For example, as shown in Figures 4(b) and (c), based on the similarity information calculated by the calculation unit 1723, the determination unit 1724 determines the blending method when the similarity indicated by the similarity information is lower than a threshold, and determines the non-blending method when the similarity indicated by the similarity information is higher than the threshold. This threshold may be a predetermined fixed value, or may be a variable value transmitted from the encoding side to the decoding side.

The combining unit 1725 combines the two region predicted images output by the generating unit 1722 using the combining method determined by the determining unit 1724, and outputs a predicted block corresponding to the prediction target block. In the present embodiment, the combining unit 1725 combines the two area predicted images using the blending method when the determining unit 1724 determines the blending method, and combines the two area predicted images using the blending method when the determining unit 1724 determines the non-blending method. Images are combined using a non-blending method.

FIG. 5 is a diagram showing the operation of combining two region predicted images P ₁ and P ₂ . As shown in FIG. 5A, when combining two region predicted images P ₁ and P ₂ using the blending method, the combining unit 1725 creates a weight coefficient map (Weight) according to the block size and block shape of the prediction target block. map), the two region predicted images P ₁ and P ₂ are combined by weighted averaging. Thereby, the boundary regions R1 and R2 shown in FIG. 4(a) are adjusted to be continuous. On the other hand, as shown in FIG. 5B, when combining the two region predicted images P ₁ and P ₂ using the Non-blending method, the composition unit 1725 performs a weighted average of the two region prediction images P ₁ and P ₂ . Synthesize without

<Configuration of Decoding Device>
Next, the configuration of a decoding device according to this embodiment will be described, focusing mainly on the differences from the configuration of the encoding device described above. Fig. 6 is a diagram showing 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. 6, the decoding device 2 has an entropy decoding unit 200, an inverse quantization/inverse transform unit 210, a synthesis unit 220, a memory 230, and a prediction unit 240.

The entropy decoding unit 200 decodes the coding stream generated by the coding device 1, obtains quantized transform coefficients, and outputs the obtained transform coefficients to the inverse quantization/inverse transform unit 210 (inverse quantization unit 211). The entropy decoding unit 200 also obtains various types of signaling information. For example, the entropy decoding unit 200 obtains information related to the prediction process to be applied to the block to be decoded, and outputs the obtained information to the prediction unit 240.

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

The inverse quantization unit 211 performs an inverse quantization process corresponding to the quantization process performed by the quantization unit 122 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 by the entropy decoding unit 200 using a quantization parameter (Qp) 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 121 of the encoding device 1. The inverse transform unit 212 performs inverse transform processing on the transform coefficients output by 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 reconstructs (decodes) the block to be decoded by synthesizing the prediction residual output by the inverse transform unit 212 and the prediction block output by the prediction unit 240 on a pixel-by-pixel basis, and outputs the reconstructed block to the memory 230.

The memory 230 stores the restored blocks output by the combining unit 220 in units of frames as decoded images. The memory 230 outputs the decoded image in units of frames to the outside of the decoding device 2 . Note that a loop filter may be interposed between the synthesis section 220 and the memory 230.

The prediction unit 240 performs prediction on a block-by-block basis. The prediction unit 240 has an intra prediction unit 241, an inter prediction unit 242, and a switching unit 243.

The intra prediction unit 241 refers to reference pixels adjacent to the decoding target block in the decoded image stored in the memory 230, and predicts the decoding target block by intra prediction based on the information output from the entropy decoding unit 200. Then, the intra prediction unit 241 generates an intra prediction block and outputs the generated intra prediction block to the switching unit 243.

The inter prediction unit 242 predicts the block to be decoded by inter prediction using the decoded image stored in the memory 230 as a reference image. The inter prediction unit 242 generates an inter prediction block by performing inter prediction using the motion vector information output by the entropy decoding unit 200, and outputs the generated inter prediction block to the switching unit 243.

In this embodiment, the inter prediction unit 242 includes a triangulation prediction unit 242a that performs triangulation prediction (TPM) (see FIG. 7). If the entropy decoding unit 200 acquires a triangulation prediction application flag indicating that triangulation prediction is to be applied to the prediction target block, the inter prediction unit 242 applies triangulation prediction to the prediction target block. Details of the triangulation prediction unit 242a will be described later.

The switching unit 243 switches between the intra prediction block output by the intra prediction unit 241 and the inter prediction block output by the inter prediction unit 242, and outputs one of the prediction blocks to the combining unit 220.

Next, the triangular decomposition prediction unit 242a according to this embodiment will be described. FIG. 7 is a diagram showing the configuration of the triangular decomposition prediction unit 242a according to this embodiment.

As shown in FIG. 7, the triangulation prediction unit 242a includes a division unit 2421, a generation unit 2422, a calculation unit 2423, a determination unit 2424, and a synthesis unit 2425.

The dividing unit 2421 divides the prediction target block diagonally based on the dividing direction flag acquired by the entropy decoding unit 200, and outputs two triangular areas to the generating unit 2422.

The generation unit 2422 generates a region predicted image for each of the two triangular regions based on the motion vector information acquired by the entropy decoding unit 200, and outputs the two region prediction images to the calculation unit 2423 and the composition unit 2425.

The calculation unit 2423 calculates similarity information indicating the similarity between the two region predicted images generated by the generation unit 2422, and outputs the calculated similarity information to the determination unit 2424. The method of calculating the similarity information is the same as that on the encoding side, and the similarity information is calculated using a common rule (algorithm) on the encoding side and the decoding side.

The determination unit 2424 determines the synthesis method to be used for synthesizing the two area prediction images from among a plurality of methods based on the similarity information calculated by the calculation unit 2423, and outputs information indicating the determined synthesis method to the synthesis unit 2425. In this embodiment, the determination unit 2424 determines whether to use the blending method or the non-blending method as the synthesis method to be used for synthesizing the two area prediction images based on the similarity information calculated by the calculation unit 2423. The method of determining the synthesis method is the same as that on the encoding side, and the synthesis method is determined using a rule common to the encoding side and the decoding side. In other words, the determination unit 2424 determines the synthesis method based on the similarity information, without being based on a flag indicating whether the blending method or the non-blending method is to be applied.

The combining unit 2425 combines the two region predicted images output by the generating unit 2422 using the combining method determined by the determining unit 2424, and outputs a predicted block corresponding to the prediction target block. In this embodiment, the combining unit 2425 combines two area predicted images using a blending method when the determining unit 2424 determines a blending method, and combines two area predicted images using a blending method when the determining unit 2424 determines a non-blending method. Images are combined using a non-blending method.

<Operation of Triangulation Prediction Unit>
Next, the operation of the triangular decomposition prediction units 172a and 242a according to this embodiment will be described. Since the triangular decomposition prediction units 172a and 242a perform similar operations, the triangular decomposition prediction unit 242a will be taken as an example for description here. Fig. 8 is a diagram showing the operation flow of the triangular decomposition prediction unit 242a according to this embodiment.

As shown in FIG. 8, in step S1, the division unit 2421 divides the prediction target block diagonally based on the division direction flag acquired by the entropy decoding unit 200, and outputs two triangular regions to the generation unit 2422.

In step S2, the generation unit 2422 generates a region prediction image for each of the two triangular regions based on the motion vector information acquired by the entropy decoding unit 200, and outputs the two region prediction images to the calculation unit 2423 and the synthesis unit 2425.

In step S3, the calculation unit 2423 calculates similarity information indicating the similarity between the two region predicted images generated by the generation unit 2422, and outputs the calculated similarity information to the determination unit 2424.

In step S4, the determination unit 2424 determines whether the blending method or the non-blending method is to be used as the synthesis method for synthesizing the two area prediction images based on the similarity information calculated by the calculation unit 2423.

In step S5, the synthesis unit 2425 synthesizes the two area prediction images output by the generation unit 2422 using the synthesis method determined by the determination unit 2424, and outputs a prediction block corresponding to the block to be predicted.

In this way, according to this embodiment, a common method can be implicitly determined on the encoding side and the decoding side without transmitting a flag indicating whether the blending method or the non-blending method is to be applied to the decoding side. This makes it unnecessary to transmit (signal) a flag indicating the blending method to the decoding side even when multiple blending methods are introduced into the triangulation prediction, and can suppress an increase in the amount of flags to be signaled.

<Other embodiments>
In the above-mentioned embodiment, an example is described in which the blending method or the non-blending method is determined as the synthesis method used for synthesizing two area prediction images based on similarity information. However, when introducing a plurality of blending methods, it is also possible to determine which of the plurality of blending methods is used based on similarity information.

The plurality of blending methods may have different parameters for blending the boundary areas. Here, the parameter may be at least one of a boundary region range and a set of weighting factors. For example, the range of the boundary area to which the blending process is applied is the first range (narrow range), the first blending method uses the first set of weighting coefficients, and the range of the boundary area to which the blending process is applied is the second range (wide range). ) and a second blending scheme using a second set of weighting coefficients. Here, the number of weighting coefficients forming the second set is larger than that of the first set.

Under such a premise, the determining units 1724 and 2424 determine the second blending method when the similarity indicated by the similarity information calculated by the calculating units 1723 and 2423 is lower than the threshold, and If the degree is higher than the threshold, the first blending method is determined.

Alternatively, a first threshold and a second threshold greater than the first threshold may be introduced. The decision units 1724 and 2424 decide on the non-blending method if the similarity indicated by the similarity information calculated by the calculation units 1723 and 2423 is higher than the second threshold, decide on the second blending method if the similarity indicated by this similarity information is lower than the first threshold, and decide on the first blending method if the similarity indicated by this similarity information is within the range from the first threshold to the second threshold.

In the above-described embodiment, an example was described in which the prediction target block is divided along diagonal lines to output two triangular areas, but the divided shape is not limited to triangles, and may be any shape that is divided along straight lines. Further, the number of divisions of the prediction target block can also be three or more. In that case, the degree of similarity near the boundary for each pair of regions in three or more prediction regions is calculated, and the weighting coefficient used for compositing the pair of regions is determined based on that information. .

Note that a program that causes a computer to execute each process performed by the encoding device 1 may be provided. A program that causes a computer to execute each process performed by the decoding device 2 may be provided. The program may be recorded on a computer readable medium. Computer-readable media allow programs to be installed 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, and may be, for example, a recording medium such as a CD-ROM or a DVD-ROM.

The circuits that execute the processes 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 execute the processes 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).

Although the embodiments have been described above in detail with reference to the drawings, the specific configuration is not limited to that described above, and various design changes can be made without departing from the scope of the invention.

1: Encoding device 2: Decoding device 100: Block division unit 110: Subtraction unit 120: Transformation and quantization unit 121: Transformation unit 122: Quantization unit 130: Entropy encoding unit 140: Inverse quantization and inverse transform unit 141: Inverse quantization unit 142: Inverse transform unit 150: Synthesis unit 160: Memory 170: Prediction unit 171: Intra prediction unit 172: Inter prediction unit 172a: Triangular partition prediction unit 173: Switching unit 200: Entropy decoding unit 210: Inverse transformation unit 211: Inverse quantization unit 212: Inverse transformation unit 220: Synthesis unit 230: Memory 240: Prediction unit 241: Intra prediction unit 242: Inter prediction unit 242a: Triangular partition prediction unit 243: Switching unit 1721: Division unit 1722 : Generation unit 1723 : Calculation unit 1724 : Determination unit 1725 : Combination unit 2421 : Division unit 2422 : Generation unit 2423 : Calculation unit 2424 : Determination unit 2425 : Combination unit

Claims (7)

A prediction device that performs prediction in units of blocks obtained by dividing an image,
a dividing unit that divides the prediction target block along straight lines and outputs at least two prediction regions;
a generation unit that generates a region predicted image using a motion vector for each of the two prediction regions;
a calculation unit that calculates similarity information indicating the similarity between the two region predicted images generated by the generation unit;
a determining unit that determines a combination method to be used for combining the two region predicted images from among a plurality of methods, based on the similarity information calculated by the calculation unit;
A prediction device comprising: a synthesis section that synthesizes the two region predicted images using the synthesis method determined by the determination section and outputs a prediction block corresponding to the prediction target block. The plurality of methods include a first method of mixing boundary regions between the two region predicted images by weighted synthesis using weighting coefficients for each pixel position,
The prediction device according to claim 1, wherein the determining unit determines whether to use the first method as the combining method based on the similarity information calculated by the calculating unit. The plurality of methods further includes a second method of not blending the boundary region by the weighted blending,
The prediction device according to claim 2 , wherein the determination unit determines which of the first method and the second method to use as the combining method based on the similarity information calculated by the calculation unit. The first method includes a plurality of first methods having mutually different parameters for mixing the boundary regions,
The parameter is at least one of the range of the boundary region and the set of weighting coefficients,
When the first method is used, the determining section determines which of the plurality of first methods is to be used as the combining method based on the similarity information calculated by the calculating section. The prediction device according to claim 2 or 3. An encoding device comprising the prediction device according to any one of claims 1 to 4. A decoding device comprising the prediction device according to any one of claims 1 to 4. A program that causes a computer to function as a prediction device according to any one of claims 1 to 4.

Images