WO2012090425A1

WO2012090425A1 - Moving image encoding device, moving image encoding method, and moving image encoding program, as well as moving image decoding device, moving image decoding method, and moving image decoding program

Info

Publication number: WO2012090425A1
Application number: PCT/JP2011/007065
Authority: WO
Inventors: 上田　基晴; 英樹竹原; 博哉中村; 和美荒蔭; 智坂爪
Original assignee: 株式会社Ｊｖｃケンウッド
Priority date: 2010-12-27
Filing date: 2011-12-19
Publication date: 2012-07-05

Abstract

A motion vector detection unit of the present invention detects a motion vector from a first reference image with respect to a block to be encoded. A first reference image combination unit generates a first combined reference block, in which a first reference block which has been extracted from the first reference image using the motion vector and a predetermined area of another at least one reference image, have been combined. A second reference image combination unit generates a second combined reference block in which, by way of information required for generation of the combined reference block of the areas which have been coded, a second reference block and a predetermined area of the reference image used for the combination are identified and combined. An encoding unit encodes a predicted difference block in which a predicted block, which has been selected from a plurality of predicted blocks which include at least the first combined reference block and the second combined reference block, has been subtracted from the block to be encoded.

Description

Moving picture encoding apparatus, moving picture encoding method, moving picture encoding program, moving picture decoding apparatus, moving picture decoding method, and moving picture decoding program

The present invention relates to a video signal encoding and decoding technique.

In recent years, services that deliver digital image and sound content via broadcast waves such as satellite and terrestrial waves and networks have been put into practical use, and content with a huge amount of information can be efficiently recorded and transmitted. In order to do so, a high-efficiency encoding technique is required. High-efficiency coding of moving images includes the correlation between pixels that are spatially adjacent in the same frame of a moving image signal, and the correlation between temporally adjacent frames and fields, as represented by MPEG4-AVC. A method of compressing information using it is used.

In MPEG4-AVC, as a compression using temporal correlation, a local decoded image of an already encoded frame is used as a reference image for a target image that is a target frame to be encoded, and a two-dimensional block having a predetermined size is used. A motion amount (hereinafter referred to as “motion vector”) between the target image and the reference image is detected in units (hereinafter referred to as “target block”), and a predicted image based on the target block and the motion vector is detected. The generated motion compensated prediction is used.

In MPEG4-AVC, the size of a target block in a 16 × 16 pixel two-dimensional block (hereinafter referred to as “macroblock”), which is a unit of encoding processing, is made variable, and a motion vector for each target block is obtained. By using a method of predicting using, a method of storing a plurality of reference images and selecting a reference image used for prediction, and a method of generating a motion predicted image by obtaining a motion vector between two reference images and a target block, It is possible to improve the prediction accuracy of motion compensated prediction, thereby reducing the amount of information.

Also, in motion compensated prediction, it is necessary to encode and transmit the generated motion vector, and in order to prevent an increase in the amount of information due to the motion vector, a predicted motion vector predicted from the motion vector for a decoded block around the target block By encoding using values, it is possible to use motion compensated prediction called direct mode in which no motion vector is transmitted.

However, since the prediction of the motion vector cannot always be obtained with high accuracy, as shown in Patent Document 1, both the encoding side and the decoding side detect a motion vector between reference images, and the motion vector is A method of generating a predicted motion vector of a target block and assuming a direct mode on the assumption that it is continuous in time is also presented.

JP 2008-154015 A

In the motion compensated prediction in the conventional moving picture coding represented by MPEG4-AVC, the following problems cannot be solved, so that the improvement of the coding efficiency is hindered.

The first problem is the degradation of the quality of the motion compensated predicted image due to the degradation of the quality of the decoded image used as the reference image, especially the degradation mixed in the motion compensated predicted image when high-compression encoding is performed. While the component deteriorates the prediction accuracy, it is necessary to encode information for restoring the deteriorated component as a prediction difference, and the amount of information is increasing.

The second problem is that the motion vector prediction is not accurate enough for image signals with little temporal and spatial motion continuity, and the predicted image quality when using the direct mode is effective. It is a point not to do. This degradation occurs when adjacent blocks have different motions across the target object, and the motion vector used for prediction when the motion is large in time has moved corresponding to the motion of the original target block This degradation occurs because a block of positions is assumed. Similarly, when the motion changes with time, the prediction is not successful and deterioration occurs.

The third problem is an increase in the amount of code required for motion vector transmission when using prediction using two reference images or motion compensated prediction in units of fine blocks. When two reference images are used, prediction deterioration is smoothed by adding the reference images, and the influence of the deterioration component can be reduced. To increase.
Also, in motion compensation in fine block units, it is possible to obtain appropriate motion according to the boundary of the object, and the accuracy of the predicted image is improved, but it is necessary to transmit motion vectors in fine units. The amount of code increases.

Patent Document 1 is a technique presented to solve the second problem described above. When a spatially uniform motion is present, a motion vector obtained between reference images is a target block. The motion vector prediction accuracy is improved because the motion passes through the position, but if the motion is not spatially uniform, it is the predicted motion vector obtained without using the target block information Therefore, the motion is different from that of the target block, and the prediction is not sufficient. In addition, in order to capture a large motion, motion vector detection processing over a wide range between reference images is required for both the encoding device and the decoding device, which causes a problem that the amount of calculation increases.

Therefore, the present invention provides a technique for improving the efficiency of motion compensated prediction by improving the quality of a predicted image while suppressing an increase in the amount of calculation in the encoding device and decoding device without increasing the motion vector to be transmitted. The purpose is to do.

In order to solve the above-described problem, a moving image encoding device according to an aspect of the present invention includes a motion vector detection unit (117) that detects a motion vector from a first reference image for an encoding target block; A first reference image synthesizing unit that generates a first synthesized reference block by synthesizing a first reference block extracted from the first reference image using a motion vector and a predetermined region of at least one other reference image; 121), a reference image synthesis parameter storage unit (122) for storing information necessary for generating a synthesized reference block calculated by the first reference image synthesis unit, and an encoding stored in the reference image synthesis parameter storage unit Based on the information necessary for generating the synthesized reference block of the completed region, the second synthesized reference block synthesized by specifying the predetermined region of the reference image used for synthesis with the second reference block is generated. A second reference image synthesizing unit (123), selecting a prediction block for the encoding target block from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block, and a prediction block The encoding part which encodes the prediction difference block which subtracted from the encoding object block is provided.

By using the first reference image synthesis motion compensation prediction, a highly efficient prediction block using inter-reference image motion vector detection can be generated without performing additional motion vector transmission, and the second reference image synthesis motion compensation prediction can be performed. By using, it is possible to extend a highly efficient prediction block with a small amount of information, and to significantly reduce a region for performing inter-reference image motion vector detection necessary for performing reference image synthesis motion compensation prediction in a decoding device. And has the effect of greatly reducing the amount of computation.

Another aspect of the present invention is also a moving picture coding apparatus. The apparatus includes a motion vector detection unit (117) that detects a motion vector from a first reference image with respect to an encoding target block, and a first extracted from the first reference image using the motion vector. The first reference image synthesizing unit (121) that generates a first synthesized reference block obtained by synthesizing the reference block and a predetermined region of at least one other reference image, and the synthesized reference block calculated by the first reference image synthesizing unit. A reference image synthesis parameter storage unit (122) that stores information necessary for generation, and a second reference based on information necessary for generating a synthesized reference block of an encoded region stored in the reference image synthesis parameter storage unit. A second reference image synthesizing unit (123) for generating a second synthesized reference block obtained by identifying and synthesizing a predetermined region of the reference image used for synthesis with the block; A reference image synthesis selection unit (1224) for inputting information necessary for generating a synthesized reference block of the encoded region stored in the meter storage unit and determining a synthesis method for the second reference image synthesis unit; A reference image synthesis selection unit needs a correlation value between a plurality of reference blocks generated using information necessary for generating the synthesized reference block for the encoding target block and the synthesized reference block for the encoded region. A function of comparing correlation values between a plurality of reference blocks generated using information, and selecting an output from the second reference image combining unit between the second combined reference block and the second reference block (S1414 to S1417) , S1418-S1421), and a plurality of prediction blocks including at least a first synthesized reference block and a second synthesized reference block or a second reference block. More, select a prediction block for said current block comprises a coding unit configured to code a prediction difference block a prediction block is subtracted from the encoding target block.

An error value between the first predicted image and the second predicted image in the adjacent block referred to, and an error value between the first predicted image and the second predicted image with respect to the encoding target block are calculated, and the continuity between the reference images is calculated. When the continuity is lost, the first prediction image and the motion compensated prediction image are output as the motion compensated prediction image, so that the reference image synthesis motion compensation can be appropriately performed only when the continuity is maintained. Prediction can be made to function, and a more accurate motion compensated prediction image can be generated and encoding efficiency can be improved without giving additional information.

Another aspect of the present invention is also a moving picture coding apparatus. The apparatus includes a motion vector detection unit (117) that detects a motion vector from a first reference image with respect to an encoding target block, and a first extracted from the first reference image using the motion vector. A first reference image synthesizing unit (121) that generates a first synthesized reference block by synthesizing a predetermined area of the reference block and at least one other reference image, and a synthesized reference block generation calculated by the first reference image synthesizing unit A reference image synthesis parameter storage unit (122) for storing information necessary for the second reference block, and a second reference block based on information necessary for generating a synthesized reference block for the encoded region stored in the reference image synthesis parameter storage unit. And a second reference image synthesis unit (123) that generates a second synthesized reference block that is synthesized by specifying a predetermined area of the reference image and a reference image synthesis parameter storage unit. A reference image synthesis selection unit (1224) that inputs the information necessary for generating the synthesized reference block of the encoded region and determines a synthesis method for the second reference image synthesis unit, and includes a second reference image synthesis Or a first reference image synthesis unit has a function of generating a third synthesized reference block obtained by synthesizing the second reference block and a predetermined region of at least one other reference image, and the reference image synthesis selection unit A correlation value between a plurality of reference blocks generated using information necessary for generating the synthesized reference block for the encoding target block, and a plurality generated using information necessary for generating the synthesized reference block for the encoded region And comparing the correlation values between the reference blocks and outputting the second synthesized reference block from the second reference image synthesizer, or from the second reference image synthesizer or the first reference image synthesizer, 3 synthesis reference blocks to be output (S1414-S1416, S1522-S1523, S1418-S1420, S1524-S1525), and at least the first synthesis reference block and the second synthesis reference block An encoding unit is provided that selects a prediction block for an encoding target block from a plurality of prediction blocks including a block or a third synthesis reference block, and encodes a prediction difference block obtained by subtracting the prediction block from the encoding target block. .

An error value between the first predicted image and the second predicted image in the adjacent block referred to, and an error value between the first predicted image and the second predicted image with respect to the encoding target block are calculated, and the continuity between the reference images is calculated. If there is no continuity, a third predicted image is generated by calculating inter-reference image motion vector information between the first predicted image and the second reference image. By generating the synthesized motion compensated predicted image by synthesizing the first predicted image and the third predicted image, the motion compensated prediction of the decoding side motion vector calculation type can be performed again only when necessary. Thus, it is possible to further improve the coding efficiency without using additional information while suppressing an increase in the amount of calculation.

The moving image decoding apparatus according to an aspect of the present invention includes a motion vector decoding unit (212) that decodes a motion vector for a decoding target block from an encoded stream, and a first extracted from the first reference image using the motion vector. The first reference image composition unit (215) that generates a first synthesized reference block obtained by synthesizing the reference block and the predetermined region of at least one other reference image, and the synthesized reference calculated by the first reference image synthesis unit The reference image synthesis parameter storage unit (216) that stores information necessary for block generation and the information necessary for generating the synthesized reference block of the decoded area stored in the reference image synthesis parameter storage unit are used as the second reference. A second reference image synthesis unit (217) that generates a second synthesized reference block obtained by identifying and synthesizing a predetermined area of the reference image used for synthesis with the block. And a prediction mode decoding unit (203) that decodes prediction mode selection information selected from a plurality of prediction blocks including at least a first combined reference block and a second combined reference block as a decoding target block from the encoded stream. ), And a decoding unit that generates a decoded image by adding the prediction block selected by the prediction mode selection information and the prediction difference block decoded from the decoding target block.

Another aspect of the present invention is also a video decoding device. This apparatus includes a motion vector decoding unit (212) that decodes a motion vector for a decoding target block from an encoded stream, a first reference block extracted from a first reference image using a motion vector, and at least one other A first reference image synthesis unit (215) that generates a first synthesized reference block obtained by synthesizing predetermined areas of two reference images, and information necessary for generating a synthesized reference block calculated by the first reference image synthesis unit is stored. The reference image used for combining with the second reference block based on the information necessary for generating the combined reference block of the decoded area stored in the reference image combining parameter storage unit (216) and the reference image combining parameter storage unit The second reference image synthesis unit (217) that generates a second synthesized reference block that is synthesized by specifying a predetermined area of A prediction mode decoding unit (203) for decoding prediction mode selection information selected from a plurality of prediction blocks in which the target block includes at least a first synthesized reference block and a second synthesized reference block; and reference image synthesis parameter storage A reference image synthesis selection unit (1318) that inputs information necessary for generating a synthesized reference block of a decoded area stored in the unit and determines a synthesis method for the second reference image synthesis unit, and includes a reference image synthesis Generated using a correlation value between a plurality of reference blocks generated using information necessary for generating the synthesized reference block for the decoding target block and information necessary for generating the synthesized reference block for the decoded region The correlation values between the plurality of reference blocks thus obtained are compared, and the output from the second reference image synthesizing unit is converted into the second synthesized reference block and the second reference block. A function (S1426 to S1429) for selecting at least one of a plurality of blocks including at least a first combined reference block and a second combined reference block or a second reference block as a prediction block based on prediction mode selection information. A decoding unit is provided that generates a decoded image by adding a prediction block selected from the prediction blocks and a prediction difference block decoded from the decoding target block.

Another aspect of the present invention is also a video decoding device. The apparatus includes a motion vector decoding unit (212) that decodes a motion vector for a decoding target block from an encoded stream, a first reference block extracted from the first reference image using a motion vector, A first reference image synthesis unit (215) that generates a first synthesized reference block obtained by synthesizing a predetermined region of at least one reference image, and information necessary for generating a synthesized reference block calculated by the first reference image synthesis unit. The reference image synthesis parameter storage unit (216) to be stored and the information necessary for generating the synthesized reference block of the decoded area stored in the reference image synthesis parameter storage unit are used for synthesis with the second reference block. A second reference image synthesis unit (217) that generates a second synthesized reference block that is synthesized by specifying a predetermined region of the reference image, and an encoded stream. A prediction mode decoding unit (203) that decodes prediction mode selection information in which a decoding target block is selected from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block; and reference image synthesis A reference image synthesis selection unit (1318) for inputting information necessary for generating a synthesized reference block of the decoded area stored in the parameter storage unit and determining a synthesis method for the second reference image synthesis unit; A second reference image synthesis unit or the first reference image synthesis unit has a function of generating a third synthesized reference block obtained by synthesizing the second reference block and a predetermined region of at least one other reference image; Correlation values between a plurality of reference blocks generated by the image synthesis selection unit using information necessary for generating a synthesis reference block for the decoding target block, and a decoded area A correlation value between a plurality of reference blocks generated using information necessary for generating a synthesized reference block is compared, and a second synthesized reference block is output from the second reference image synthesizer, or a second reference image synthesized Or the first reference image synthesis unit has a function (S1426-S1428, S1530-S1531) for selecting whether to output the third synthesis reference block, and at least as a prediction block from the prediction mode selection information A decoded image by adding a prediction block selected from a plurality of prediction blocks including the first synthesis reference block and the second synthesis reference block or the third synthesis reference block and a prediction difference block decoded from the decoding target block The decoding part which produces | generates is provided.

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

According to the present invention, it is possible to improve the quality of motion-compensated prediction by improving the quality of a predicted image while suppressing an increase in the amount of calculation in the encoding device / decoding device.

It is a block diagram of the moving image encoder in Embodiment 1 of this invention. It is a block diagram of the moving image decoding apparatus in Embodiment 1 of this invention. It is a conceptual diagram which shows the operation | movement of the reference image synthetic | combination motion compensation prediction in this invention. It is a flowchart of the prediction block production | generation process in Embodiment 1 of this invention. It is a process flowchart of the 1st reference picture synthetic | combination motion compensation prediction in Embodiment 1 of this invention. It is a block diagram of the motion vector detection part between reference images in Embodiment 1 of this invention. It is a figure which shows an example of the motion vector detection range between the reference images in Embodiment 1 of this invention. It is a process flowchart of the motion vector detection part between reference images in Embodiment 1 of this invention. It is a process flowchart of the 2nd reference picture synthetic | combination motion compensation prediction in Embodiment 1 of this invention. It is a figure which shows the management form of the reference image synthetic | combination parameter storage part in Embodiment 1 of this invention. It is a conceptual diagram of the selection result of the 1st, 2nd reference image synthetic | combination motion compensation in Embodiment 1 of this invention. It is a block diagram of the moving image encoder in Embodiment 2 of this invention. It is a block diagram of the moving image decoding apparatus in Embodiment 2 of this invention. It is a flowchart of the reference image synthetic | combination selection process by the side of an encoding in Embodiment 2 of this invention. It is a flowchart of the reference image synthetic | combination selection process by the side of decoding in Embodiment 2 of this invention. It is a flowchart of the reference image synthetic | combination selection process by the side of an encoding in Embodiment 3 of this invention. It is a flowchart of the reference image synthetic | combination selection process by the side of decoding in Embodiment 3 of this invention.

EMBODIMENT OF THE INVENTION Below, the form for inventing is demonstrated with reference to drawings.
(Embodiment 1)
First, the moving picture coding apparatus according to the first embodiment will be described. FIG. 1 is a configuration diagram showing the configuration of the moving picture encoding apparatus according to the first embodiment.

As shown in FIG. 1, the moving picture coding apparatus according to Embodiment 1 includes an input terminal 100, an input picture buffer 101, a block division unit 102, a subtractor 103, an orthogonal transformation unit 104, a quantization unit 105, and an inverse quantization. Unit 106, inverse orthogonal transform unit 107, adder 108, intra-frame decoded image memory 109, decoded reference image memory 110, entropy encoding unit 111, stream buffer 112, output terminal 113, code amount control unit 114, prediction mode determination unit 115, intra-frame prediction unit 116, motion vector detection unit 117, motion compensation prediction unit 118, motion vector prediction unit 119, inter-reference image motion vector detection unit 120, first reference image synthesis motion compensation prediction unit 121, reference image synthesis parameter The storage unit 122 and the second reference image synthesis motion compensation prediction unit 123 are configured.

Operation of this processing block in that an inter-reference image motion vector detection unit 120, a first reference image synthesis motion compensation prediction unit 121, a reference image synthesis parameter storage unit 122, and a second reference image synthesis motion compensation prediction unit 123 are provided. However, this is a feature of the first embodiment of the present invention, and the same processing as the processing blocks constituting the encoding processing in the moving image encoding apparatus such as MPEG4-AVC can be applied to the other processing blocks.

The digital image signal input from the input terminal 100 is stored in the input image buffer 101. The digital image signal stored in the input image buffer 101 is supplied to the block dividing unit 102, and is cut out as an encoding target block in units of blocks composed of horizontal N pixels × vertical M pixels. The values of N and M can be selectively configured from a plurality of pixels that can be set in advance. However, in the description of the first embodiment, it is assumed that N = 16 and M = 16 are fixed. The block dividing unit 102 extracts the encoded target block from the intra-frame prediction unit 116, the motion vector detection unit 117, the motion compensation prediction unit 118, the first reference image synthesis motion compensation prediction unit 121, and the second reference image synthesis motion compensation. This is supplied to the prediction unit 123 and the subtracter 103.

The subtractor 103 calculates a difference between the encoding target block supplied from the block dividing unit 102 and the predicted image block supplied from the prediction mode determining unit 115, and supplies the result to the orthogonal transform unit 104 as a difference block. . The operation of the prediction mode determination unit 115 will be described later.

The orthogonal transform unit 104 generates DCT coefficients corresponding to the orthogonally transformed frequency component signal by performing DCT transform on the difference block in a predetermined unit. In the description of the first embodiment, it is assumed that the unit for performing DCT conversion is a 4 × 4 pixel unit or an 8 × 8 pixel unit. Further, the orthogonal transform unit 104 collects the generated DCT coefficients in units of encoding target blocks and outputs them to the quantization unit 105.

The quantization unit 105 performs quantization processing by dividing the DCT coefficient by a different value for each frequency component. The quantization unit 105 supplies the quantized DCT coefficient to the inverse quantization unit 106 and the entropy coding unit 111.

The inverse quantization unit 106 performs inverse quantization by multiplying the quantized DCT coefficient input from the quantization unit 105 by a value divided at the time of quantization, and the result of the inverse quantization is obtained. The decoded DCT coefficient is output to the inverse orthogonal transform unit 107.

The inverse orthogonal transform unit 107 performs inverse DCT processing to generate a decoded difference block. The inverse orthogonal transform unit 107 supplies the decoded difference block to the adder 108.

The adder 108 adds the prediction image block supplied from the prediction mode determination unit 115 and the decoded difference block supplied from the inverse orthogonal transform unit 107 to generate a local decoding block. The local decoded block generated by the adder 108 is stored in the intra-frame decoded image memory 109 and the decoded reference image memory 110 in a form subjected to inverse block conversion. In the case of MPEG-4 AVC, before local decoding blocks are input to the decoded reference image memory 110, adaptive filtering is applied to block boundaries where coding distortion for each block tends to appear as a boundary. In some cases, processing to be performed is performed.

The entropy encoding unit 111 receives the quantized DCT coefficient supplied from the quantization unit 105, the prediction mode information supplied from the prediction mode determination unit 115, and additional information that needs to be transmitted according to the prediction mode. In contrast, variable-length coding of each piece of information is performed. Specifically, in the case of intra-frame prediction, the intra prediction mode and the prediction block size information are used. In the case of motion compensated prediction and reference image synthesized image motion compensated prediction, the prediction block size, reference image designation information, and motion are used. The difference value between the vector and the predicted motion vector value is information that requires encoding. The information subjected to variable length coding is output as a coded bit stream from the entropy coding unit 111 to the stream buffer 112.

The encoded bit stream stored in the stream buffer 112 is output to a recording medium or a transmission path via an output terminal 113. Regarding the code amount control of the encoded bit stream, the code amount control unit 114 is supplied with the code amount of the encoded bit stream stored in the stream buffer 112 and compared with the target code amount. In order to approach the target code amount, the fineness of quantization (quantization scale) of the quantization unit 105 is controlled.

The prediction mode determination unit 115 is a prediction mode for each prediction method input from the intra-frame prediction unit 116, the motion compensation prediction unit 118, the first reference image synthesis motion compensation prediction unit 121, and the second reference image synthesis motion compensation prediction unit 123. From the prediction block, the prediction signal with the least difference information to be encoded is selected for the encoding target block input from the block dividing unit 102, and the subtraction unit 103 adds the prediction image block for the selected prediction method. To the entropy coding unit 111, the motion vector prediction unit 119, and the reference image synthesis parameter storage unit 122, the prediction mode information as additional information and the information that needs to be transmitted according to the prediction mode. Output. Regarding the motion vector value, difference information from the prediction vector value calculated by the motion vector prediction unit 119 described later is sent to the entropy encoding unit 111 and encoded.

In the intra-frame prediction unit 116, the decoding target image input from the block division unit 102 and the decoded image of the area that has been encoded with respect to the periphery of the encoding target block, stored in the intra-frame decoding image memory 109. Is input, and prediction using correlation within the frame is performed. In the case of MPEG4-AVC, pixel values are predicted in a plurality of predetermined directions in units of 4 × 4 pixels, 8 × 8 pixels, and 16 × 16 pixels for the encoding target block, and prediction processing is performed. The prediction using the correlation between adjacent pixels in the screen is performed using a method called intra prediction that generates a prediction block together with information indicating the unit and the selected direction (intra prediction mode). The predicted image block and the selected intra prediction mode are output from the intra-frame prediction unit 116 to the prediction mode determination unit 115.

In the motion vector detection unit 117, an encoding target block input from the block division unit 102 and a decoded image of a frame that has been encoded on the entire screen and stored in the decoded reference image memory 110 are input as a reference image. Then, motion estimation is performed between the encoding target block and the reference image. As a general motion estimation process, a reference image at a position moved by a predetermined movement amount from the same position on the screen is cut out, and the movement amount that minimizes the prediction error when the image is used as a prediction block is determined as a motion vector. As a value, a block matching process that is obtained while changing the movement amount is used. As the evaluation value of the prediction error, an absolute value sum of errors (SAD), a sum of squares of error (SSE), or the like is used, and further evaluation is performed by taking into account the amount of code of additional information such as a motion vector. When it is converted into a motion vector, it is possible to calculate an efficient motion vector. The detected motion vector value is output to the motion compensation prediction unit 118 and the inter-reference image motion vector detection unit 120.

The motion compensation prediction unit 118 receives the motion vector value obtained by the motion vector detection unit 117, and obtains motion compensation prediction images for a plurality of block sizes and a plurality of reference images that are equal to or less than the encoding target block, and the motion vector detection unit 117. The prediction block with the least difference information to be encoded is selected and the difference information to be encoded is the least with respect to the encoding target block acquired from the decoded reference image memory 110 via and input from the block dividing unit 102 Select a prediction block. The motion compensation prediction unit 118 outputs the selected motion compensation prediction mode, the motion vector value, and the prediction block to the prediction mode determination unit 115.

The motion vector prediction unit 119 calculates a predicted motion vector value using the motion vectors of the surrounding encoded blocks, and sends the motion vector detection unit 117, the motion compensation prediction unit 118, and the first reference image synthesis motion compensation prediction unit 121 to each other. Supply.

Using the predicted motion vector value, the motion vector detection unit 117 detects an optimal motion vector value in consideration of the code amount necessary for encoding the difference between the motion vector predicted value and the motion vector value. To do. Similarly, the motion compensation prediction unit 118 takes into account the amount of code required when encoding the difference between the motion vector prediction value and the transmitted motion vector value, and is used as a block unit for optimal motion compensation prediction. Select image and motion vector values.

The inter-reference image motion vector detection unit 120 extracts the first reference block from the reference image corresponding to the motion vector, from the motion vector input from the motion vector detection unit 117. The inter-reference image motion vector detection unit 120 calculates an error value by block matching or the like for the motion vector between the extracted first reference block and another reference image, and uses the motion vector having the small value as the inter-reference image motion. Calculate as a vector. The inter-reference image motion vector detection unit 120 calculates a motion vector value for generating the first reference block and a motion vector value between the other reference image and the encoding target block calculated based on the inter-reference image motion vector. It outputs to the 1st reference image synthetic | combination motion compensation prediction part 121. FIG.

The first reference image synthesis motion compensation prediction unit 121 includes a motion vector value input from the inter-reference image motion vector detection unit 120, and another reference image calculated based on the inter-reference image motion vector and an encoding target block. The first prediction block and the second prediction block are acquired from the decoded reference image memory 110 based on the motion vector values of the two, and a reference image synthesis motion compensation prediction block is generated by synthesizing these prediction blocks. Similarly to the motion compensation prediction unit 118, the first reference image synthesis motion compensation prediction unit 121 also calculates the difference between the motion vector prediction value and the motion vector value to be transmitted for the encoding target block input from the block division unit 102. A motion vector value to be used as a block unit for optimal reference image synthesis motion compensation prediction is selected in consideration of a code amount necessary for encoding.

The first reference image synthesis motion compensation prediction unit 121 outputs the selected motion compensation prediction mode, the motion vector value, and the prediction block to the prediction mode determination unit 115, and other reference images calculated from the inter-reference image motion vector. And a motion vector value between the encoding target block and the encoding target block are output to the reference image synthesis parameter storage unit 122. Details of the operation of the first reference image synthesis motion compensation prediction unit 121 will be described later.

The reference image synthesis parameter storage unit 122 is input from the first reference image synthesis motion compensation prediction unit 121, and a motion vector value between another reference image calculated from the inter-reference image motion vector and the encoding target block. Prediction mode information as additional information input from the prediction mode determination unit 115, information that needs to be transmitted according to the prediction mode, and second reference image synthesis input from the second reference image synthesis motion compensation prediction unit 123 Parameters necessary for reference image synthesis prediction for motion compensated prediction are stored, and necessary for reference image synthesis prediction in the encoded peripheral blocks of the target block for the second reference image synthesis motion compensation prediction unit 123. Output parameters.

The second reference image synthesis motion compensation prediction unit 123 performs the encoding in the peripheral blocks that have been encoded of the encoding target block input from the reference image synthesis parameter storage unit 122 with respect to the encoding target block input from the block division unit 102. The first prediction block and the second prediction block are obtained from the decoded reference image memory 110 using the parameters used for the reference image synthesis, and a reference image synthesis motion compensation prediction block is generated by synthesizing these prediction blocks. To do.

The second reference image synthesis motion compensation prediction unit 123 outputs the selected motion compensation prediction mode and the prediction block to the prediction mode determination unit 115. Since the prediction block generated by the second reference image synthesis motion compensation prediction unit 123 is generated from the peripheral information of the decoded block, a motion vector is not transmitted as additional information.

Also, the second reference image synthesis motion compensation prediction unit 123 outputs a parameter used for reference image synthesis used in the selected motion compensation prediction mode to the reference image synthesis parameter storage unit 122. Details of the operation of the second reference image synthesis motion compensation prediction unit 123 will be described later.

Subsequently, a moving picture decoding apparatus that decodes the encoded bitstream generated by the moving picture encoding apparatus according to Embodiment 1 will be described. FIG. 2 is a configuration diagram of the moving picture decoding apparatus according to the first embodiment.

As shown in FIG. 2, the moving picture decoding apparatus according to Embodiment 1 includes an input terminal 200, a stream buffer 201, an entropy decoding unit 202, a prediction mode decoding unit 203, a prediction image selection unit 204, an inverse quantization unit 205, and an inverse. Orthogonal transformation unit 206, adder 207, intra-frame decoded image memory 208, decoded reference image memory 209, output terminal 210, intra-frame prediction unit 211, motion vector prediction decoding unit 212, motion compensation prediction unit 213, inter-reference image motion vector The detection unit 214 includes a first reference image synthesis motion compensation prediction unit 215, a reference image synthesis parameter storage unit 216, and a second reference image synthesis motion compensation prediction unit 217.

The reference inter-motion vector detection unit 214, the first reference image synthesis motion compensation prediction unit 215, the reference image synthesis parameter storage unit 216, and the second reference image synthesis motion compensation prediction unit 217 are provided, and in these processing blocks The operation is a feature of the decoding apparatus according to the first embodiment of the present invention. These operations are paired with the same functional block of the moving picture coding apparatus shown in FIG. 1 to generate a motion compensated prediction block with reduced transmission of additional information. For the other processing blocks, the same processing as the processing blocks constituting the decoding processing in the moving image decoding apparatus such as MPEG4-AVC can be applied.

The encoded bit stream input from the input terminal 200 is supplied to the stream buffer 201, and the stream buffer 201 absorbs the code amount variation of the encoded bit stream and is supplied to the entropy decoding unit 202 in a predetermined unit such as a frame. The The entropy decoding unit 202 performs variable-length decoding on the encoded prediction mode information, the additional information corresponding to the prediction mode, and the quantized DCT coefficient from the encoded bitstream input via the stream buffer 201. Then, the quantized DCT coefficient is output to the inverse quantization unit 205, and the prediction mode information and additional information corresponding to the prediction mode are output to the prediction mode decoding unit 203.

Regarding the inverse quantization unit 205, the inverse orthogonal transform unit 206, the adder 207, the intra-frame decoded image memory 208, and the decoded reference image memory 209, the local decoding process of the moving image coding apparatus according to the first embodiment of the present invention. Processing similar to that of a certain inverse quantization unit 106, inverse orthogonal transform unit 107, adder 108, intra-frame decoded image memory 109, and decoded reference image memory 110 is performed. The decoded image stored in the decoded reference image memory 209 is displayed as a decoded image signal on the display device via the output terminal 210.

In the prediction mode decoding part 203, when motion compensation prediction is selected as a prediction mode from the prediction mode information input from the entropy decoding part 202 and the additional information corresponding to the prediction mode, the motion vector prediction decoding part 212 The information indicating the predicted block unit, the information indicating the motion compensation prediction mode, the first reference image synthesis motion compensation prediction mode, the second reference image synthesis motion compensation prediction mode, and the decoded difference vector value are output and predicted. Prediction mode information is output to the image selection unit 204 and the reference image synthesis parameter storage unit 216. The prediction mode decoding unit 203 also determines the intra-frame prediction unit 211, the motion compensation prediction unit 213, the first reference image synthesis motion compensation prediction unit 215, and the second reference image synthesis motion compensation prediction according to the decoded prediction mode information. Information indicating the selection and additional information according to the prediction mode are output to the unit 217.

The prediction image selection unit 204, according to the prediction mode information input from the prediction mode decoding unit 203, the intra-frame prediction unit 211, the motion compensation prediction unit 213, the first reference image synthesis motion compensation prediction unit 215, and the second reference The predicted image for the decoding target block output from any of the image synthesis motion compensation prediction unit 217 is selected and output to the adder 207.

When the decoded prediction mode indicates intra-frame prediction, the intra-frame prediction unit 211 receives the intra prediction mode as additional information according to the prediction mode from the prediction mode decoding unit 203, and according to the intra prediction mode. The decoded image of the region where decoding is completed is input to the periphery of the decoding target block stored in the intra-frame decoded image memory 208, and prediction using the intra-frame correlation is performed in the same intra prediction mode as the encoding device. Done. The intra-frame prediction unit 211 outputs the intra-frame prediction block generated by the prediction to the predicted image selection unit 204.

The motion vector predictive decoding unit 212 uses the motion vector of the neighboring decoded block for the decoded difference vector value input from the prediction mode decoding unit 203, and performs the motion prediction using the same method as that performed by the encoding device. A value obtained by calculating a vector value and adding the difference vector value and the predicted motion vector value is used as a motion vector value of the decoding target block, as a motion compensated prediction unit 213, a reference image motion vector detection unit 214, and a reference image synthesis parameter storage unit. To 216. The motion vectors are decoded by the number encoded according to the block unit of the prediction process indicated in the motion compensation prediction mode or the first reference image synthesized image motion compensation prediction mode.

The motion compensated prediction unit 213 generates a motion compensated prediction block from the decoded reference image memory 209 from the motion vector value input from the motion vector predictive decoding unit 212, and sends the generated motion compensated prediction block to the predicted image selection unit 204. Output.

The inter-reference image motion vector detection unit 214 extracts a first reference block used for motion compensated prediction from the decoded reference image memory 209 from the motion vector value input from the motion vector prediction decoding unit 212. Subsequently, an error value is calculated by block matching or the like for the motion vector between the extracted first reference block and another reference image, and a motion vector having a small value is calculated as a motion vector between reference images.

Subsequently, the inter-reference image motion vector detection unit 214 generates a motion vector value between the motion vector value for generating the first reference block and another reference image calculated from the inter-reference image motion vector and the encoding target block. The value is output to the first reference image synthesis motion compensation prediction unit 215.

The first reference image synthesis motion compensation prediction unit 215 includes a motion vector value input from the inter-reference image motion vector detection unit 214 and between another reference image calculated based on the inter-reference image motion vector and the encoding target block. The first prediction block and the second prediction block are obtained from the decoded reference image memory 209 based on the motion vector values of the two, and a reference image synthesis motion compensation prediction block is generated by synthesizing these prediction blocks. The generated reference image synthesis motion compensation prediction block is output to the prediction image selection unit 204.

In addition, the first reference image synthesis motion compensation prediction unit 215 outputs a motion vector value between another reference image and a decoding target block calculated by the inter-reference image motion vector to the reference image synthesis parameter storage unit 216. . Details of the operation of the first reference image synthesis motion compensation prediction unit 215 will be described later.

The reference image synthesis parameter storage unit 216 receives a motion vector value between another reference image calculated from the inter-reference image motion vector and the decoding target block, which is input from the first reference image synthesis motion compensation prediction unit 215, Prediction mode information as additional information input from the prediction mode decoding unit 203, motion vector information input from the motion vector prediction decoding unit 212, and a second reference input from the second reference image synthesis motion compensation prediction unit 217 Parameters necessary for the reference image synthesis prediction for the image synthesis motion compensation prediction are stored, and the second reference image synthesis motion compensation prediction unit 217 stores the parameters necessary for the reference image synthesis prediction in the decoded peripheral block of the decoding target block. Output parameters.

The second reference image synthesis motion compensation prediction unit 217 uses the parameters used for reference image synthesis in the decoded peripheral blocks of the decoding target block input from the reference image synthesis parameter storage unit 216, and uses the parameters used for reference image synthesis from the decoded reference image memory 209. 1 prediction block and 2nd prediction block are acquired, and a reference image synthetic | combination motion compensation prediction block is produced | generated by synthesize | combining these prediction blocks. The generated reference image synthesis motion compensation prediction block is output to the prediction image selection unit 204.

Also, the second reference image synthesis motion compensation prediction unit 217 outputs a parameter used for reference image synthesis used in the selected motion compensation prediction mode to the reference image synthesis parameter storage unit 216. Details of the operation of the second reference image synthesis motion compensation prediction unit 217 will be described later.

Hereinafter, the prediction image generation operation of the reference image synthesis motion compensation prediction that operates in the moving image encoding device and the moving image decoding device of Embodiment 1 will be described using the conceptual diagram of FIG.

FIGS. 3b) and 3c) are conceptual diagrams showing reference image synthesis motion compensation prediction in the invention. FIG. 3a) is a conceptual diagram of bidirectional motion compensated prediction using two reference images used in MPEG4-AVC.

FIG. 3a) detects a motion vector between two reference images with a target block, transmits a motion vector for each reference image, and averages the reference blocks indicated by the two motion vectors. Is a prediction image. A prediction image having a function of removing an encoding degradation component as a motion adaptive filter in the time direction by synthesizing two reference images and a function of following a minute luminance change component of an encoding object by averaging. Can be generated.

In MPEG4-AVC, a plurality of decoded reference images are stored, and a reference image number and a motion vector used for prediction are transmitted in a predetermined block unit to adaptively select the reference image. Yes. In the case of FIG. 3 a), four decoded images are secured as reference images, and two prediction images are acquired using the reference image 1 and the reference image 3, and bidirectional prediction is performed. .
As the motion vectors, mvL0 and mvL1 are transmitted after taking a difference value from the predicted motion vector.

In contrast, in FIG. 3b), a reference image serving as a reference is specified, a motion vector is detected, and a reference image between the first reference block acquired using the motion vector mvL0 and another reference image By obtaining the inter-motion vector mvInterRef, a method of generating the motion mvL1 between the encoding target image and another reference image without transmitting the motion vector is adopted. On the decoding side, mvL1 can be generated by performing similar processing using the transmitted motion vector mvL0.

The method disclosed in Patent Document 1 evaluates an error value of a block that is symmetrical with an encoding target block between two reference images when the motion of an object in an adjacent image across the encoding target image is uniform. By doing so, it is a method to generate mvL0 and mvL1 without transmitting motion vectors, but it is effective when mvL0 and mvL1 are generated with a limited motion and there is little continuity in time Therefore, it is impossible to generate a prediction block with a sufficiently small prediction error. Also, a large motion search range is required to generate an appropriate motion for an image with a large motion, and the amount of calculation required for motion search processing in the encoding device / decoding device increases.

In the configuration of FIG. 3b), other reference images are obtained from mvL0 that transmits an appropriate reference block as a prediction block of an encoding target block, and the generated prediction image block is both encoded and decoded. Are detected by detecting a motion vector between reference images, and the other motion vector value is implicitly calculated, thereby obtaining a motion vector suitable for prediction of a coding target block with a small search range. This makes it possible to achieve appropriate bi-directional prediction without involving motion vectors even in moving image signals with little spatial and temporal continuity, and greatly improve coding efficiency. .

In addition, as in the configuration of FIG. 3c), it is also possible to generate three or more reference images for generating a predicted image and detect each of the motion vectors mvL1 and mvL2 by detecting a motion vector between the reference images. It is possible to transmit a single motion vector in the same way as in FIG. 3b) by reducing the efficiency reduction of the predicted image due to motion accompanied by coding distortion and deformation by combining with many reference images. The encoding efficiency can be further improved.

In the present invention, in addition to the improvement of the encoding efficiency by the reference image synthesis motion compensation prediction, the parameters necessary for the reference image synthesis motion compensation prediction used in the encoded / decoded block are used on the decoding side. By introducing a second reference image synthesis motion compensation prediction process that calculates a motion vector between reference pictures without using a motion vector detection process and performs a reference image synthesis motion compensation prediction, motion continuity is spatially reduced. For a certain image, a motion vector generated using motion vector detection between reference images in a small area can be applied in a wider range, greatly increasing the amount of computation required for motion vector detection in a decoding device. In addition, it is possible to realize encoding with improved efficiency of a predicted image by transmitting fewer motion vectors.

Next, the operation of the prediction block generation process in Embodiment 1 will be described using the flowchart of FIG.

On the encoding side, intra-frame prediction is performed (S400), and in a frame for which inter-frame prediction is performed, a first motion vector is detected between the encoding target block and the decoded reference image (S401). One-way and two-way motion compensation prediction, which is a conventional motion compensation prediction, is performed using a motion vector (S402). Subsequently, a second motion vector is calculated between the first prediction block generated using the first motion vector and a plurality of reference images (S403).

The first reference image synthesis motion compensation prediction is performed using the first motion vector and the calculated second motion vector (S404). Subsequently, the second reference image synthesis motion compensation prediction is performed using the first motion vector and the second motion vector calculated in the decoded peripheral block (S405).

Additional information necessary for each prediction from intra-frame prediction / one-way / bidirectional motion compensation prediction / first reference image synthesis motion compensation prediction / second reference image synthesis motion compensation prediction obtained by these processes. Is selected and output based on the error amount calculated from the encoding amount and the encoding distortion (S406).

Finally, parameters necessary for generating the selected prediction block are recorded in the reference image synthesis parameter storage unit (S407). Information to be recorded includes information indicating an optimal prediction block, reference image information used for motion compensation prediction, and a motion vector.

On the decoding side, the prediction mode information is decoded, and when the prediction mode is the intra-frame prediction mode (S410: YES), intra-frame prediction is performed (S411). If not (S410: NO), if the prediction mode is the second reference image synthesis motion compensation prediction mode (S412: YES), decoding is performed with the decoded peripheral blocks recorded in the reference image synthesis parameter storage unit. The second reference image synthesis motion compensation prediction is performed using the first motion vector and the second motion vector (S413).

When the prediction mode is not the second reference image synthesis motion compensation prediction mode (S412: NO), and when the prediction mode is not the first reference image synthesis motion compensation prediction mode (S414: NO), the conventional motion compensation prediction process is performed. Is performed, and motion compensation prediction of one-way or bidirectional prediction is performed using the decoded first motion vector (S415). When the prediction mode is the first reference image synthesis motion compensation prediction mode (S414: YES), the second prediction is performed between the first prediction block generated using the decoded first motion vector and the plurality of reference images. A motion vector is calculated (S416). Then, using the calculated second motion vector and first motion vector, first reference image synthesis motion compensation prediction is performed (S417).

The prediction block generated by these prediction processes is output as the prediction image information at the time of decoding (S418), and the parameters necessary for generating the selected prediction block are used for the decoding process of the subsequent decoding block. It is recorded in the reference image synthesis parameter storage unit (S419).

Next, the operation of the first reference image synthesis motion compensation prediction process in Embodiment 1 will be described using the flowchart of FIG. As a block of the encoding / decoding device, an inter-reference image motion vector detection unit (120 in FIG. 1, 214 in FIG. 2) and a first reference image synthesis motion compensation prediction unit (121 in FIG. 1, 215 in FIG. 2) are used. This is a description of the detailed operations performed.

The encoding side first determines a standard reference image for the encoding target block (S500). Since the reference image can be selectively specified by transmitting information indicating the reference image to the predictable reference image at the time of encoding, all the reference images are set as the standard reference image and the subsequent processing is performed. It is possible to generate an optimal prediction block.

Subsequently, the first motion vector value mvL0 base for the standard reference image detected by the motion vector detection unit (117 in FIG. 1) is input (S501). As an embodiment, a motion vector value used for normal motion compensation prediction is input, but it is also possible to input a motion vector value specified by another method.

Next, a prediction block using mvL0 rounded, which is a value obtained by rounding mvL0 base to integer pixel precision, is generated (S502). For example, when the input motion vector value mvL0 base has a 1/4 pixel accuracy, mvL0 rounded is obtained by calculation as follows.
mvL0 rounded = (mvL0 base +2) >> 2 (calculated separately for horizontal and vertical)

The inter-reference image motion vector value mvInterRef is calculated between the prediction block generated in this way and another reference image (S503). Details of this processing will be described in the description of the inter-reference image motion vector detection unit (120 in FIG. 1 and 214 in FIG. 2). Here, by adding the generated mvL0edrounded and mvInterRef, mvL1 base serving as a reference for the second motion vector is calculated (S504).

Since mvL0 base is rounded to integer pixel accuracy in S502, the information for the accuracy of the motion vector transmitted at the time of encoding can be used as information for increasing the accuracy of the multi-reference image prediction block. Define the motion in the range that becomes 0 when rounded to integer pixel precision as a phase shift vector phase vector, and use the motion vector value obtained by adding phase vector to mvL0 rounded and mvL1 base respectively, to predict a prediction block from two reference images A prediction error value with respect to the encoding target block is calculated as a prediction block that is generated and the average of which is a candidate (S505).

Specifically, when the phase vector is 1/4 pixel precision, both horizontal and vertical can be transmitted in addition to the motion vector value within the range of -1 / 2 ≦ phase vector <1/2. By moving the phase vector within this range, each prediction error value is calculated. During decoding, the phase vector can be reproduced as the difference between the transmitted mvL0 and the mvL0 rounded rounded to integer pixel precision.

Finally, the first motion vector value mvL0 and the second motion vector value mvL1 in the first reference image synthesis motion compensated prediction block are obtained by adding the motion vector value obtained by adding the phase vector that takes the minimum prediction error value to mvL0 rounded and mvL1 base. Are output together with the prediction block bidirectionally predicted from mvL0 and mvL1 (S506). In the case of the first reference image synthesis motion compensated prediction, only the output mvL0 information is transmitted as a motion vector value, the mvL1 information is not transmitted, and is generated by detecting the inter-reference image motion vector on the decoding side. It becomes the composition to do.

In the operation of the first reference image synthesis motion compensation prediction on the decoding side, first, a reference reference image for the encoding target block is determined (S510). In this case, when the decoded prediction mode is the first reference image synthesis motion compensated prediction, information specifying the reference image is decoded as information indicating the unidirectional prediction of the normal motion compensated prediction. Can be confirmed.

Subsequently, the first motion vector value mvL0 decoded for the standard reference image is input from the motion vector prediction decoding unit (212 in FIG. 2) (S511), and the value mvL0vrounded by rounding mvL0 to integer pixel precision. Is used to generate a prediction block (S512).

The inter-reference image motion vector value mvInterRef is calculated between the prediction block generated in this way and another reference image (S513). This process is the same as the motion vector detection between reference images on the encoding side. Here, by adding the generated mvL0 rounded and mvInterRef, mvL1 base serving as a reference for the second motion vector is calculated (S514).

Next, the difference between mvL0 and mvL0 rounded is calculated as a phase vector (S515). MvL1 is calculated by adding the calculated phase vector to mvL1 base (S516). Finally, using the calculated mvL0 and mvL1, bi-predicted prediction blocks are output together with mvL0 and mvL1 (S517).

In the first reference image synthesis motion compensated prediction in the first embodiment, mvL0 rounded, which is a reference for obtaining the inter-reference image motion vector value, is an integer pixel. It is possible to perform image synthesis motion compensated prediction, and when using a motion vector value with the same accuracy as that of the finally transmitted motion vector as mvL0 rounded, the phase vector is fixed to 0. The motion vector value cannot be detected in consideration of the error evaluation at the time of synthesizing the reference image on the encoding side, but the inter-reference image motion vector detection is performed based on the prediction block with high accuracy. Is called.

In the first embodiment, since mvL0 rounded is an integer pixel, it is not necessary to perform a filtering process for generating a prediction block with a small pixel precision on the first prediction block when detecting a motion vector between reference images. There is an advantage that a motion vector value can be generated in consideration of an error evaluation at the time of synthesizing a reference image on the encoding side.

Next, the description of the inter-reference image motion vector detection unit (120 in FIG. 1, 214 in FIG. 2) used in the first reference image synthesis motion compensation prediction process in encoding / decoding will be described with reference to the configuration diagram and FIG. This is performed using the processing flowchart shown in FIG.

The inter-reference image motion vector detection unit (120 in FIG. 1 and 214 in FIG. 2) includes a standard reference image acquisition unit 600, a motion vector detection range setting unit 601, a standard reference image memory 602, a reference image acquisition unit 603, and block matching. The evaluation unit 604 is configured.

First, based on the reference ID and motion vector value mvL0 of the reference reference image input from the reference vector acquisition unit 600 from the motion vector detection unit (117 in FIG. 1) or the motion vector prediction decoding unit (212 in FIG. 2). MvL0 rounded is calculated from the decoded reference image memory (110 in FIG. 1, 209 in FIG. 2), and the image block at the position moved mvL0 rounded from the encoding target block of the standard reference image is cut out, and the first reference block is extracted. Obtain (S800). The acquired reference block is stored in the standard reference image memory 602.

Subsequently, the motion vector detection range setting unit 601 determines the reference ID of the second reference image, and sets a detection range of the inter-reference image motion vector that is detected between the standard reference image and the second reference image ( S801). With respect to the other reference ID for the reference reference image, it is possible to adopt a configuration in which information is transmitted in units similar to the slice header in MPEG4-AVC as information in units of screens. A configuration is adopted in which the reference ID of the second reference image is implicitly determined for each reference ID of the reference image.

Regarding the detection range, it is also possible to take the entire area of the second reference image as the motion vector detection range for the first reference block, and by performing detection processing with the same definition as the encoding device and the decoding device. Although functioning, in order to reduce the amount of calculation in detecting a motion vector between reference images, a detection range as shown in FIG. 7 is set.

FIG. 7 is an example of a motion vector detection range between reference images in the first embodiment. When the input time of the encoding target image is Poc Cur, the input time of the standard reference image is Poc Ref1, and the input time of the second reference image is Poc Ref2, the motion vector mvL0 rounded from the standard reference image for the encoding target block is If the search range of the second reference image is based on the position of the encoding target block, the search center position is expressed as α = mvL0 rounded × (Poc Cur−Poc Ref2) / (Poc Cur−Poc Ref1). In addition, the motion vector prediction value between the encoding target block and the second reference image when it is assumed that the motion is temporally continuous is set.

However, since there are many situations that are not temporally continuous changes, such as camera movements and object movements, an appropriate reference block of the second reference image can be obtained by searching for a motion vector for a specific region with the search position as the center. To be able to get In the example shown in FIG. 7, an area of ± 8 pixels is specified as the specific area.

The reference image acquisition unit 603 obtains the reference block of the second reference image in the motion vector detection range designated by the motion vector detection range setting unit 601 from the decoded reference image memory (110 in FIG. 1, 209 in FIG. 2). Obtain (S802) and output to the block matching evaluation unit 604.

The block matching evaluation unit 604 calculates an error sum for each pixel between the first reference block stored in the standard reference image memory 602 and the reference block of the second reference image input from the reference image acquisition unit 603. Then, a reference block having a smaller total sum and a motion vector value when the reference block is acquired are stored (S803). As a block matching method, evaluation is performed using evaluation values such as SAD and SSE as in the case of a normal motion vector detection unit. However, since a motion vector value derived by detecting a motion vector between reference images is not transmitted during encoding. The evaluation is made without considering the code amount of the motion vector.

After calculating the error sum for each pixel for all the motion vector detection ranges (S804: YES), the stored reference block is set as the second reference block, and the motion vector value is set as the inter-reference image motion vector. The value is mvInterRef (S805).

Regarding the detection accuracy of motion vectors between reference images, it is possible to apply a method of implicitly detecting motion vectors with the same detection accuracy in the encoding device and the decoding device, but detection of motion vectors for each frame or reference image used. It is also possible to use a technique of transmitting accuracy as encoded information. Here, as an implicit setting, the detection accuracy is 1/4 pixel accuracy. The calculated inter-reference image motion vector value mvInterRef is sent to the first reference image synthesis motion compensation prediction unit (121 in FIG. 1, 215 in FIG. 2) together with mvL0, mvL0 rounded, the first reference block, and the second reference block. It is output (S806).

Next, the operation of the second reference image synthesis motion compensated prediction in Embodiment 1 will be described using the flowchart of FIG. As a block of the encoding / decoding device, a reference image synthesis parameter storage unit (122 in FIG. 1, 216 in FIG. 2) and a second reference image synthesis motion compensation prediction unit (123 in FIG. 1, 217 in FIG. 2) are used. The detailed operation will be described.

On the encoding side, first, reference image synthesis parameters for the decoded block above the encoding target block are acquired (S900). The decoded block adjacent to the encoding target block has a relationship as shown in FIG. 10a), and the parameter at the position B is acquired as the upper block. Information as shown in FIG. 10b) is stored as parameters. In the first embodiment, basic prediction mode / reference image synthesis mode / motion vector information (two) and synthesis mode motion vector information are stored.

As the basic prediction mode, a prediction mode indicating the number of motion vectors predicted and transmitted by a prediction motion vector as intra prediction or motion compensated prediction is stored. As the reference image synthesis mode, a flag indicating whether or not the reference image synthesis motion compensation prediction is used is stored. When the reference image synthesis mode is On, unidirectional prediction is set as the basic prediction mode. (To transmit only one motion vector)

As the motion vector information, a reference image ID for specifying the used reference image and a motion vector value are stored, and as the synthesis mode motion vector information, a motion vector generated in the reference image synthesis motion compensation and not transmitted as encoded information is stored. Stores the value. In the first embodiment, since the reference image synthesis motion compensation is configured to use motion vectors for two reference images, one motion vector is stored for each target block as synthesis mode motion vector information. However, when three or more reference images are used, two or more motion vectors are stored.

Next, when the prediction mode of the acquired decoded image synthesis parameter is the intra-frame prediction mode (S901: YES), the second reference image synthesis motion compensation prediction referring to the upper decoded block is invalid, and the prediction block is Not generated. When the prediction mode is not the intra-frame prediction mode (S901: NO), and when the parameter reference image synthesis mode is not On (S902: NO), the motion vector is used as the parameter prediction mode to perform one-way or two-way. A motion compensated prediction block is generated (S903).

If the parameter reference image synthesis mode is On (S902: YES), a bi-directional motion compensated prediction block is generated using the stored motion vector and synthesis mode motion vector information (S904). As a result, a prediction block having the same condition as the result obtained by applying the same process as the multiple composite image motion compensation calculated in the upper part to the encoding target block is generated.

The prediction error evaluation value extended above is calculated between the prediction block generated by these and the encoding target block (S905).

Subsequently, the reference image synthesis parameter for the decoded block at the left of the encoding target block is acquired (S906). In FIG. 10a), the parameter at position A is acquired as the left block.

The prediction block generation using the information on the left part is also performed by the same process. When the prediction mode of the acquired decoded image synthesis parameter is the intra-frame prediction mode (S907: YES), the upper decoded block is referred to. The second reference image synthesis motion compensated prediction is invalid and a prediction block is not generated. When the prediction mode is not the intra-frame prediction mode (S907: NO), and when the parameter reference image synthesis mode is not On (S908: NO), a motion compensated prediction block is generated using a motion vector as the parameter prediction mode. If the parameter reference image synthesis mode is On (S908: YES), a bidirectional motion compensated prediction block is generated using the stored motion vector and synthesis mode motion vector information (S910). .

A prediction error evaluation value extended 生成 left is calculated with respect to the prediction block generated as described above with respect to the encoding target block (S911).

When upper and left prediction blocks are generated, both prediction error evaluation values extended above and extended left are compared, a prediction block with a smaller evaluation value is selected, and output along with information indicating the direction left flag (S912). When only one of the prediction blocks is generated, the prediction block in the generated direction is output. When the prediction block in either direction is not generated, the second decoded image synthesis motion compensated prediction is , Not used in the encoding target block.

Next, processing on the decoding side will be described. On the decoding side, when both the upper and left prediction modes are intra prediction, the second reference image synthesis motion compensation prediction process is not allowed to function, and therefore, when the block is not one of the upper or left intra prediction. When there is one effective block, information indicating this direction is decoded. When there are two effective blocks, left flag is decoded as information indicating the encoded direction, and a reference image in the direction specified thereby. A synthesis parameter is acquired (S920).

If the acquired parameter reference image synthesis mode is not On (S921: NO), a motion compensated prediction block is generated using a motion vector as the parameter prediction mode (S922), and the parameter reference image synthesis mode is On. If there is (S921: YES), a bidirectional motion compensated prediction block is generated using the stored motion vector and synthesized mode motion vector information (S923). The selected prediction block is output (S924), and the reference image synthesis parameter in the selected direction is output to the reference image synthesis parameter storage unit (S925), and the second reference image synthesis motion compensation prediction process for the decoding target block is performed. finish.

In the above flowchart, the selectable adjacent decoded blocks are two points, the upper part and the left part. However, it is also possible to select by adding a direction, and the upper left part (D) or FIG. It is also possible to perform the second reference image synthesis motion compensation prediction by selecting when the upper right part (C) or the like is a decoded block.

Intra prediction, motion compensation prediction, first reference image synthesis motion compensation prediction, and second reference image synthesis motion compensation prediction generated according to the first embodiment take into account additional information necessary for decoding a prediction block in the encoding device. Then, the prediction mode determination unit 115 performs selection based on the code amount and the distortion amount due to encoding. In the first embodiment, information indicating the second reference image synthesis motion compensated prediction is encoded as On / Off bit information, and in the case of On, information indicating the direction in which the reference image synthesis parameter of the decoded block is taken over. The left flag is encoded when there are multiple candidates.

If it is not the second reference image synthesis motion compensation prediction, information indicating the second reference image synthesis motion compensation prediction is encoded as Off, and information indicating intra prediction / unidirectional prediction / bidirectional prediction is encoded as the basic prediction mode. In the case of unidirectional prediction, On / Off information is sent as reference image synthesis mode information. When the reference image synthesis mode information is On, the first reference image synthesis motion compensation prediction is performed, and when it is Off, the conventional motion compensation prediction is performed. In the case of conventional motion compensation prediction, one motion vector is encoded together with information indicating a reference image in the case of unidirectional prediction and two motion vectors in the case of bidirectional prediction. In the case of compensation prediction, only one reference motion vector is encoded together with information indicating a reference image.

FIG. 11 shows a conceptual diagram of the selection results of the first and second reference image synthesis motion compensation predictions generated by the encoding device / decoding device of Embodiment 1, and the effects thereof will be described.

When the object is moving with time, the prediction part that has multiple reference images for the background part has the target prediction area. However, since the motion vector detection between the reference images can be acquired without transmitting an effective motion vector by the first reference image synthesis motion compensation prediction and an effective prediction block can be generated, the first reference image synthesis motion compensation prediction is performed. Selected.

In addition, when the motion is spatially continuous, the parameter generated in the first reference image synthesis motion compensated prediction functions effectively in an adjacent block, and therefore indicates that the parameter is extended. The second reference image synthesized motion compensated prediction that can generate the reference image synthesized motion compensated prediction block is selected by transmitting only the information indicating the direction.

When the prediction from multiple reference images does not function sufficiently due to the overlap between the background and the moving object, etc., normal motion compensation prediction and intra-frame prediction are selected adaptively. A prediction block is configured by the first reference image synthesis motion compensation prediction and the second reference image synthesis motion compensation prediction of the remaining region.

By using the first reference image synthesized motion compensated prediction, a highly efficient prediction block can be generated, and by using the second reference image synthesized motion compensated prediction, a highly efficient predicted block can be extended with less information and decoded. In the apparatus, it is possible to drastically reduce a region in which the inter-reference image motion vector detection necessary for performing the reference image synthesis motion compensation prediction is performed, and it is possible to greatly reduce the amount of calculation.

(Embodiment 2)
Next, a video encoding device and a video decoding device according to Embodiment 2 will be described. FIG. 12 is a configuration diagram showing the configuration of the video encoding device according to the second embodiment, and FIG. 13 is a configuration diagram showing the configuration of the video decoding device according to the second embodiment.

As shown in FIG. 12, the moving picture coding apparatus according to Embodiment 2 includes an input terminal 100, an input picture buffer 101, a block division unit 102, a subtractor 103, an orthogonal transformation unit 104, a quantization unit 105, and an inverse quantization. Unit 106, inverse orthogonal transform unit 107, adder 108, intra-frame decoded image memory 109, decoded reference image memory 110, entropy encoding unit 111, stream buffer 112, output terminal 113, code amount control unit 114, prediction mode determination unit 115, intra-frame prediction unit 116, motion vector detection unit 117, motion compensation prediction unit 118, motion vector prediction unit 119, inter-reference image motion vector detection unit 120, first reference image synthesis motion compensation prediction unit 121, reference image synthesis parameter Storage unit 122, second reference image synthesis motion compensation prediction unit 123, and reference image synthesis selection unit 122 It consists of.

The point that the reference image composition selection unit 1224 is provided and the operation in this processing block are the features in the second embodiment of the present invention, and the same operation as in the first embodiment is performed for the other processing blocks.

Similarly, as illustrated in FIG. 13, the moving picture decoding apparatus according to Embodiment 2 includes an input terminal 200, a stream buffer 201, an entropy decoding unit 202, a prediction mode decoding unit 203, a prediction image selection unit 204, and an inverse quantization unit. 205, inverse orthogonal transform unit 206, adder 207, intra-frame decoded image memory 208, decoded reference image memory 209, output terminal 210, intra-frame prediction unit 211, motion vector prediction decoding unit 212, motion compensation prediction unit 213, reference image The inter-frame motion detection unit 214, the first reference image synthesis motion compensation prediction unit 215, the reference image synthesis parameter storage unit 216, the second reference image synthesis motion compensation prediction unit 217, and the reference image synthesis selection unit 1318 are configured.

The point that the reference image composition selection unit 1318 is provided and the operation in this processing block are the characteristics of the decoding apparatus according to the second embodiment of the present invention. The other processing blocks perform the same operation as in the first embodiment. Is called. Whether the reference image synthesis selection unit (1224, 1318) has the same function in encoding / decoding, and generates a prediction block using a reference image synthesis parameter from an adjacent block without transmission of additional information when paired It has a function of determining whether or not.

Subsequently, FIGS. 14A and 14B show a flowchart of the reference image composition selection process in Embodiment 2, and the operation will be described. On the encoding side, in addition to the operation of the flowchart (S900 to S913) in the first embodiment, the operation of S1414 to S1421 performed by the reference image composition selection unit 1224 is added.

When the reference image synthesis mode is On (S902: YES), the prediction block from the two reference images is generated using the stored motion vector and the synthesis mode motion vector, and the two prediction blocks generated The error value between them is calculated as extended interref error (S1414). As the evaluation value in this case, the sum of errors such as SAD and SSE is used as in the evaluation in the inter-reference image motion vector detection unit.

Subsequently, a prediction block from two reference images is generated for the upper adjacent block, and an error evaluation value interref error between the two reference images is calculated (S1415). With regard to this error evaluation value, it is possible to calculate interref 、 error when generating a prediction block of the upper adjacent block and store it as a reference image synthesis parameter. The processing configuration is calculated when used as

Then, by comparing the extended interref error and the interref error, the characteristic change with the adjacent block is evaluated. In particular,
If extended interref error> interref error + α (threshold) (S1416: YES), it is assumed that there is little continuity with respect to the motion vector between reference images generated in adjacent blocks, and only the stored motion vector value is used. A one-way motion compensation prediction block is generated (S1417).

On the other hand, if the above condition is not satisfied (S1416: NO), it is assumed that the continuity with respect to the inter-reference image motion vector generated in the adjacent block is maintained, and the stored motion is the same as in the first embodiment. A bidirectional motion compensated prediction block is generated using the vector and the composite mode motion vector information (S904). Regarding the threshold value α, for example, when the error evaluation is SAD, a value such that the average error is 4 (4 × the number of pixels of the block) can be taken, but the quantization value at the time of encoding, interref, By switching α according to the magnitude of error, it is possible to more effectively determine continuity with adjacent blocks.

On the encoding side, for the adjacent block on the left side as well, when the reference image synthesis mode is On (S908: YES), two reference images are used using the stored motion vector and synthesis mode motion vector. Each of the prediction blocks is generated, an extended interref error is calculated as an error value between the two generated prediction blocks (S1418), and prediction blocks from two reference images are generated for the adjacent blocks. An error evaluation value interref error between two reference images is calculated (S1419),
If extended interref error> interref error + α (threshold) (S1420: YES), it is assumed that there is little continuity with respect to the inter-reference image motion vector generated in the adjacent block, and only the stored motion vector value is used. Then, a one-way motion compensated prediction block is generated (S1421). Otherwise (S1420: NO), a bidirectional motion compensated prediction block is generated using the stored motion vector and synthesized mode motion vector information ( S910).

On the decoding side, the same determination as that of the first embodiment is added to the reference image composition selection process. On the decoding side, in addition to the operation of the flowchart (S920 to S925) in the first embodiment, the operation of S1426 to S1429 performed by the reference image composition selection unit 1318 is added.

When the reference image synthesis mode is On (S921: YES), the prediction block from the two reference images is generated using the stored motion vector and the synthesis mode motion vector, and the two prediction blocks generated are generated. An extended interref error is calculated as an error value between them (S1426), prediction blocks from two reference images are generated for adjacent blocks, and an error evaluation value interref error between the two reference images is calculated (S1427).
When extended interref error> interref error + α (threshold) (S1428: YES), it is assumed that there is little continuity with respect to the inter-reference image motion vector generated in the adjacent block, and only the stored motion vector value is used. Then, a one-way motion compensated prediction block is generated (S1429). Otherwise (S1428: NO), a bidirectional motion compensated prediction block is generated using the stored motion vector and synthesized mode motion vector information ( S923).

In Embodiment 2, by using the first and second reference image synthesis motion compensated prediction in Embodiment 1, a highly efficient prediction block is encoded with a small amount of information, and the inter-reference image motion vector in the decoding apparatus is used. In addition to the effect of greatly reducing the calculation amount of detection, in the second reference image synthesis motion compensated prediction, an error value between the first predicted image and the second predicted image in the referenced decoded adjacent block, and encoding When the error value of the first prediction image and the second prediction image for the decoding target block is calculated and the error value for the encoding / decoding target block is sufficiently larger than the error value of the referenced decoded block, To generate a motion compensated prediction image composed of only the first prediction image for the encoding / decoding target block. Determines the continuity of the motion compensation prediction image to generate a more accurate motion compensated prediction image without using the additional information it is possible to improve the coding efficiency.

(Embodiment 3)
Next, a video encoding device and a video decoding device according to Embodiment 3 will be described. In the third embodiment, the configuration of the video encoding device and the configuration of the video decoding device are the same as those in the second embodiment, and the operation of the reference image synthesis selection unit (FIG. 12: 1224, FIG. 13: 1318). Are in different forms. FIG. 15A and FIG. 15B show a flowchart of the reference image composition selection process in Embodiment 3, and the operation will be described.

On the encoding side in the third embodiment, the operation of S1522-S1525 is added to the operation of the flowchart (S900-S913, S1414-S1421) in the second embodiment instead of S1417, S1421. Become. On the decoding side, the operations of S1530-S1531 are added to the operations of the flowcharts (S920-S925, S1426-S1429) in the second embodiment instead of S1429.

On the encoding side, for the upper adjacent block,
If extended interref error> interref error + α (threshold) (S1416: YES), the second motion vector is detected between reference images from the first prediction block generated from the motion vector (S1522), and the first A reference image synthesis motion compensation prediction block is generated using the prediction block and the second prediction block (S1523).

These processes are equivalent to the first reference image synthesis motion compensated prediction, but the same process as that on the decoding side is performed on the encoding side, so that the second motion is determined from the identified first motion vector. The vector is calculated, and the processing described in the processing flowchart of FIG.

With regard to the above processing, the reference image synthesis selection unit 1224 may be provided, but the first reference image synthesis motion compensation prediction unit 121 has a function, and by giving an instruction from the reference image synthesis selection unit 1224, It is also possible to calculate a reference image synthesis motion compensation prediction block.

On the encoding side, for the left adjacent block,
If extended interref error> interref error + α (threshold) (S1420: YES), a second motion vector is detected between reference images from the first prediction block generated from the motion vector (S1524), and the first A reference image synthesis motion compensated prediction block is generated using the prediction block and the second prediction block (S1525).

On the decoding side, similar processing is added to the second embodiment, and the adjacent blocks in the extended direction are
If extended interref error> interref error + α (threshold) (S1428: YES), the second motion vector is detected between the reference images from the first prediction block generated from the motion vector (S1530), and the first A reference image synthesis motion compensation prediction block is generated using the prediction block and the second prediction block (S1531).

In Embodiment 3, by using the first and second reference image synthesized motion compensated prediction in Embodiment 1, a highly efficient prediction block is encoded with a small amount of information, and the inter-reference image motion vector in the decoding device In addition to the effect of greatly reducing the calculation amount of detection, in the second reference image synthesis motion compensated prediction, the error value between the first predicted image and the second predicted image in the referenced decoded adjacent block, When the error value of the first prediction image and the second prediction image for the decoding target block is calculated and the error value for the encoding / decoding target block is sufficiently larger than the error value in the referenced decoded block, It is determined that there is no continuity, and motion vector information between the first predicted image and the second reference image for the encoding / decoding target block is calculated. It is necessary to adaptively determine the continuity of the motion-compensated predicted image by generating the third predicted image and generating the motion-compensated predicted image by synthesizing the first predicted image and the third predicted image. Only when this is the case, it is possible to perform the motion compensation prediction of the decoding side motion vector calculation type again, while suppressing an increase in the amount of computation and generating a more appropriate motion compensation prediction image without using additional information, thereby improving the encoding efficiency. Make it possible.

The moving picture encoding apparatus and moving picture decoding apparatus presented as the first, second, and third embodiments are physically a CPU (central processing unit), a recording device such as a memory, and a display device such as a display. , And a computer having a communication means for a transmission path, and the means having each of the presented functions can be realized as a program on the computer and executed. In addition, the program can be provided by being recorded on a computer-readable recording medium, provided from a server through a wired or wireless network, or provided as data broadcasting of terrestrial or satellite digital broadcasting. is there.

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

DESCRIPTION OF SYMBOLS 100 Input terminal 101 Input image buffer 102 Block division part 103 Subtractor 104 Orthogonal transformation part 105 Quantization part 106 Inverse quantization part 107 Inverse orthogonal transformation part 108 Adder 109 In-frame decoded image memory 110 Decoded reference image memory 111 Entropy encoding Unit 112 stream buffer 113 output terminal 114 code amount control unit 115 prediction mode determination unit 116 intra-frame prediction unit 117 motion vector detection unit 118 motion compensation prediction unit 119 motion vector prediction unit 120 inter-picture motion vector detection unit 121 first reference image Synthetic motion compensation prediction unit 122 Reference image synthesis parameter storage unit 123 Second reference image synthesis motion compensation prediction unit 200 Input terminal 201 Stream buffer 202 Entropy decoding unit 203 Prediction mode decoding unit 204 Predicted image selection unit 205 Inverse amount Conversion unit 206 Inverse orthogonal transform unit 207 Adder 208 Intra-frame decoded image memory 209 Decoded reference image memory 210 Output terminal 211 Intra-frame prediction unit 212 Motion vector prediction decoding unit 213 Motion compensation prediction unit 214 Inter-reference image motion vector detection unit 215 1 reference image synthesis motion compensation prediction unit 216 reference image synthesis parameter storage unit 217 second reference image synthesis motion compensation prediction unit 600 standard reference image acquisition unit 601 motion vector detection range setting unit 602 standard reference image memory 603 reference image acquisition unit 604 block Matching evaluation unit 1224 Reference image composition selection unit 1318 Reference image composition selection unit

The present invention can be used for a video signal encoding technique.

Claims

A motion vector detection unit that detects a motion vector from the first reference image with respect to the encoding target block;
A first reference image synthesizing unit that generates a first synthesized reference block obtained by synthesizing a first reference block extracted from the first reference image using the motion vector and a predetermined region of at least one other reference image. When,
A reference image synthesis parameter storage unit that stores information necessary for generating a synthesized reference block calculated by the first reference image synthesis unit;
A second reference block and a predetermined region of the reference image used for the synthesis are identified and synthesized based on information necessary for generating the synthesized reference block of the encoded region stored in the reference image synthesis parameter storage unit. A second reference image synthesis unit for generating two synthesized reference blocks;
Selecting a prediction block for the encoding target block from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block;
A moving picture encoding apparatus comprising: an encoding unit that encodes a prediction difference block obtained by subtracting the prediction block from the encoding target block.
A motion vector detection unit that detects a motion vector from the first reference image with respect to the encoding target block;
A first reference image synthesizing unit that generates a first synthesized reference block obtained by synthesizing a first reference block extracted from the first reference image using the motion vector and a predetermined region of at least one other reference image. When,
A reference image synthesis parameter storage unit that stores information necessary for generating a synthesized reference block calculated by the first reference image synthesis unit;
A second reference block and a predetermined region of the reference image used for the synthesis are identified and synthesized based on information necessary for generating the synthesized reference block of the encoded region stored in the reference image synthesis parameter storage unit. A second reference image synthesis unit for generating two synthesized reference blocks;
A reference image synthesis selection unit configured to input information necessary for generating a synthesized reference block of an encoded region stored in the reference image synthesis parameter storage unit, and to determine a synthesis method for the second reference image synthesis unit; ,
Necessary for the reference image synthesis selection unit to generate a correlation value between a plurality of reference blocks generated using information necessary for generating the synthesized reference block for the encoding target block and the synthesized reference block for the encoded region. A function of comparing correlation values between a plurality of reference blocks generated using various information and selecting an output from the second reference image combining unit between the second combined reference block and the second reference block Have
Selecting a prediction block for the encoding target block from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block or a second reference block;
A moving picture encoding apparatus comprising: an encoding unit that encodes a prediction difference block obtained by subtracting the prediction block from the encoding target block.
A motion vector detection unit that detects a motion vector from the first reference image with respect to the encoding target block;
A first reference image synthesizing unit that generates a first synthesized reference block obtained by synthesizing a first reference block extracted from the first reference image using the motion vector and a predetermined region of at least one other reference image. When,
A reference image synthesis parameter storage unit that stores information necessary for generating a synthesized reference block calculated by the first reference image synthesis unit;
The second synthesized reference block synthesized by specifying the second reference block and the predetermined area of the reference image based on the information necessary for generating the synthesized reference block of the encoded area stored in the reference picture synthesis parameter storage unit A second reference image composition unit for generating
A reference image synthesis selection unit configured to input information necessary for generating a synthesized reference block of an encoded region stored in the reference image synthesis parameter storage unit, and to determine a synthesis method for the second reference image synthesis unit; ,
The second reference image synthesis unit or the first reference image synthesis unit has a function of generating a third synthesized reference block obtained by synthesizing the second reference block and a predetermined region of at least one other reference image;
Necessary for the reference image synthesis selection unit to generate a correlation value between a plurality of reference blocks generated using information necessary for generating the synthesized reference block for the encoding target block and the synthesized reference block for the encoded region. A correlation value between a plurality of reference blocks generated using various information is compared, and the second reference image synthesis unit outputs the second synthesis reference block, or the second reference image synthesis unit or the second reference block A function to select whether to output the third synthesized reference block from one reference image synthesis unit;
Selecting a prediction block for the encoding target block from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block or a third synthesis reference block;
A moving picture encoding apparatus comprising: an encoding unit that encodes a prediction difference block obtained by subtracting the prediction block from the encoding target block.
Detecting a motion vector from the first reference image for the encoding target block;
Generating a first reference block obtained by combining the first reference block extracted from the first reference image using the motion vector and a predetermined region of at least one other reference image;
Storing information necessary for generating the synthesized reference block;
A step of generating a second synthesized reference block obtained by identifying and synthesizing a predetermined area of the reference image used for synthesis with the second reference block from information necessary for generating a synthesized reference block of the stored encoded area Comprising
Selecting a prediction block for the encoding target block from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block;
A video encoding method comprising: encoding a prediction difference block obtained by subtracting the prediction block from the encoding target block.
A function of detecting a motion vector from the first reference image for the encoding target block;
A first reference block extracted from the first reference image using the motion vector and a first synthesized reference block obtained by synthesizing a predetermined region of at least one other reference image;
A function for storing information necessary for generating the synthetic reference block;
A function for generating a second synthesized reference block obtained by identifying and synthesizing a predetermined area of a reference image used for synthesis with the second reference block from information necessary for generating a synthesized reference block of the stored encoded area Comprising
Selecting a prediction block for the encoding target block from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block;
A moving picture encoding program comprising a function of encoding a prediction difference block obtained by subtracting the prediction block from the encoding target block.
A motion vector decoding unit for decoding a motion vector for a decoding target block from the encoded stream;
A first reference image synthesizing unit that generates a first reference block extracted from a first reference image using the motion vector, and a first synthesized reference block obtained by synthesizing a predetermined region of at least one other reference image; ,
A reference image synthesis parameter storage unit that stores information necessary for generating a synthesized reference block calculated by the first reference image synthesis unit;
A second region obtained by specifying and synthesizing a predetermined region of the reference image used for the synthesis with the second reference block based on the information necessary for generating the synthesized reference block of the decoded region stored in the reference image synthesis parameter storage unit; A second reference image synthesis unit for generating a synthesized reference block of
A decoding mode decoding unit that decodes prediction mode selection information selected from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block from a coded stream;
A moving picture decoding apparatus comprising: a decoding unit that generates a decoded image by adding a prediction block selected by the prediction mode selection information and a prediction difference block decoded from the decoding target block.
A motion vector decoding unit for decoding a motion vector for a decoding target block from the encoded stream;
A first reference image synthesizing unit that generates a first reference block extracted from a first reference image using the motion vector, and a first synthesized reference block obtained by synthesizing a predetermined region of at least one other reference image; ,
A reference image synthesis parameter storage unit that stores information necessary for generating a synthesized reference block calculated by the first reference image synthesis unit;
A second region obtained by specifying and synthesizing a predetermined region of the reference image used for the synthesis with the second reference block based on the information necessary for generating the synthesized reference block of the decoded region stored in the reference image synthesis parameter storage unit; A second reference image synthesis unit for generating a synthesized reference block of
A prediction mode decoding unit that decodes prediction mode selection information selected from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block as a decoding target block from an encoded stream;
A reference image synthesis selection unit configured to input information necessary for generating a synthesized reference block of a decoded area stored in the reference image synthesis parameter storage unit, and to determine a synthesis method for the second reference image synthesis unit;
Correlation values between a plurality of reference blocks generated using the information necessary for generating the synthesized reference block for the decoding target block by the reference image synthesis selecting unit, and information necessary for generating the synthesized reference block for the decoded region A function of comparing correlation values between a plurality of reference blocks generated by using and selecting an output from the second reference image synthesis unit between the second synthesized reference block and the second reference block. And
A prediction block selected from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block or a second reference block as a prediction block based on the prediction mode selection information, and decoding from the decoding target block A moving image decoding apparatus comprising: a decoding unit that generates a decoded image by adding the predicted difference blocks.
A motion vector decoding unit for decoding a motion vector for a decoding target block from the encoded stream;
A first reference image synthesizing unit that generates a first reference block extracted from a first reference image using the motion vector, and a first synthesized reference block obtained by synthesizing a predetermined region of at least one other reference image; ,
A reference image synthesis parameter storage unit that stores information necessary for generating a synthesized reference block calculated by the first reference image synthesis unit;
A second region obtained by specifying and synthesizing a predetermined region of the reference image used for the synthesis with the second reference block based on the information necessary for generating the synthesized reference block of the decoded region stored in the reference image synthesis parameter storage unit; A second reference image synthesis unit for generating a synthesized reference block of
A prediction mode decoding unit that decodes prediction mode selection information selected from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block as a decoding target block from an encoded stream;
A reference image synthesis selection unit configured to input information necessary for generating a synthesized reference block of a decoded area stored in the reference image synthesis parameter storage unit, and to determine a synthesis method for the second reference image synthesis unit;
The second reference image synthesis unit or the first reference image synthesis unit has a function of generating a third synthesized reference block obtained by synthesizing the second reference block and a predetermined region of at least one other reference image. ,
Correlation values between a plurality of reference blocks generated using the information necessary for generating the synthesized reference block for the decoding target block by the reference image synthesis selecting unit, and information necessary for generating the synthesized reference block for the decoded region Are used to compare correlation values between a plurality of reference blocks generated by using the second reference image combining unit and output the second combined reference block from the second reference image combining unit or the second reference image combining unit or the first reference A function of selecting whether to output the third synthesis reference block from the image synthesis unit;
From the prediction mode selection information, a prediction block selected from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block or a third synthesis reference block as a prediction block, and the decoding target block A moving picture decoding apparatus comprising: a decoding unit that generates a decoded image by adding the decoded prediction difference blocks.
Decoding a motion vector for a decoding target block from an encoded stream;
Generating a first reference block extracted from a first reference image using the motion vector and a first reference block obtained by combining predetermined regions of at least one other reference image;
Storing information necessary for generating the synthesized reference block;
Generating a second synthesized reference block obtained by identifying and synthesizing a predetermined area of the reference image used for synthesis with the second reference block from information necessary for generating a synthesized reference block of the stored decoded area; ,
Decoding a prediction mode selection information selected from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block as a decoding target block from an encoded stream;
A moving picture decoding method comprising: generating a decoded image by adding a prediction block selected by the prediction mode selection information and a prediction difference block decoded from the decoding target block.
A function of decoding a motion vector for a decoding target block from an encoded stream;
A function of generating a first reference block extracted from a first reference image using the motion vector and a first synthesized reference block obtained by synthesizing a predetermined region of at least one other reference image;
A function for storing information necessary for generating the synthetic reference block;
A function for generating a second synthesized reference block obtained by identifying and synthesizing a predetermined area of the reference image used for synthesis with the second reference block from information necessary for generating a synthesized reference block of the stored decoded area; ,
The decoding target block has a function of decoding prediction mode selection information selected from a plurality of prediction blocks including at least a first synthesis reference block and a second synthesis reference block from an encoded stream,
A moving picture decoding program comprising a function of generating a decoded image by adding a prediction block selected by the prediction mode selection information and a prediction difference block decoded from the decoding target block.