JP2016195441A - Moving-image decoding device, moving-image decoding method, moving-image decoding program, receiving device, receiving method and receiving program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize motion compensation prediction highly efficiently without transmitting an encoded vector.SOLUTION: In generating a candidate list, a connection motion information candidate list generating unit 230 selects a decoded prediction block whose motion information including information on a motion vector and information on a reference image is effective from among a plurality of prediction blocks adjacent to a prediction block to be decoded, and makes the motion information of the selected decoded prediction block as a selection candidate, and does not make as the selection candidate the motion information of a specific prediction block in the plurality of prediction blocks adjacent to the prediction block to be decoded on the basis of a division type of the prediction block to be decoded and a position in the decoded block of the prediction block to be decoded. A code string analysis unit 201 decodes information indicating a position in the candidate list. A connection motion information selecting unit 231 selects motion information used for motion compensation prediction of the prediction block to be decoded from the candidate list on the basis of information indicating a position in the decoded candidate list.SELECTED DRAWING: Figure 23

Description

  The present invention relates to a moving picture decoding technique using motion compensated prediction, and more particularly to a moving picture decoding technique for decoding motion information used in motion compensated prediction.

  In general video compression coding, motion compensation prediction is used. The motion compensated prediction is performed by dividing the target image into fine blocks, using the decoded image as a reference image, and moving from the processing target block of the target image to the reference block of the reference image based on the amount of motion indicated by the motion vector. This is a technique for generating a signal as a prediction signal. There are two types of motion compensated prediction, one for single prediction using one motion vector and the other for bi-prediction using two motion vectors.

  For a motion vector, a motion vector of an encoded block adjacent to the processing target block is set as a prediction motion vector (also simply referred to as “prediction vector”), and a difference between the motion vector of the processing target block and the prediction vector is obtained. The compression efficiency is improved by transmitting the vector as an encoded vector.

  MPEG-4 AVC / H. In moving picture compression coding such as H.264 (hereinafter referred to as MPEG-4 AVC), motion compensation prediction with high accuracy is made possible by making the block size for motion compensation prediction fine and diverse. On the other hand, there is a problem that the code amount of the encoded vector is increased by reducing the block size.

  Therefore, in MPEG-4 AVC, paying attention to the continuity of motion in the time direction, using the motion vector of the block of the reference image at the same position as the processing target block as the motion vector of the processing target block, Temporal direct motion compensated prediction is used to achieve motion compensated prediction without transmission.

  Also, in Patent Document 1, paying attention to the continuity of motion in the spatial direction, the encoded vector is transmitted using the motion vector of the processed block adjacent to the processing target block as the motion vector of the processing target block. A method for realizing motion compensated prediction without the need is disclosed.

JP-A-10-276439

  In the method described in Patent Document 1, when the block size for motion compensation prediction is variable, the minimum block size is used as a reference, and various block sizes are defined in advance as in MPEG-AVC. There is a problem that cannot be applied as it is when applied to such moving image compression coding.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide a moving picture decoding technique capable of realizing motion compensation prediction with high efficiency without transmitting an encoded vector.

  In order to solve the above problems, a video decoding device according to an aspect of the present invention performs motion compensation prediction by dividing a decoded block into one or a plurality of prediction blocks based on a division type, and encodes a video A video decoding device for decoding a stream, wherein a decoded prediction block in which motion information including motion vector information and reference image information is valid from a plurality of prediction blocks adjacent to a prediction block to be decoded A candidate derivation unit (230) that selects and selects motion information of the selected decoded prediction block as a selection candidate; a division type of the prediction block to be decoded; and a position in the decoding block of the prediction block to be decoded And invalidating the motion information of a specific prediction block in a plurality of prediction blocks adjacent to the prediction block to be decoded as the selection candidate (230), a candidate list generation unit (230) that generates a candidate list including the selection candidates, and an addition unit that adds motion information having a predetermined motion vector as a new selection candidate to the candidate list ( 230), a code string analysis unit (201) for decoding information indicating the position in the candidate list, and motion compensated prediction of the prediction block to be decoded from the candidate list based on the decoded information indicating the position. A selection unit (231) that selects motion information to be used.

  Another aspect of the present invention is also a video decoding device. This apparatus is a moving picture decoding apparatus that performs motion compensation prediction by dividing a decoded block into one or a plurality of prediction blocks based on a division type, and for specifying a combined motion information candidate in a combined motion information candidate list. A decoding unit (201) that decodes the index from the code string in which the index is encoded, and motion information of a plurality of neighboring blocks that have been decoded adjacent to the prediction block to be decoded are used for the prediction block to be decoded As a combined motion information candidate to be added to the combined motion information candidate list, a combined motion information candidate list generation unit (230) that generates the combined motion information candidate list, a division type of the prediction block to be decoded, and the A specific block included in the combined motion information candidate list based on a position of the prediction block to be decoded in the decoded block. A combined motion information candidate rearranging unit (162, 168) that selects a combined motion information candidate and changes the position of the specific combined motion information candidate in the combined motion information candidate list, and the combined motion based on the index A combined motion information selection unit (231) that selects one combined motion information candidate from the information candidate list and uses it as motion information of the prediction block to be decoded;

  Yet another aspect of the present invention is a video decoding method. This method is a moving picture decoding method in which a decoded block is divided into one or a plurality of prediction blocks based on a division type, motion compensated prediction is performed, and a coded stream of a moving picture is decoded. Select a decoded prediction block for which motion information including motion vector information and reference image information is valid from a plurality of prediction blocks adjacent to the block, and select motion information of the selected decoded prediction block In a plurality of prediction blocks adjacent to the prediction block to be decoded based on the candidate derivation step as a candidate, the division type of the prediction block to be decoded and the position of the prediction block to be decoded in the decoding block A candidate for generating a candidate list including the selection candidates, and a step of invalidating the motion information of the specific prediction block of the selected prediction block as the selection candidates An add step for adding motion information having a predetermined predetermined motion vector as a new selection candidate to the candidate list, a code string analyzing step for decoding information indicating a position in the candidate list, And a selection step of selecting motion information used for motion compensation prediction of the prediction block to be decoded from the candidate list based on the decoded information indicating the position.

  Yet another embodiment of the present invention is a receiving device. This apparatus is a receiving apparatus that performs decoding processing by performing motion compensation prediction by dividing a decoded block into one or a plurality of prediction blocks in an encoded stream of a received moving image, and the encoded block based on a division type A receiving unit that receives encoded data in which an encoded stream of a moving image that is encoded by performing motion compensation prediction by dividing the image into one or a plurality of prediction blocks, and the received encoded data Motion information including motion vector information and reference image information is valid from a restoration unit that performs packet processing to restore the original encoded stream and a plurality of prediction blocks adjacent to the prediction block to be decoded. A candidate derivation unit (230) that selects motion information of the selected decoded prediction block as a selection candidate, and a prediction target to be decoded. Based on the block division type and the position of the prediction block to be decoded in the decoding block, the motion information of a specific prediction block in a plurality of prediction blocks adjacent to the prediction block to be decoded is selected as the selection candidate. An invalidating unit (230), a candidate list generating unit (230) for generating a candidate list including the selection candidates, and motion information having a predetermined motion vector set in advance as new selection candidates in the candidate list Based on the addition unit (230) to be added, the code string analysis unit (201) for decoding the information indicating the position in the candidate list from the restored encoded stream, and the candidate list based on the information indicating the decoded position To a selection unit (231) that selects motion information used for motion compensation prediction of the decoding target prediction block.

  Yet another embodiment of the present invention is a reception method. This method is a receiving method of performing decoding processing by performing motion compensation prediction by dividing a decoded block into one or a plurality of prediction blocks in an encoded stream of a received moving image, and the encoded block based on the division type Receiving a coded data in which an encoded stream of a moving image encoded by performing motion compensation prediction by dividing the image into one or a plurality of prediction blocks is received, and the received encoded data Motion information including motion vector information and reference image information from a plurality of prediction blocks adjacent to the prediction block to be decoded and a decoding step that recovers the original encoded stream by packet processing has been decoded. A candidate derivation step in which motion information of the selected decoded prediction block is selected as a selection candidate, and the decoding The motion information of a specific prediction block in a plurality of prediction blocks adjacent to the prediction block to be decoded is based on the division type of the prediction block of the elephant and the position of the prediction block to be decoded in the decoding block. An invalidation step that does not select a selection candidate, a candidate list generation step that generates a candidate list that includes the selection candidate, and an addition step that adds motion information having a predetermined motion vector that is determined in advance to the candidate list as a new selection candidate A code sequence analyzing step for decoding information indicating the position in the candidate list from the restored encoded stream, and a motion of the prediction block to be decoded from the candidate list based on the information indicating the decoded position A selection step of selecting motion information used for compensation prediction.

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

  According to the present invention, motion compensated prediction can be realized with high efficiency without transmitting an encoded vector.

FIGS. 1A and 1B are diagrams for explaining an encoded block. FIGS. 2A to 2D are diagrams for explaining the prediction block size type. It is a figure explaining a prediction block size type. It is a figure explaining prediction coding mode. It is a figure explaining the relationship between a merge index and a code sequence. It is a figure explaining an example of the syntax of a prediction block. 1 is a diagram illustrating a configuration of a moving image encoding device according to Embodiment 1. FIG. It is a figure which shows the structure of the motion information generation part of FIG. It is a figure explaining the structure of the merge mode determination part of FIG. It is a figure which shows the adjacent block of the prediction block of a process target in case the prediction block size of a process target is 16 pixels x 16 pixels. It is a figure which shows the block in the prediction block on ColPic in the same position as the prediction block of a process target in case the prediction block size of a process target is 16 pixels x 16 pixels, and its peripheral block. It is a figure explaining the structure of the joint motion information candidate list production | generation part of FIG. 10 is a flowchart illustrating an operation of a combined motion information candidate list generation unit in FIG. 9. It is a flowchart explaining operation | movement of the space joint motion information candidate production | generation part of FIG. It is a flowchart explaining operation | movement of the time combination motion information candidate production | generation part of FIG. It is a flowchart explaining operation | movement of the recessive combined motion information candidate rearrangement part of FIG. FIGS. 17A to 17D are diagrams for explaining recessive combined motion information candidates. It is a flowchart explaining the determination process of a recessive joint motion information candidate. It is a flowchart explaining the operation | movement of the 1st joint motion information candidate supplement part of FIG. It is a flowchart explaining the operation | movement of the 2nd joint motion information candidate supplement part of FIG. 1 is a diagram illustrating a configuration of a video decoding device according to Embodiment 1. FIG. It is a figure which shows the structure of the motion information reproduction | regeneration part of FIG. It is a figure which shows the structure of the joint motion information reproduction | regeneration part of FIG. FIG. 10 is a diagram for explaining the effect of the first embodiment. FIG. 10 is a diagram illustrating a configuration of a combined motion information candidate list generation unit according to the second embodiment. FIG. 10 is a diagram for explaining an operation of a combined motion information candidate list generation unit according to the second embodiment. It is a figure explaining the effect by Embodiment 2. FIG. FIG. 10 is a diagram illustrating a configuration of a combined motion information candidate list generation unit according to a third embodiment. FIG. 10 is a diagram for explaining an operation of a combined motion information candidate list generation unit according to the third embodiment. FIG. 15 is a diagram for explaining the operation of the spatially coupled motion information candidate generation unit according to the third embodiment. It is a figure explaining operation | movement of the recessive combined motion information candidate addition part of FIG. It is a figure explaining the effect by Embodiment 3. FIG. FIG. 10 is a diagram for describing a configuration of a combined motion information candidate list generation unit according to a fourth embodiment. FIG. 20 is a diagram for explaining the operation of the combined motion information candidate list generation unit according to the fourth embodiment. It is a figure explaining operation | movement of the recessive combined motion information candidate deletion part of FIG. It is a figure explaining operation | movement of the recessive combined motion information candidate deletion part in the modification of Embodiment 4. FIG. It is a figure explaining a duplication joint motion information candidate. FIG. 25 is a diagram for describing a configuration of a combined motion information candidate list generation unit according to the fifth embodiment. FIG. 25 is a diagram for explaining the operation of the combined motion information candidate list generation unit according to the fifth embodiment. It is a figure explaining the operation | movement of the dominant joint motion information candidate rearrangement part of FIG. It is a figure explaining a dominant joint motion information candidate. It is a figure explaining the effect by Embodiment 5. FIG. It is a figure explaining the relationship between the frequency | count of combination inspection, combined motion information candidate M, and combined motion information candidate N.

  First, a technique that is a premise of the embodiment of the present invention will be described.

  Currently, apparatuses and systems that comply with an encoding method such as MPEG (Moving Picture Experts Group) are widely used. In such an encoding method, a plurality of images that are continuous on the time axis are handled as digital signal information. At that time, for the purpose of broadcasting, transmitting or storing information with high efficiency, motion compensated prediction using the temporal redundancy by dividing the image into a plurality of blocks and the discrete cosine using the redundancy in the spatial direction Compression encoding is performed using orthogonal transformation such as transformation.

  In 2003, a joint effort between the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC) Joint Technical Committee (ISO / IEC) and the International Telecommunication Union Telecommunication Standardization Sector (ITU-T) AVC / H. An encoding method called H.264 (the ISO / IEC has a standard number of 14496-10 and ITU-T has an H.264 standard number, hereinafter referred to as MPEG-4AVC) has been established as an international standard. In MPEG-4 AVC, basically, a median value of motion vectors of a plurality of adjacent blocks of a processing target block is used as a prediction vector. When the prediction block size is not square and the reference index of a specific adjacent block of the processing target block matches the reference index of the processing target block, the motion vector of the specific adjacent block is set as a prediction vector.

  Coding currently called HEVC by the joint work of the International Technical Organization (ISO) and the International Electrotechnical Commission (IEC) Joint Technical Committee (ISO / IEC) and the International Telecommunication Union Telecommunication Standardization Sector (ITU-T) Standardization of the method is being studied.

  In the standardization of HEVC, a plurality of adjacent blocks and another decoded image block are used as a candidate block group, and one candidate block is selected from these candidate block groups, and information on the selected candidate block is encoded and decoded. A merge mode is being considered.

[Embodiment 1]
(Encoding block)
In the present embodiment, the input image signal is divided into units of maximum encoding blocks, and the divided maximum encoding blocks are processed in a raster scan order. The encoded block has a hierarchical structure, and can be made smaller encoded blocks by sequentially equally dividing into 4 in consideration of the encoding efficiency. Note that the encoded blocks divided into four are encoded in the zigzag scan order. An encoded block that cannot be further reduced is called a minimum encoded block. An encoded block is a unit of encoding, and the maximum encoded block is also an encoded block when the number of divisions is zero. In this embodiment, the maximum coding block is 64 pixels × 64 pixels, and the minimum coding block is 8 pixels × 8 pixels.

  FIGS. 1A and 1B are diagrams for explaining an encoded block. In the example of FIG. 1A, the encoded block is divided into ten. CU0, CU1 and CU9 are 32 × 32 coding blocks, CU2, CU3 and CU8 are 16 × 16 coding blocks, and CU4, CU5, CU6 and CU7 are 8 × 8 coding blocks. It has become. In the example of FIG. 1B, the encoded block is divided into one.

(Prediction block)
In the present embodiment, the encoded block is further divided into prediction blocks. The encoded block is divided into one or more prediction blocks according to a prediction block size type (also referred to as “division type”). FIGS. 2A to 2D are diagrams for explaining the prediction block size type. 2 (a) is 2N × 2N that does not divide the encoded block, FIG. 2 (b) is 2N × N that is horizontally divided, FIG. 2 (c) is N × 2N that is vertically divided, and FIG. d) shows N × N which is divided into 4 parts horizontally and vertically. 2N × 2N is one prediction block 0, 2N × N and N × 2N are two prediction blocks 0 and 1, and N × N is four prediction blocks 0, prediction block 1, prediction block 2, and prediction It consists of block 3.

  FIG. 3 is a diagram for explaining the prediction block size according to the number of divisions of the encoded block and the prediction block size type. The prediction block size in the present embodiment is from 64 pixels × 64 pixels in which the number of CU divisions is 0 and the prediction block size type is 2N × 2N to 3 in the number of CU divisions, and the prediction block size type is N × N. There are 13 predicted block sizes up to 4 pixels x 4 pixels.

  In the present embodiment, the maximum encoding block is 64 pixels × 64 pixels and the minimum encoding block is 8 pixels × 8 pixels, but the present invention is not limited to this combination. In addition, although the prediction block division pattern is shown in FIGS. 2A to 2D, the combination is not limited to this as long as the combination is divided into one or more.

(Predictive coding mode)
In the present embodiment, the motion compensation prediction and the number of encoded vectors can be switched for each block of the prediction block. Here, an example of a predictive coding mode in which motion compensation prediction is associated with the number of coding vectors will be briefly described with reference to FIG. FIG. 4 is a diagram for explaining the predictive coding mode.

  In the predictive coding mode shown in FIG. 4, the prediction direction of motion compensation prediction is single prediction (L0 prediction) and the number of coding vectors is PredL0, and the prediction direction of motion compensation prediction is single prediction (L1 prediction). PredL1 in which the number of encoding vectors is 1, PredBI in which the prediction direction of motion compensation prediction is bi-prediction (BI prediction) and the number of encoding vectors is 2, and the prediction direction in motion compensation prediction is single prediction ( There is a merge mode (MERGE) which is L0 prediction / L1 prediction) or bi-prediction (BI prediction) and the number of encoding vectors is zero. There is also an intra mode (Intra) which is a predictive coding mode in which motion compensation prediction is not performed. Here, PredL0, PredL1, and PredBI are prediction vector modes.

  In the merge mode, the prediction direction is any of L0 prediction / L1 prediction / BI prediction. This is because the prediction direction of the merge mode takes over the prediction direction of the candidate block selected from the candidate block group as it is, or is decoded information. This is because it is derived from In the merge mode, the encoded vector is not encoded. This is because the merge mode encoding vector inherits the motion vector of the candidate block selected from the candidate block group as it is or is derived from the decoded information.

(Reference index)
In this embodiment, in order to improve the accuracy of motion compensation prediction, it is possible to select an optimal reference image from a plurality of reference images in motion compensation prediction. For this reason, the reference image used in the motion compensation prediction is encoded as a reference image index together with the encoded vector. The reference image index used in motion compensation prediction is a numerical value of 0 or more. If the motion compensation prediction is uni-prediction, one reference index is used, and if the motion compensation prediction is bi-prediction, two reference indexes are used (FIG. 4).

  In merge mode, the reference index is not encoded. This is because the reference index of the merge mode inherits the reference index of the candidate block selected from the candidate block group as it is or is derived from the decoded information.

(Reference index list)
In the present embodiment, a plurality of reference images that can be used in motion compensation prediction are registered in the reference index list, and the reference image is confirmed by indicating the reference image registered in the reference index list by the reference index. Used in motion compensated prediction. The reference index list includes a reference index list L0 and a reference index list L1. When the motion compensation prediction is simple prediction, either L0 prediction using a reference image in the reference index list L0 or L1 prediction using a reference image in the reference index list L1 is used. In the case of bi-prediction, BI prediction using the reference index list L0 and the reference index list L1 is used. The maximum number of reference images that can be registered in each reference index list is 16.

(Merge index)
In the present embodiment, in the merge mode, a plurality of adjacent blocks in the processing target image and blocks around the same position as the processing target block in another encoded image are used as candidate block groups. A candidate block having the optimal predictive coding mode, motion vector, and reference index is selected from the block group, and a merge index for indicating the selected candidate block is encoded and decoded. Only one merge index is used in the merge mode (FIG. 4). The maximum number of merge indexes (also referred to as the maximum number of merge candidates) is 5, and the merge index is an integer from 0 to 4. Here, the maximum number of merge indexes is set to 5, but it may be 2 or more, and is not limited to this.

  Hereinafter, the motion information of the candidate block that is the target of the merge index is referred to as a combined motion information candidate, and the aggregate of combined motion information candidates is referred to as a combined motion information candidate list. Hereinafter, the motion information includes a prediction direction, a motion vector, and a reference index.

  The relationship between the merge index and the code string will be described with reference to the drawings. FIG. 5 is a diagram for explaining the relationship between the merge index and the code string. The code string when the merge index is 0 is '0', the code string when the merge index is 1 is '10', the code string when the merge index is 2 is '110', and the code string when the merge index is 3 The code string when the column is “1110” and the merge index is 4 is “11110”, and the code string is set longer as the merge index becomes larger. Therefore, it is possible to improve the encoding efficiency by assigning a small merge index to a candidate block with a high selection rate.

(Predicted vector index)
In the present embodiment, in order to improve the accuracy of the prediction vector, a plurality of adjacent blocks and blocks around the same position as the processing target block of another encoded image are used as candidate block groups, and the candidate block group is used. A candidate block having an optimal motion vector is selected as a prediction vector, and a prediction vector index for indicating the selected candidate block is encoded and decoded. If the motion compensation prediction is uni-prediction, one prediction vector index is used, and if the motion compensation prediction is bi-prediction, two prediction vector indexes are used (FIG. 4). The maximum number of prediction vector indexes (also referred to as the maximum number of prediction vector candidates) is 2, and the prediction vector index is an integer of 0 or 1. Although the maximum number of prediction vector candidates is 2 here, it may be 2 or more, and is not limited to this.

  Hereinafter, a motion vector of a candidate block that is a target of a prediction vector index is referred to as a prediction vector candidate, and a set of prediction vector candidates is referred to as a prediction vector candidate list.

(Syntax)
An example of the syntax of the prediction block according to the present embodiment will be described with reference to FIG. Whether the prediction block is intra or inter is specified by a higher-order encoding block, and FIG. 6 shows only the syntax of the prediction block when the prediction block is inter. In addition, it is assumed that the prediction block size type is also specified in the encoded block.

  The prediction block (PU of FIG. 6) includes a merge flag (merge_flag), a merge index (merge_idx), an inter prediction type (inter_pred_type), an L0 prediction reference index (ref_idx_10), an L0 prediction difference vector (mvd_10 [0]), mvd_l0 [1]), prediction vector index for L0 prediction (mvp_idx_l0), reference index for L1 prediction (ref_idx_l1), difference vector for L1 prediction (mvd_l1 [0], mvd_l1 [1]), and prediction vector index for L1 prediction ( mvp_idx_l1) is installed. [0] of the difference vector indicates a horizontal component and [1] indicates a vertical component.

  Here, inter_pred_type indicates a prediction direction (also referred to as inter prediction type) of motion compensation prediction, and three types of Pred_L0 (uni prediction of L0 prediction), Pred_L1 (uni prediction of L1 prediction) and Pred_BI (bi prediction of BI prediction). There is. When inter_pred_type is Pred_L0 or Pred_BI, information about L0 prediction is installed, and when inter_pred_type is Pred_L1 or Pred_BI, information about L1 prediction is installed.

  In addition, although the syntax of the prediction block by this Embodiment was set like FIG. 6, it is not limited to this.

  Hereinafter, with reference to the drawings, details of a moving image encoding apparatus, a moving image encoding method, a moving image encoding program, a moving image decoding apparatus, a moving image decoding method, and a moving image decoding program according to a preferred embodiment of the present invention will be described. explain. In the description of the drawings, the same elements are denoted by the same reference numerals and redundant description is omitted.

(Configuration of moving picture coding apparatus 100)
FIG. 7 shows the configuration of moving picture coding apparatus 100 according to the first embodiment. The moving image encoding apparatus 100 is an apparatus that encodes a moving image signal in units of prediction blocks for performing motion compensated prediction. The coding block division, the prediction block size type determination, the prediction block size and the position of the prediction block in the coding block (position information of the prediction block), and the determination whether the prediction coding mode is intra are not shown. It is assumed that it is determined by the higher-order coding control unit, and in the first embodiment, a case where the predictive coding mode is not intra will be described. In the first embodiment, a B picture corresponding to bi-prediction will be described, but L1 prediction may be omitted for a P picture not corresponding to bi-prediction.

  The moving image encoding apparatus 100 is realized by hardware such as an information processing apparatus including a CPU (Central Processing Unit), a frame memory, and a hard disk. The moving image encoding apparatus 100 realizes functional components described below by operating the above components. Note that the position information of the prediction block to be processed, the prediction block size, and the prediction direction of motion compensated prediction are assumed to be shared in the video encoding device 100 and are not shown.

  The moving image encoding apparatus 100 according to Embodiment 1 includes a prediction block image acquisition unit 101, a subtraction unit 102, a prediction error encoding unit 103, a code string generation unit 104, a prediction error decoding unit 105, a motion compensation unit 106, and an addition unit. 107, a motion vector detection unit 108, a motion information generation unit 109, a frame memory 110, and a motion information memory 111.

(Function and operation of moving picture coding apparatus 100)
Hereinafter, the function and operation of each unit will be described. The prediction block image acquisition unit 101 acquires the image signal of the prediction block to be processed from the image signal supplied from the terminal 10 based on the position information and the prediction block size of the prediction block, and subtracts the image signal of the prediction block 102, and supplied to the motion vector detection unit 108 and the motion information generation unit 109.

  The motion vector detection unit 108 obtains motion vectors and reference images of the L0 prediction and the L1 prediction from the image signal supplied from the prediction block image acquisition unit 101 and image signals corresponding to a plurality of reference images stored therein. Detect the reference index indicated. The motion vector of the L0 prediction and the L1 prediction and the reference index of the L0 prediction and the L1 prediction are supplied to the motion information generation unit 109. Here, the motion vector detection unit 108 uses image signals corresponding to a plurality of reference images stored therein as reference images, but a reference image stored in the frame memory 110 can also be used.

  A general motion vector detection method calculates an error evaluation value for an image signal of a target image and a prediction signal of a reference image moved by a predetermined movement amount from the same position, and a movement amount that minimizes the error evaluation value. Is a motion vector. When there are a plurality of reference images, a motion vector is detected for each reference image, and a reference image having a minimum error evaluation value is selected. As the error evaluation value, SAD (Sum of Absolute Difference) indicating the sum of absolute differences, MSE (Mean Square Error) indicating the mean square error, or the like can be used. It is also possible to evaluate by adding the motion vector code amount to the error evaluation value.

  The motion information generation unit 109 indicates L0 prediction and L1 prediction motion vectors supplied from the motion vector detection unit 108, reference indexes for L0 prediction and L1 prediction, candidate block groups supplied from the motion information memory 111, and reference indexes. The prediction encoding mode is determined from the reference image in the frame memory 110 and the image signal supplied from the prediction block image acquisition unit 101.

  Based on the determined predictive coding mode, the merge flag, the merge index, the prediction direction of motion compensation prediction, the reference index of L0 prediction and L1 prediction, the difference vector of L0 prediction and L1 prediction, and the prediction vector of L0 prediction and L1 prediction The index is supplied to the code string generation unit 104 as necessary. The motion compensation prediction direction, the reference index of L0 prediction and L1 prediction, and the motion vector of L0 prediction and L1 prediction are supplied to the motion compensation unit 106 and the motion information memory 111. Details of the motion information generation unit 109 will be described later.

  If the prediction direction of the motion compensation prediction supplied from the motion information generation unit 109 is LN prediction, the motion compensation unit 106 stores the frame compensation information in the frame memory 110 indicated by the LN prediction reference index supplied from the motion information generation unit 109. The reference image is motion-compensated based on the LN prediction motion vector supplied from the motion information generation unit 109 to generate a prediction signal for LN prediction. N is 0 or 1. Here, if the prediction direction of motion compensation prediction is bi-prediction, the average value of the prediction signals of L0 prediction and L1 prediction is the prediction signal. Note that the prediction signals of the L0 prediction and the L1 prediction may be weighted. The motion compensation unit 106 supplies the prediction signal to the subtraction unit 102.

  The subtraction unit 102 subtracts the image signal supplied from the prediction block image acquisition unit 101 and the prediction signal supplied from the motion compensation unit 106 to calculate a prediction error signal, and calculates the prediction error signal to the prediction error encoding unit 103. To supply.

  The prediction error encoding unit 103 performs processing such as orthogonal transformation and quantization on the prediction error signal supplied from the subtraction unit 102 to generate prediction error encoded data, and encodes the prediction error encoded data. The data is supplied to the column generation unit 104 and the prediction error decoding unit 105.

  The code string generation unit 104 includes prediction error encoded data supplied from the prediction error encoding unit 103, a merge flag, a merge index, and a motion compensation prediction prediction direction (inter prediction type) supplied from the motion information generation unit 109. , The reference index of L0 prediction and L1 prediction, the difference vector of L0 prediction and L1 prediction, and the prediction vector index of L0 prediction and L1 prediction are entropy-coded according to the syntax sequence shown in FIG. The column is supplied to terminal 11. Entropy coding is performed by a method including variable length coding such as arithmetic coding or Huffman coding.

  The prediction error decoding unit 105 performs a process such as inverse quantization or inverse orthogonal transform on the prediction error encoded data supplied from the prediction error encoding unit 103 to generate a prediction error signal, and the prediction error signal Is supplied to the adder 107. The addition unit 107 adds the prediction error signal supplied from the prediction error decoding unit 105 and the prediction signal supplied from the motion compensation unit 106 to generate a decoded image signal, and supplies the decoded image signal to the frame memory 110. To do.

The frame memory 110 stores the decoded image signal supplied from the adding unit 107. In addition, for a decoded image in which decoding of the entire image is completed, a predetermined number of images of 1 or more is stored as a reference image. The frame memory 110 supplies the stored reference image signal to the motion compensation unit 106 and the motion information generation unit 109. The storage area for storing the reference image is a FIFO (First
In First Out) method.

  The motion information memory 111 stores the motion information supplied from the motion information generation unit 109 for a predetermined number of images in units of the minimum predicted block size. The motion information of the adjacent block of the prediction block to be processed is set as a space candidate block group.

  In addition, the motion information memory 111 sets the motion information of the block on the ColPic located in the same position as the prediction block to be processed and its neighboring blocks as a time candidate block group. The motion information memory 111 supplies the spatial candidate block group and the temporal candidate block group to the motion information generation unit 109 as candidate block groups. The motion information memory 111 is synchronized with the frame memory 110 and is controlled by a FIFO (First In First Out) method.

  Here, ColPic is a decoded image different from the prediction block to be processed, and is stored in the frame memory 110 as a reference image. In Embodiment 1, ColPic is a reference image decoded immediately before the processing target image. In Embodiment 1, ColPic is a reference image decoded immediately before the processing target image. However, it may be a decoded image, for example, the reference image immediately before in the display order or the reference image immediately after in the display order. Alternatively, it can be specified in the encoded stream.

  Here, a method for managing motion information in the motion information memory 111 will be described. The motion information is stored in each memory area in units of the smallest prediction block. Each memory area stores at least a prediction direction, a motion vector for L0 prediction, a reference index for L0 prediction, a motion vector for L1 prediction, and a reference index for L1 prediction.

  When the predictive coding mode is the intra mode, (0, 0) is stored as a motion vector for L0 prediction and L1 prediction, and “−1” is stored as a reference index for L0 prediction and L prediction. Hereinafter, in the motion vector (H, V), H represents a horizontal component and V represents a vertical component. The reference index “−1” may be any value as long as it can be determined that the mode in which motion compensation prediction is not performed. From this point onward, unless expressed otherwise, the term “block” refers to the smallest predicted block unit when expressed simply as a block. Also, in the case of a block outside the area, (0, 0) is stored as a motion vector for L0 prediction and L1 prediction, and “−1” is stored as a reference index for L0 prediction and L1 prediction, as in the intra mode. The When the LX direction (X is 0 or 1) is valid, the reference index in the LX direction is 0 or more, and when the LX direction is invalid (not valid), the reference index in the LX direction is “−1”. That is. Note that “outside area” generally indicates an area outside a picture to which a prediction block to be processed belongs or an area outside a slice.

(Configuration of the motion information generation unit 109)
Next, a detailed configuration of the motion information generation unit 109 will be described. FIG. 8 shows a configuration of the motion information generation unit 109. The motion information generation unit 109 includes a prediction vector mode determination unit 120, a merge mode determination unit 121, and a prediction encoding mode determination unit 122. The terminal 12 is in the motion information memory 111, the terminal 13 is in the motion vector detection unit 108, the terminal 14 is in the frame memory 110, the terminal 15 is in the prediction block image acquisition unit 101, the terminal 16 is in the code string generation unit 104, and the terminal 50 Are connected to the motion compensation unit 106, and the terminal 51 is connected to the motion information memory 111, respectively.

(Function and operation of motion information generation unit 109)
Hereinafter, the function and operation of each unit will be described. The prediction vector mode determination unit 120 includes a candidate block group supplied from the terminal 12, a motion vector for L0 prediction and L1 prediction supplied from the terminal 13, a reference index for L0 prediction and L1 prediction, and a reference index supplied from the terminal 14. The inter-prediction type is determined from the reference image shown in FIG. 5 and the image signal supplied from the terminal 15, and the prediction vector index of the L0 prediction and the L1 prediction is selected according to the inter prediction type, and the difference vector between the L0 prediction and the L1 prediction is selected. , A prediction error is calculated, and a rate distortion evaluation value is calculated. Then, motion information, a difference vector, a prediction vector index, and a rate distortion evaluation value based on the inter prediction type are supplied to the prediction coding mode determination unit 122.

  The merge mode determination unit 121 generates a combined motion information candidate list from the candidate block group supplied from the terminal 12, the reference image supplied from the terminal 14, and the image signal supplied from the terminal 15, and the combined motion information One merged motion information candidate is selected from the candidate list, a merge index is determined, and a rate distortion evaluation value is calculated. Then, the motion information of the combined motion information candidate, the merge index, and the rate distortion evaluation value are supplied to the predictive coding mode determination unit 122. Details of the merge mode determination unit 121 will be described later.

  The prediction coding mode determination unit 122 compares the rate distortion evaluation value supplied from the prediction vector mode determination unit 120 and the rate distortion evaluation value supplied from the merge mode determination unit 121 to determine a merge flag.

  If the predicted vector mode rate distortion evaluation value is less than the merge mode rate distortion evaluation value, the merge flag is set to “0”. The prediction encoding mode determination unit 122 supplies the merge flag, the inter prediction type, the reference index, the difference vector, and the prediction vector index supplied from the prediction vector mode determination unit 120 to the terminal 16, and the prediction vector mode determination unit 120 The supplied motion information is supplied to the terminal 50 and the terminal 51.

  If the merge mode rate distortion evaluation value is less than or equal to the predicted vector mode rate distortion evaluation value, the merge flag is set to “1”. The predictive coding mode determination unit 122 supplies the merge flag and the merge index supplied from the merge mode determination unit 121 to the terminal 16, and supplies the motion information supplied from the merge mode determination unit 121 to the terminal 50 and the terminal 51. To do. Although a specific calculation method of the rate distortion evaluation value is not the main point of the present invention, detailed description is omitted, but the prediction error amount per code amount is calculated from the prediction error and the code amount, and the rate distortion evaluation value is calculated. The evaluation value has a characteristic that the encoding efficiency increases as the value decreases. Therefore, encoding efficiency can be improved by selecting a predictive encoding mode with a small rate distortion evaluation value.

(Configuration of merge mode determination unit 121)
Next, a detailed configuration of the merge mode determination unit 121 will be described. FIG. 9 is a diagram for explaining the configuration of the merge mode determination unit 121. The merge mode determination unit 121 includes a combined motion information candidate list generation unit 140 and a combined motion information selection unit 141. The combined motion information candidate list generation unit 140 is also installed in the video decoding device 200 that decodes the code sequence generated by the video encoding device 100 according to Embodiment 1 in the same manner. The moving image decoding apparatus 200 generates the same combined motion information list.

(Function and operation of merge mode determination unit 121)
Hereinafter, the function and operation of each unit will be described. The combined motion information candidate list generation unit 140 generates a combined motion information candidate list including the maximum number of merge candidate motion candidates from the candidate block group supplied from the terminal 12, and uses the combined motion information candidate list as the combined motion information. This is supplied to the selection unit 141. A detailed configuration of the combined motion information candidate list generation unit 140 will be described later.

  The combined motion information selection unit 141 selects the optimum combined motion information candidate from the combined motion information candidate list supplied from the combined motion information candidate list generation unit 140, and is information indicating the selected combined motion information candidate. A certain merge index is determined and the merge index is supplied to the terminal 17.

  Here, a method for selecting an optimal combined motion information candidate will be described. A prediction error amount is calculated from the reference image supplied from the terminal 14 obtained by motion compensation prediction based on the prediction direction, motion vector, and reference index of the combined motion information candidate, and the image signal supplied from the terminal 15. . A rate distortion evaluation value is calculated from the code amount of the merge index and the prediction error amount, and a combined motion information candidate that minimizes the rate distortion evaluation value is selected as an optimal combined motion information candidate.

(Candidate block group supplied to combined motion information candidate list generation unit 140)
Next, candidate block groups supplied to the combined motion information candidate list generation unit 140 will be described with reference to FIGS. 10 and 11. The candidate block group includes a spatial candidate block group and a temporal candidate block group.

  FIG. 10 shows adjacent blocks of a prediction block to be processed when the prediction block size to be processed is 16 pixels × 16 pixels. In the first embodiment, the space candidate block group is assumed to be five blocks of block A1, block C, block D, block B1, and block E shown in FIG. Here, the spatial candidate block group is five blocks of block A1, block C, block D, block B1, and block E, but the spatial candidate block group is at least one or more processed adjacent to the prediction block to be processed. Any block may be used, and the present invention is not limited to these. For example, all of block A1, block A2, block A3, block A4, block B1, block B2, block B3, block B4, block C, block D, and block E may be spatial candidate blocks.

  Next, the time candidate block group will be described with reference to FIG. FIG. 11 shows a block in a prediction block on ColPic and its peripheral blocks at the same position as the prediction block to be processed when the prediction block size to be processed is 16 pixels × 16 pixels. In the first embodiment, the time candidate block group includes two blocks, block H and block I6 shown in FIG. Here, the time candidate block group is two blocks of block H and block I6 on ColPic, but the time candidate block group is at least one block on a decoded image different from the prediction block to be processed. There is no limitation to these. For example, only the block H may be used. Hereinafter, the block A4 is expressed as a block A, the block B4 is expressed as a block B, the block I6 is expressed as a block I, and the blocks H and I6 are expressed as a time block.

(Configuration of combined motion information candidate list generation unit 140)
Next, a detailed configuration of the combined motion information candidate list generation unit 140 will be described. FIG. 12 is a diagram for explaining the configuration of the combined motion information candidate list generation unit 140. The terminal 19 is connected to the combined motion information selection unit 141. The combined motion information candidate list generating unit 140 includes a spatially combined motion information candidate generating unit 160, a temporally combined motion information candidate generating unit 161, a recessive combined motion information candidate rearranging unit 162, a redundant combined motion information candidate deleting unit 163, and a first combination. A motion information candidate supplementing unit 164 and a second combined motion information candidate supplementing unit 165 are included.

(Function and operation of combined motion information candidate list generation unit 140)
Hereinafter, the function and operation of each unit will be described. FIG. 13 is a flowchart for explaining the operation of the combined motion information candidate list generation unit 140. First, the combined motion information candidate list generation unit 140 initializes a combined motion information candidate list (S100). There are no combined motion information candidates in the initialized combined motion information candidate list.

  Next, the spatially coupled motion information candidate generation unit 160 generates a spatially coupled motion information candidate from the candidate block group supplied from the terminal 12 and adds the spatially coupled motion information candidate to the combined motion information candidate list (S101). And the candidate block group are supplied to the time combination motion information candidate generation unit 161. The detailed operation of the spatially coupled motion information candidate generation unit 160 will be described later.

  Next, the temporal combination motion information candidate generation unit 161 generates temporal combination motion information candidates from the candidate block group supplied from the spatial combination motion information candidate generation unit 160 and is supplied from the spatial combination motion information candidate generation unit 160. The combined motion information candidate list is added to the combined motion information candidate list (S102), and the combined motion information candidate list is supplied to the recessive combined motion information candidate rearranging unit 162. The detailed operation of the time combination motion information candidate generation unit 161 will be described later.

  Next, the recessive combined motion information candidate rearranging unit 162 orders the recessive combined motion information candidates registered in the combined motion information candidate list supplied from the temporal combined motion information candidate generating unit 161 in the combined motion information candidate list. (S103), and supplies the combined motion information candidate list to the redundant combined motion information candidate deletion unit 163. The detailed operation of the recessive combined motion information candidate rearranging unit 162 will be described later. Details of the recessive combined motion information candidates will be described later.

  Next, the redundant combined motion information candidate deletion unit 163 checks the combined motion information candidates registered in the combined motion information candidate list supplied from the recessive combined motion information candidate rearranging unit 162, and has the same motion information. If there are a plurality of combined motion information candidates, one combined motion information candidate is left and other combined motion information candidates are deleted (S104), and the combined motion information candidate list is stored in the first combined motion information candidate supplementing unit 164. Supply. Accordingly, all of the combined motion information candidates registered in the combined motion information candidate list become different combined motion information candidates.

  Next, the first combined motion information candidate supplementing unit 164 generates a first supplemental combined motion information candidate from the combined motion information candidates registered in the combined motion information candidate list supplied from the redundant combined motion information candidate deleting unit 163. Are added to the combined motion information candidate list (S105), and the combined motion information candidate list is supplied to the second combined motion information candidate supplementing unit 165. The detailed operation of the first combined motion information candidate supplement unit 164 will be described later.

  Next, the second combined motion information candidate supplementing unit 165 generates a second supplemental combined motion information candidate that does not depend on the combined motion information candidate list supplied from the first combined motion information candidate supplementing unit 164 to generate the first combined motion information. The information is added to the combined motion information candidate list supplied from the information candidate supplementing unit 164 (S106), and the combined motion information candidate list is supplied to the terminal 19. The detailed operation of the second combined motion information candidate supplement unit 165 will be described later.

(Detailed operation of spatially coupled motion information candidate generation unit 160)
Next, a detailed operation of the spatially coupled motion information candidate generation unit 160 will be described. FIG. 14 is a flowchart for explaining the operation of the spatially coupled motion information candidate generation unit 160. The spatially coupled motion information candidate generating unit 160 repeatedly performs the following processing in the order of block A, block B, block C, block E, and block D, which are candidate blocks included in the spatial candidate block group of the candidate block group (from S110) S114).

  First, it is checked whether the candidate block is valid (S111). That the candidate block is valid means that at least one of the reference indexes of the L0 prediction and the L1 prediction of the candidate block is 0 or more. If the candidate block is valid (Y in S111), the motion information of the candidate block is added to the combined motion information candidate list as a spatially combined motion information candidate (S112). If the candidate block is not valid (N in S111), step S112 and step S113 are skipped and the next candidate block is inspected (S114). Following step S112, it is checked whether the number of spatially combined motion information candidates added to the combined motion information candidate list is the maximum number of spatially combined motion information candidates (S113). Here, the maximum number of spatially coupled motion information candidates is 4. If the number of spatially combined motion information candidates added to the combined motion information candidate list is not the maximum number of spatially combined motion information candidates (N in S113), the next candidate block is examined (S114). If the number of spatially combined motion information candidates added to the combined motion information candidate list is the maximum number of spatially combined motion information candidates (Y in S113), the process ends.

  Here, the processing order is blocked so that block A and block B, which are generally considered to have a long tangent to the processing target block and generally have high correlation with the processing target block, can be preferentially registered in the combined motion information candidate list. Although A, block B, block C, block E, and block D are used, the combined motion information candidates may be registered in the combined motion information candidate list in the order of high correlation, and the present invention is not limited to this. Further, although the maximum number of spatially coupled motion information candidates is four, the maximum number of spatially coupled motion information candidates may be 1 or more and less than or equal to the number of candidate blocks included in the spatial candidate block group, and is not limited to this.

(Detailed operation of time combination motion information candidate generation unit 161)
Subsequently, a detailed operation of the time combination motion information candidate generation unit 161 will be described. FIG. 15 is a flowchart for explaining the operation of the time combination motion information candidate generation unit 161. The following processing is repeated for each prediction direction LX of L0 prediction and L1 prediction (S120 to S127). Here, X is 0 or 1. Further, the following processing is repeated in the order of block H and block I, which are candidate blocks included in the time candidate block group of the candidate block group (S121 to S126).

  The temporal combination motion information candidate generation unit 161 checks whether the LN prediction of the candidate block is valid (S122). Here, N is 0 or 1. Here, N is the same as X. That the LN prediction of the candidate block is valid means that the reference index of the LN prediction of the candidate block is 0 or more. If the LN prediction of the candidate block is valid (Y in S122), the LN prediction motion vector of the candidate block is set as the reference motion vector (S123). If the LN prediction of the candidate block is not valid (N in S122), step S123 to step S126 are skipped and the next candidate block is inspected (S126).

  Subsequent to step S123, a reference image for LX prediction of a temporally combined motion information candidate is determined (S124). Here, the LX prediction reference image of the temporal combination motion information candidate is a reference image that is most frequently used as a reference image for LX prediction of candidate blocks included in the spatial candidate block group. Next, a scaling vector is calculated by scaling the reference motion vector to match the distance between the processing target image and the LX prediction reference image of the temporally combined motion information candidate, and the scaling vector is used for the temporally combined motion information candidate LX prediction. A motion vector is used (S125), and the next prediction direction is processed (S127). Here, the scaling vector calculation formula is the same as the scaling vector calculation formula for temporal direct motion compensated prediction in MPEG-4 AVC. Subsequent to step S127 in which the processing for the L0 prediction and the L1 prediction is completed, it is checked whether at least one of the L0 prediction and the L1 prediction of the temporally coupled motion information candidate is valid (S128). The fact that the LN prediction of the temporally combined motion information candidate is effective means that the reference image for the LN prediction of the temporally combined motion information candidate is determined. If at least one of the L0 prediction and the L1 prediction of the temporally combined motion information candidate is valid (Y in S128), the inter prediction type of the temporally combined motion information candidate is determined, and the temporally combined motion information candidate is combined with the motion. The information is added to the information candidate list (S129). Here, in the determination of the inter prediction type, if only the L0 prediction is valid, the inter prediction type of the temporally combined motion information candidate is Pred_L0, and if only the L1 prediction is valid, the inter prediction type of the temporally combined motion information candidate is Is Pred_L1, and if both the L0 prediction and the L1 prediction are valid, the inter prediction type of the temporally combined motion information candidate is Pred_BI.

  Here, N is the same as X, but N may be different from X and is not limited to this. In addition, the reference image for LX prediction of temporal combination motion information candidates is a reference image that is most frequently used as a reference image for LX prediction of candidate blocks included in the spatial candidate block group, but is not limited thereto, and is not limited to this. The reference image may be determined to be 0. Further, although the inter prediction type is any one of Pred_L0, Pred_L1, and Pred_BI, it may be only Pred_BI and is not limited to this.

(Detailed operation of recessive combined motion information candidate rearranging unit 162)
Next, detailed operation of the recessive combined motion information candidate rearranging unit 162 will be described. FIG. 16 is a flowchart for explaining the operation of the recessive combined motion information candidate rearranging unit 162. First, it is checked whether the candidate block A is a recessive combined motion information candidate (S140). Details of the recessive combined motion information candidate will be described later. If the candidate block A is a recessive combined motion information candidate (Y in S140), the combined motion information candidate corresponding to the candidate block A that is a recessive combined motion information candidate is moved to the tail of the combined motion information candidate list (S141). ), The process is terminated. If the candidate block A is not a recessive combined motion information candidate (N in S140), it is checked whether the candidate block B is a recessive combined motion information candidate (S142). If the candidate block B is a recessive combined motion information candidate (Y in S142), the combined motion information candidate corresponding to the candidate block B that is a recessive combined motion information candidate is moved to the tail of the combined motion information candidate list (S143). . If the candidate block B is not a recessive combined motion information candidate (N in S142), the process ends.

(About recessive combined motion information candidates)
Hereinafter, the recessive combined motion information candidates will be described. FIGS. 17A to 17D are diagrams for explaining a recessive combined motion information candidate. It has been mentioned above that an encoded block can be divided into 1 or 2 or 4 prediction blocks. FIG. 17 shows an example in which the encoded block is 16 × 16. By dividing the coding block into a number of prediction blocks, the prediction error related to the coding block can be minimized, while the syntax overhead related to the prediction block shown in FIG. 6 increases by the number of prediction blocks. Therefore, a plurality of prediction blocks having the same motion information and the same prediction error can be combined into one prediction block, and the coding efficiency can be improved by suppressing the number of divisions of the prediction block.

  FIG. 17A is a diagram for explaining Example 1 in which candidate block A is a recessive combined motion information candidate. The encoded block is the merge mode, and the position of the spatially coupled motion information candidate of the prediction block 1 when the prediction block size type is encoded as N × 2N is shown. Here, if the motion information of the prediction block 0 and the motion information of the prediction block 1 are the same, the prediction block 0 and the prediction block 1 are collectively encoded as 2N × 2N, thereby generating syntax overhead related to the prediction block. And the coding efficiency can be improved. That is, when the prediction block size type is N × 2N and the processing target block is the prediction block 1, the candidate block A corresponding to the candidate block when encoded as 2N × 2N is selected as a combined motion information candidate. Unlikely. Therefore, when the prediction block size type is N × 2N and the processing target block is the prediction block 1, the candidate block A is set as a recessive combined motion information candidate.

  FIG. 17B is a diagram for explaining an example 2 in which the candidate block A is a recessive combined motion information candidate. The encoded block is the merge mode, and the position of the spatially coupled motion information candidate of the prediction block 3 when the prediction block size type is encoded as N × N is shown. If the motion information of the prediction block 0 and the motion information of the prediction block 1 are the same, and the motion information of the prediction block 2 and the motion information of the prediction block 3 are the same, the prediction block 0 and the prediction block 1 are combined. By encoding the prediction block 2 and the prediction block 3 together as 2N × N, it is possible to reduce the syntax overhead related to the prediction block and improve the encoding efficiency. Therefore, when the prediction block size type is N × N and the motion information of the prediction block 0 and the motion information of the prediction block 1 are the same and the prediction block 3, the candidate block A is set as a recessive combined motion information candidate.

  FIG. 17C is a diagram for explaining an example 1 in which the candidate block B is a recessive combined motion information candidate. The coding block is in the merge mode, and the prediction block size type is 2N × N. Here, if the motion information of the prediction block 0 and the motion information of the prediction block 1 are the same, the prediction block 0 and the prediction block 1 are collectively encoded as 2N × 2N, thereby generating syntax overhead related to the prediction block. And the coding efficiency can be improved. Therefore, when the prediction block size type is 2N × N and the processing target block is the prediction block 1, the candidate block B is set as a recessive combined motion information candidate.

  FIG. 17D is a diagram for explaining Example 2 in which candidate block B is a recessive combined motion information candidate. The encoded block is the merge mode, and the position of the spatially coupled motion information candidate of the prediction block 3 when the prediction block size type is encoded as N × N is shown. Here, if the motion information of the prediction block 0 and the motion information of the prediction block 2 are the same, and the motion information of the prediction block 1 and the motion information of the prediction block 3 are the same, the prediction block 0 and the prediction block 2 are combined. By encoding the prediction block 1 and the prediction block 3 together as N × 2N, it is possible to reduce the syntax overhead related to the prediction block and improve the encoding efficiency. Therefore, when the prediction block size type is N × N and the motion information of the prediction block 0 and the motion information of the prediction block 2 are the same and the prediction block 3, the candidate block B is set as a recessive combined motion information candidate.

  As described above, the recessive combined motion information candidate rearrangement unit 162 determines the encoding target prediction block based on the division type of the prediction block to be encoded and the position of the prediction block to be encoded in the encoding block. A combined motion information candidate that is more appropriate in terms of coding efficiency to perform motion compensation prediction with a prediction block having a size larger than the size is selected as a recessive combined motion information candidate. As described above, the recessive combined motion information candidate is a combined motion information candidate that improves the coding efficiency when combined with a larger prediction block size.

  FIG. 18 is a flowchart for explaining the process of determining the recessive combined motion information candidate.

  First, it is checked whether the prediction block size type is 2N × 2N (S150). If the predicted block size type is 2N × 2N (Y in S150), the process ends. If the predicted block size type is not 2N × 2N (N in S150), the predicted block size type is checked (S151). If the predicted block size type is N × 2N (N × 2N in S151), it is checked whether it is a predicted block 1 (S152). If the predicted block size type is 2N × N (2N × N in S151), it is checked whether it is a predicted block 1 (S154). If the predicted block size type is N × N (N × N in S151), it is checked whether it is a predicted block 3 (S156). If the predicted block size type is N × 2N and the predicted block is 1 (Y in S152), block A is set as a recessive combined motion information candidate (S153). If the predicted block size type is N × 2N and the predicted block size is not 1 (N in S152), the process ends. If the predicted block size type is 2N × N and the predicted block is 1 (Y in S154), block B is set as a recessive combined motion information candidate (S155). If the predicted block size type is 2N × N and the predicted block size is not 1 (N in S154), the process ends.

  If the prediction block size type is N × N and the prediction block 3 (Y in S156), it is checked whether the motion information of the prediction block 0 and the prediction block 2 is the same (S157). If the predicted block size type is N × N and it is not predicted block 3 (N in S156), the process is terminated. If the motion information of the prediction block 0 and the prediction block 2 is the same (Y of S157), the block A is set as a recessive combined motion information candidate (S158). If the motion information of the prediction block 0 and the prediction block 2 is not the same (N in S157), it is checked whether the motion information of the prediction block 0 and the prediction block 1 is the same (S159). If the motion information of the prediction block 0 and the prediction block 1 is the same (Y in S159), the block B is set as a recessive combined motion information candidate (S160). If the motion information of the prediction block 0 and the prediction block 1 is not the same (N in S159), the process is terminated.

  FIG. 18 shows an example of the process of determining the recessive combined motion information candidate, and the recessive combined motion information using at least the number of divisions of the encoded block into the prediction block and / or the position of the prediction block in the encoded block. The candidate is not limited to this as long as the candidate can be determined.

  For example, in order to simplify the condition determination when the prediction block type is N × N, the motion block identity determination between the prediction block 0 and the prediction block 1 (step S159) or the prediction block 0 and the prediction block 2 (step S157). It is also possible to omit the motion information identity determination for the prediction block 3 and always use blocks A and B as inferior combined motion information candidates. Furthermore, in prediction block 1, block A can always be an inferior combined motion information candidate, and in prediction block 2, block B can always be an inferior combined motion information candidate.

  Further, as another method for simplifying condition determination when the prediction block type is N × N, when the prediction block type is N × N, the recessive combined motion information candidates are not rearranged. You can also. This controls the rearrangement of the recessive joint motion information candidates based on the number of divisions of the encoded block into the prediction blocks.

  In general, candidate block A and candidate block B are combined motion information candidates having a relatively high selection rate compared to other blocks, and therefore, a merge index with a short code length is assigned. However, when candidate block A and candidate block B meet the condition of the recessive combined motion information candidate as described above, encoding efficiency is higher when candidate block A and candidate block B are combined with a larger prediction block. Since it is good, the possibility of being selected as a combined motion information candidate is low. Therefore, in the recessive combined motion information candidate rearrangement unit 162, the recessive combined motion information candidate is moved to the tail of the combined motion information candidate list, and the candidate block other than the recessive combined motion information candidate having a relatively high selectivity is obtained. Encoding efficiency can be improved by assigning a merge index having a short code length.

(Detailed operation of the first combined motion information candidate supplement unit 164)
Next, a detailed operation of the first combined motion information candidate supplement unit 164 will be described. FIG. 19 is a flowchart for explaining the operation of the first combined motion information candidate supplementing unit 164. First, from the number of combined motion information candidates (NumCandList) registered in the combined motion information candidate list supplied from the redundant combined motion information candidate deletion unit 163 and the maximum number of merge candidates (MaxNumMergeCand), the first supplemental combined motion information candidate MaxNumGenCand, which is the maximum number for generating, is calculated from Equation 1 (S170).
MaxNumGenCand = MaxNumMergeCand-NumCandList; (NumCandList> 0)
MaxNumGenCand = 0; (NumCandList == 0) Equation 1

  Next, it is checked whether MaxNumGenCand is greater than 0 (S171). If MaxNumGenCand is not greater than 0 (N in S171), the process ends. If MaxNumGenCand is greater than 0 (Y in S171), the following processing is performed. First, loopTimes that is the number of combination inspections is determined. loopTimes is set to NumCandList × NumCandList. However, if loopTimes exceeds 8, loopTimes is limited to 8 (S172). Here, loopTimes is an integer from 0 to 7. The following processing is repeated for loopTimes (S172 to S180). A combination of the combined motion information candidate M and the combined motion information candidate N is determined (S173). Here, the relationship between the number of combination inspections, the combined motion information candidate M, and the combined motion information candidate N will be described. FIG. 43 is a diagram for explaining the relationship between the number of combination inspections, the combined motion information candidate M, and the combined motion information candidate N. As shown in FIG. 43, M and N are different values, and are set in order of decreasing total value of M and N. It is checked whether the L0 prediction of the combined motion information candidate M is valid and the L1 prediction of the combined motion information candidate N is valid (S174). If the L0 prediction of the combined motion information candidate M is valid and the L1 prediction of the combined motion information candidate N is valid (N in S174), the L0 prediction reference image of the combined motion information candidate M and the motion vector are combined motion information candidates. It is checked whether or not the reference image of the N L1 prediction is different from the motion vector (S175). If the L0 prediction of the combined motion information candidate M is not valid and the L1 prediction of the combined motion information candidate N is not valid (N in S174), the next combination is processed. If the reference image for L0 prediction of the combined motion information candidate M and the reference image for L1 prediction of the combined motion information candidate N are different (Y in S175), the motion vector of the combined motion information candidate M and the reference image for the L0 prediction are combined motion information. By combining the motion vector of L1 prediction of candidate N and the reference image, a bi-join motion information candidate having an inter prediction type of Pred_BI is generated (S176). Here, as the first supplemental combined motion information candidate, bi-coupled motion information is generated by combining the L0 prediction of a certain combined motion information candidate and the L1 prediction motion information of a different combined motion information candidate. If the reference image for L0 prediction of the combined motion information candidate M and the reference image for L1 prediction of the combined motion information candidate N are the same (N in S175), the next combination is processed. Subsequent to step S176, it is checked whether or not there is a combined motion information candidate in the combined motion information candidate list (S177). If the combined motion information candidate does not exist in the combined motion information candidate list (Y of S177), the combined motion information candidate is added to the combined motion information candidate list (S178). If the double coupled motion information candidate exists in the combined motion information candidate list (N in S177), step S178 is skipped. Subsequent to step S177 or step S178, it is checked whether or not the number of generated double coupled motion information is MaxNumGenCand (S179). If the number of generated double coupled motion information is MaxNumGenCand (Y in S179), the process ends. If the number of generated double coupled motion information is not MaxNumGenCand (N in S179), the next combination is processed.

  Here, the first supplemental combined motion information candidate is set as an L0 prediction motion vector of a certain combined motion information candidate registered in the combined motion information candidate list and the reference image, and the L1 prediction motion vector of another combined motion information candidate. In combination with the reference image, a bi-coupled motion information candidate in which the direction of motion compensation prediction is bidirectional is used, but the present invention is not limited to this. For example, combined motion information in which the direction of motion compensated prediction in which an offset value such as +1 is added to the motion vector of L0 prediction and the motion vector of L1 prediction of a certain combined motion information candidate registered in the combined motion information candidate list is bidirectional. Combined motion information in which the direction of motion compensated prediction obtained by adding an offset value such as +1 to the motion vector of L0 prediction or the motion vector of L1 prediction of a certain combined motion information candidate registered in the candidate, combined motion information candidate list is unidirectional Candidates may be used, or they may be arbitrarily combined.

  Here, the first supplemental combined motion information candidate is a combined motion information candidate when there is a slight difference between the motion information of the combined motion information candidate registered in the combined motion information candidate list and the motion information candidate to be processed. Coding efficiency can be improved by correcting the motion information of the combined motion information candidates registered in the list to generate effective combined motion information candidates.

  As described above, either the L0 prediction or the L1 prediction is the same as the motion information of the processing target block even if the motion information is a recessive combined motion information candidate that is unlikely to be selected as a combined motion information candidate. In some cases, the motion information of the processing target block may be slightly deviated from the motion information of the recessive combined motion information candidate. Therefore, without deleting the motion information of the recessive combined motion information candidate, the motion information of the recessive combined motion information candidate is combined with the motion information of another combined motion information candidate, or the motion information of the recessive combined motion information candidate is corrected. Then, by generating the first supplemental combined motion information candidate having a higher selectivity than the recessive combined motion information candidate, it is possible to increase the encoding efficiency. In particular, since at least two combined motion information candidates are required when using the dual combined motion information candidate, only one combined motion information candidate other than the recessive combined motion information candidate is registered in the combined motion information candidate list. If not, the motion information of the recessive combined motion information candidate is combined with the motion information of another combined motion information candidate without deleting the motion information of the recessive combined motion information candidate, so that the coding efficiency can be improved. Further, the combination of the combined motion information candidate M and the combined motion information candidate N for generating the dual combined motion information candidate is set in order of decreasing the total value of M and N, so that the combination other than the recessive combined motion information candidate A dual-coupled motion information candidate can be generated by combining motion information candidates.

(Detailed operation of second combined motion information candidate supplementing unit 165)
Next, the detailed operation of the second combined motion information candidate supplement unit 165 will be described. FIG. 20 is a flowchart for explaining the operation of the second combined motion information candidate supplementing unit 165. First, from the number of combined motion information candidates (NumCandList) registered in the combined motion information candidate list supplied from the first combined motion information candidate supplementing unit 164 and the maximum number of merge candidates (MaxNumMergeCand), the second supplemental combined motion information is obtained. MaxNumGenCand, which is the maximum number of candidates to be generated, is calculated from Equation 2 (S190).
MaxNumGenCand = MaxNumMergeCand-NumCandList; (NumCandList> 0)
MaxNumGenCand = 2; (NumCandList == 0) Equation 2

  Next, the following process is repeated MaxNumCand times for i (S191 to S195). Here, i is an integer from 0 to MaxNumGenCand-1. The second supplemental combined motion in which the motion vector for L0 prediction is (0,0), the reference index is i, the motion vector for L1 prediction is (0,0), and the inter prediction type for which the reference index is i is Pred_BI. Information candidates are generated (S192). It is checked whether the second supplementary combined motion information candidate exists in the combined motion information candidate list (S193). If the second supplementary combined motion information candidate does not exist in the combined motion information candidate list (Y in S193), the second supplemental combined motion information candidate is added to the combined motion information candidate list (S194). If the second supplementary combined motion information candidate exists in the combined motion information candidate list (N in S193), the next i is processed (S195).

  Here, the second supplementary combined motion information candidate has a motion vector for L0 prediction of (0, 0), a reference index of i, a motion vector of L1 prediction of (0, 0), and a reference index of i. The combined motion information candidate whose inter prediction type is Pred_BI is used. This is because, in a general moving image, the frequency of occurrence of combined motion information candidates in which the motion vector for L0 prediction and the motion vector for L1 prediction are (0, 0) is statistically high. The present invention is not limited to this as long as it is a combined motion information candidate that is statistically frequently used without depending on the motion information of the combined motion information candidate registered in the combined motion information candidate list. For example, the motion vectors of L0 prediction and L1 prediction may be vector values other than (0, 0), respectively, and may be set so that the reference indexes of L0 prediction and L1 prediction are different. In addition, the second supplementary combined motion information candidate can be set to motion information with a high occurrence frequency of an encoded image or a part of an encoded image.

  Here, by setting a combined motion information candidate that does not depend on the combined motion information candidate registered in the combined motion information candidate list as the second supplemental combined motion information candidate, the combined motion information candidate registered in the combined motion information candidate list When the number is zero, it is possible to use the merge mode and improve the encoding efficiency. In addition, when the motion information of the combined motion information candidate registered in the combined motion information candidate list is different from the motion information candidate motion to be processed, new combined motion information that is statistically frequently used in the combined motion information candidate list Encoding efficiency can be improved by adding candidates and expanding the range of options.

(Configuration of moving picture decoding apparatus 200)
Next, the moving picture decoding apparatus according to the first embodiment will be described. FIG. 21 is a diagram showing a configuration of the video decoding device 200 according to Embodiment 1. The video decoding device 200 is a device that generates a playback image by decoding the code string encoded by the video encoding device 100.

  The moving picture decoding apparatus 200 is realized by hardware such as an information processing apparatus including a CPU (Central Processing Unit), a frame memory, and a hard disk. The moving picture decoding apparatus 200 realizes functional components described below by operating the above components. The coding block division, the prediction block size type determination, the prediction block size and the position of the prediction block in the coding block (position information of the prediction block), and the determination whether the prediction coding mode is intra are not shown. It is assumed that it is determined by the upper control unit, and here, a case where the predictive coding mode is not intra will be described. Note that the position information and the prediction block size of the prediction block to be decoded are shared in the video decoding device 200 and are not shown.

  The moving picture decoding apparatus 200 according to Embodiment 1 includes a code string analysis unit 201, a prediction error decoding unit 202, an addition unit 203, a motion information reproduction unit 204, a motion compensation unit 205, a frame memory 206, and a motion information memory 207.

(Operation of the video decoding device 200)
Hereinafter, the function and operation of each unit will be described. The code string analysis unit 201 analyzes the code string supplied from the terminal 30 to generate prediction error encoded data, a merge flag, a merge index, a motion compensation prediction direction (inter prediction type), a reference index, a difference vector, and Entropy decodes the prediction vector index according to the syntax. Entropy decoding is performed by a method including variable length coding such as arithmetic coding or Huffman coding. Then, the prediction error encoded data is supplied to the prediction error decoding unit 202, and the merge flag, the merge index, the inter prediction type, the reference index, the difference vector, and the prediction vector index are supplied to the motion information reproduction unit 204. To do.

  The motion information reproduction unit 204 includes a merge flag, a merge index, an inter prediction type, a reference index, a difference vector, and a prediction vector index supplied from the code string analysis unit 201 and a candidate block group supplied from the motion information memory 207. The motion information is reproduced, and the motion information is supplied to the motion compensation unit 205 and the motion information memory 207. A detailed configuration of the motion information reproducing unit 204 will be described later.

  Based on the motion information supplied from the motion information reproducing unit 204, the motion compensation unit 205 performs motion compensation on the reference image indicated by the reference index in the frame memory 206 based on the motion vector to generate a prediction signal. If the prediction direction is bi-prediction, an average of the prediction signals of the L0 prediction and the L1 prediction is generated as a prediction signal, and the prediction signal is supplied to the adding unit 203.

  The prediction error decoding unit 202 performs a process such as inverse quantization or inverse orthogonal transform on the prediction error encoded data supplied from the code string analysis unit 201 to generate a prediction error signal, and the prediction error signal is It supplies to the addition part 203.

  The adding unit 203 adds the prediction error signal supplied from the prediction error decoding unit 202 and the prediction signal supplied from the motion compensation unit 205 to generate a decoded image signal, and the decoded image signal is stored in the frame memory 206 and Supply to terminal 31.

  The frame memory 206 and the motion information memory 207 have the same functions as the frame memory 110 and the motion information memory 111 of the video encoding device 100. The frame memory 206 stores the decoded image signal supplied from the adding unit 203. The motion information memory 207 stores the motion information supplied from the motion information reproducing unit 204 in units of the minimum predicted block size.

(Detailed configuration of the motion information playback unit 204)
Next, a detailed configuration of the motion information reproducing unit 204 will be described. FIG. 22 shows the configuration of the motion information playback unit 204. The motion information playback unit 204 includes an encoding mode determination unit 210, a motion vector playback unit 211, and a combined motion information playback unit 212. The terminal 32 is connected to the code string analysis unit 201, the terminal 33 is connected to the motion information memory 207, the terminal 34 is connected to the motion compensation unit 205, and the terminal 36 is connected to the motion information memory 207.

(Detailed operation of the motion information playback unit 204)
Hereinafter, the function and operation of each unit will be described. The encoding mode determination unit 210 determines whether the merge flag supplied from the code string analysis unit 201 is “0” or “1”. If the merge flag is “0”, the inter prediction type, reference index, difference vector, and prediction vector index supplied from the code string analysis unit 201 are supplied to the motion vector reproduction unit 211. If the merge flag is “1”, the merge index supplied from the code string analysis unit 201 is supplied to the combined motion information reproduction unit 212.

  The motion vector reproduction unit 211 reproduces a motion vector from the inter prediction type, the reference index, the difference vector, and the prediction vector index supplied from the encoding mode determination unit 210 and the candidate block group supplied from the terminal 33. Motion information is generated and supplied to the terminals 34 and 36.

  The combined motion information reproduction unit 212 reproduces motion information from the merge index supplied from the encoding mode determination unit 210 and the candidate block group supplied from the terminal 33 and supplies the motion information to the terminal 34 and the terminal 36.

(Detailed Configuration of Combined Motion Information Reproducing Unit 212)
Next, a detailed configuration of the combined motion information reproduction unit 212 will be described. FIG. 23 shows the configuration of the combined motion information playback unit 212. The combined motion information reproduction unit 212 includes a combined motion information candidate list generation unit 230 and a combined motion information selection unit 231. The terminal 35 is connected to the encoding mode determination unit 210.

(Detailed operation of the combined motion information reproduction unit 212)
Hereinafter, the function and operation of each unit will be described. The combined motion information candidate list generation unit 230 has the same function as the combined motion information candidate list generation unit 140 of the video encoding device 100, and is the same as the combined motion information candidate list generation unit 140 of the video encoding device 100. The combined motion information candidate list is generated by the operation, and the combined motion information candidate list is supplied to the combined motion information selection unit 231.

  The combined motion information selection unit 231 selects a combined motion information candidate indicated by the merge index supplied from the terminal 35 from the combined motion information candidate list supplied from the combined motion information candidate list generation unit 230 and combines motion The information is determined, and the motion information of the combined motion information is supplied to the terminal 34 and the terminal 36.

(Explanation of effect)
The effects of the moving picture coding apparatus and the moving picture decoding apparatus according to Embodiment 1 of the present invention will be described. FIG. 24 is a diagram for explaining the effect of the first embodiment of the present invention. FIG. 24 shows the effect of rearranging combined motion information candidates when candidate block A is a recessive combined motion information candidate. Here, the merge motion information candidate list includes merge index 0 (candidate block A), merge index 1 (candidate block B), merge index 2 (candidate block C), merge index 3 (candidate block E), merge index 4 Assume that (candidate block T) is registered. Strictly speaking, the merge index selection probability varies before and after the rearrangement, but here, for simplification of description, the description will be made assuming that the merge index selection probability does not vary before and after the rearrangement. The same applies to examples for explaining the subsequent effects. Assume that the selection probabilities of candidate block A, candidate block B, candidate block C, candidate block E, and candidate block T are 2%, 60%, 14%, 12%, and 12%, respectively. In this case, the expected code length of the merge index before rearrangement is 2.6 bits (Formula 3). On the other hand, the expected code length of the merge index after rearrangement is 1.8 bits (Formula 4). Therefore, it can be seen that the expected value of the code length of the merge index after the rearrangement is shortened by 0.8 bits, and the encoding efficiency is improved.
0.02x1 + 0.6x2 + 0.14x3 + 0.12x4 + 0.12x4 = 2.6 Equation 3
0.6x1 + 0.14x2 + 0.12x3 + 0.12x4 + 0.02x4 = 1.8 Equation 4

  As described above, the recessive combined motion information candidate is moved to the tail of the combined motion information candidate list, and a merge index with a short code length is assigned to candidate blocks other than the recessive combined motion information candidate having a relatively high selectivity. Thus, encoding efficiency can be improved.

(Modification 1 of Embodiment 1)
The configuration of the combined motion information candidate list generation unit 140 of Embodiment 1 is as shown in FIG. 12, but the recessive combined motion information candidate rearrangement unit 162 is not connected after the time combined motion information candidate generation unit 161, but is spatially combined. You may install after the motion information candidate production | generation part 160. FIG.

  In this case, in the combined motion information candidate list, the recessive combined motion information candidate cannot be registered after the temporal combined motion information candidate, but the reordering process of the recessive combined motion information candidate is performed by the spatial combined motion information candidate generation unit 160. Therefore, it is possible to facilitate circuit design and software design by increasing the independence of the spatially coupled motion information candidate generating unit 160 and the temporally coupled motion information candidate generating unit 161. .

(Modification 2 of Embodiment 1)
The operation of the recessive combined motion information candidate rearranging unit 162 according to the first embodiment is the same as that shown in FIG. 16, but steps S141 and S143 are changed as follows to obtain the same result as that of the first modification of the first embodiment. It can also be done.

  The combined motion information candidate corresponding to candidate block A, which is a recessive combined motion information candidate, is moved immediately before the time combined motion information candidate in the combined motion information candidate list or to the tail when there is no temporal combined motion information candidate (S141). . The combined motion information candidate corresponding to the candidate block B which is a recessive combined motion information candidate is moved immediately before the temporally combined motion information candidate in the combined motion information candidate list or to the tail when there is no temporally combined motion information candidate (S143). .

(Modification 3 of Embodiment 1)
The configuration of the combined motion information candidate list generation unit 140 according to the first embodiment is as shown in FIG. 12, but the first combined motion information candidate supplement unit 164 and the second combined motion information candidate supplement unit 165 do not have one or both. Also good.

[Embodiment 2]
The second embodiment will be described below. The configuration and operation of the combined motion information candidate list generation unit 140 are different from those of the first embodiment. FIG. 25 is a diagram for explaining the configuration of the combined motion information candidate list generation unit 140 according to the second embodiment. The configuration of the combined motion information candidate list generation unit 140 according to Embodiment 1 in FIG. 12 is that the recessive combined motion information candidate rearrangement unit 162 is not after the temporal combined motion information candidate generation unit 161 but the first combined motion information candidate. The difference is that it is installed after the replenishing unit 164.

  FIG. 26 is a diagram for explaining the operation of the combined motion information candidate list generation unit 140 according to the second embodiment. 13 is different from the operation of the combined motion information candidate list generation unit 140 of Embodiment 1 in that step S103 is installed after step S105, not after step S102.

  Since the first supplemental combined motion information candidate can be generated using a combined motion information candidate having a relatively higher selection probability (high reliability) than the recessive combined motion information candidate, generally, the recessive combined motion information is used. The selection probability is relatively higher than the candidate. Therefore, it is possible to improve the encoding efficiency by assigning a merge index having a code length shorter than that of the recessive combined motion information candidate to the first supplemental combined motion information candidate.

(Explanation of effect)
The effects of the moving picture coding apparatus and moving picture decoding apparatus according to Embodiment 2 of the present invention will be described. FIG. 27 is a diagram for explaining the effect of the second embodiment of the present invention. Here, it is assumed that only candidate block A and candidate block B are valid in the candidate block group, and the motion compensation prediction directions of candidate block A and candidate block B are both bidirectional. Also, it is assumed that candidate block A and candidate block B, two bi-join motion information candidates BD0 and BD1, and one second supplementary joint motion information candidate AD are registered in the joint motion information candidate list. . Merge index 0 (candidate block A), merge index 1 (candidate block B), merge index 2 (bijoin motion information candidate BD0), merge index 3 (bijoin motion information candidate BD1), merge index 4 (second supplementary join) Assume that the selection probabilities of motion information candidates AD) are 4%, 70%, 12%, 12%, and 2%, respectively. Although the calculation formula is omitted, in Embodiment 2, the expected value of the code length of the merge index before the rearrangement is 2.36 bits. On the other hand, the expected code length of the merge index after the rearrangement is 1.54 bits. Therefore, it can be seen that the expected value of the code length of the merge index after the rearrangement is 0.82 bits shorter, and the coding efficiency is improved.

(Modification of Embodiment 2)
The configuration of the combined motion information candidate list generation unit 140 according to the second embodiment is as shown in FIG. 25, but the recessive combined motion information candidate rearrangement unit 162 is not arranged after the first combined motion information candidate supplement unit 164. You may install after the 2 joint motion information candidate supplement part 165. FIG.

  In this case, the second supplemental combined motion information candidate is set using motion information that has been frequently generated in the encoded image or a part of the encoded image, as in the case where the second supplemental combined motion information candidate is set. If the selection probability of the information candidate is relatively higher than the selection probability of the recessive combined motion information candidate, encoding is performed by assigning a merge index having a code length shorter than that of the recessive combined motion information candidate to the second supplemental combined motion information candidate. Efficiency can be improved.

[Embodiment 3]
Hereinafter, the third embodiment will be described. The configuration and operation of the combined motion information candidate list generation unit 140 and the operation of the spatial combined motion information candidate generation unit 160 are different from those of the first embodiment.

  First, the configuration of the combined motion information candidate list generation unit 140 will be described. FIG. 28 is a diagram for explaining the configuration of the combined motion information candidate list generation unit 140 according to the third embodiment. 12 is different from the combined motion information candidate list generation unit 140 of Embodiment 1 in that a recessive combined motion information candidate reordering unit 162 is provided instead of the recessive combined motion information candidate rearranging unit 162.

  Next, the operation of the combined motion information candidate list generation unit 140 will be described. FIG. 29 is a diagram for explaining the operation of the combined motion information candidate list generation unit 140 according to the third embodiment. The operation of the combined motion information candidate list generation unit 140 according to the first embodiment in FIG. 13 is that step S107 is installed instead of step S103. Hereinafter, the difference of the operation of the combined motion information candidate list generation unit 140 of the third embodiment from that of the first embodiment will be described.

  The recessive combined motion information candidate adding unit 166 adds the recessive combined motion information candidate to the combined motion information candidate list supplied from the temporal combined motion information candidate generating unit 161 (S107), and makes the combined motion information candidate list redundant. The combined motion information candidate deletion unit 163 is supplied.

  Next, the operation of the spatially coupled motion information candidate generation unit 160 will be described. FIG. 30 is a diagram for explaining the operation of the spatially coupled motion information candidate generation unit 160 according to the third embodiment. 14 differs from the operation of the spatially coupled motion information candidate generation unit 160 in the first embodiment in FIG. 14 in that step S115 and step S116 are added. Hereinafter, the operation of the spatially coupled motion information candidate generation unit 160 according to the third embodiment will be described with respect to differences from the first embodiment.

  First, the temporary memory in which the recessive combined motion information candidates are stored is initialized. If the candidate block is valid (Y in S111), it is checked whether the candidate block is a recessive combined motion information candidate (S115). Here, the determination process of the recessive combined motion information candidate is illustrated in FIG. 18, but may be a simple process. If the candidate block is a recessive combined motion information candidate (Y in S115), the candidate block is stored in a temporary memory as a recessive combined motion information candidate (S116), and the next candidate block is examined (S114). If the candidate block is not a recessive combined motion information candidate (N in S115), the motion information of the candidate block is added to the combined motion information candidate list as a spatial combined motion information candidate (S112). It is assumed that the temporary memory is shared in the spatially coupled motion information candidate generation unit 160.

  Next, the operation of the recessive combined motion information candidate adding unit 166 will be described. FIG. 31 is a diagram for explaining the operation of the recessive combined motion information candidate adding unit 166. First, it is checked whether the recessive combined motion information candidate is stored in the temporary memory (S200). If recessive combined motion information candidates are stored (Y in S200), it is checked whether the number of spatially combined motion information candidates registered in the combined motion information candidate list is smaller than the maximum number of spatially combined motion information candidates (S201). ). If the recessive combined motion information candidate is not stored (N in S200), the process ends. If the number of spatially combined motion information candidates registered in the combined motion information candidate list is smaller than the maximum number of merge candidates (Y in S201), the recessive combined motion information candidate is added to the end of the combined motion information candidate list. (S202), the process ends. If the number of spatially combined motion information candidates registered in the combined motion information candidate list is not smaller than the maximum number of spatially combined motion information candidates (N in S201), the process ends.

  In Embodiment 3, when the number of spatial candidate blocks is larger than the maximum number of spatial combination motion information candidates and all the spatial candidate blocks are valid, other spatial candidate blocks are used instead of the recessive combined motion information candidates. (Candidate block D in this embodiment) can be used. Therefore, encoding efficiency can be improved by using another spatial candidate block that is not a recessive combined motion information candidate having a relatively high selectivity compared to the recessive combined motion information candidate.

(Explanation of effect)
The effects of the moving picture coding apparatus and moving picture decoding apparatus according to Embodiment 3 of the present invention will be described. FIG. 32 is a diagram for explaining the effect of the third embodiment of the present invention. Here, it is assumed that candidate block A, candidate block B, candidate block C, candidate block E, candidate block D, and candidate block T are all valid, and the same preconditions as those in FIG. 24 describing the effects of the first embodiment are used. . Therefore, the expected value of the code length of the merge index before rearrangement is 2.6 bits. On the other hand, although the calculation formula is omitted, the expected value of the code length of the merge index after processing by the recessive combined motion information candidate adding unit 166 according to Embodiment 3 is 1.9 bits. Therefore, it can be seen that the expected value of the code length of the merge index after the rearrangement is 0.7 bits shorter, and the coding efficiency is improved.

(Modification of Embodiment 3)
Although the configuration of the combined motion information candidate list generation unit 140 of Embodiment 3 is as shown in FIG. 28, the recessive combined motion information candidate addition unit 166 is not arranged after the time combined motion information candidate generation unit 161. 2, it may be installed after the first combined motion information candidate supplementing unit 164 or after the second combined motion information candidate supplementing unit 165.

[Embodiment 4]
Hereinafter, the fourth embodiment will be described. The configuration and operation of the combined motion information candidate list generation unit 140 are different from those of the first embodiment.

  First, the configuration of the combined motion information candidate list generation unit 140 will be described. FIG. 33 is a diagram for explaining the configuration of the combined motion information candidate list generation unit 140 according to the fourth embodiment. 12 differs from the combined motion information candidate list generation unit 140 of Embodiment 1 in that a recessive combined motion information candidate deletion unit 167 is added after the first combined motion information candidate supplement unit 164.

  Next, the operation of the combined motion information candidate list generation unit 140 will be described. FIG. 34 is a diagram for explaining the operation of the combined motion information candidate list generation unit 140 according to the fourth embodiment. The operation of the combined motion information candidate list generation unit 140 of Embodiment 1 in FIG. 13 is that the recessive combined motion information candidate deletion unit 167 deletes the recessive combined motion information candidate from the combined motion information candidate list following step S105. The difference is that step (S108) is installed.

  Next, the operation of the recessive combined motion information candidate deletion unit 167 will be described. FIG. 35 is a diagram for explaining the operation of the recessive combined motion information candidate deletion unit 167. The following processing is repeated for the combined motion information candidates M corresponding to the number (NumCandList) of combined motion information candidates registered in the combined motion information candidate list supplied from the first combined motion information candidate supplementing unit 164 (S210 to S213). . Here, M is an integer from 0 to NumCandList-1. Whether the candidate block M is recessive combined motion information is checked (S211). If the candidate block M is recessive combined motion information (Y in S211), the recessive combined motion information candidate is deleted from the combined motion information candidate list (S212), and the process is terminated. If the candidate block M is not the recessive combined motion information (N in S211), the next combined motion information candidate is processed (S213).

(Explanation of effect)
The effects of the moving picture coding apparatus and moving picture decoding apparatus according to Embodiment 4 of the present invention will be described. As in the case where the second supplementary combined motion information candidate is set using motion information that is frequently generated in the encoded image or a part of the encoded image, the second supplemental combined motion information candidate If the selection probability is relatively higher than the selection probability of the recessive combined motion information candidate, the recessive combined motion information candidate is deleted from the combined motion information candidate list, and the second supplemental combined motion information candidate is added to the combined motion information candidate list. Thus, the combined motion information selection unit 141 can select an optimal combined motion information candidate from the combined motion information candidate list that does not include the recessive combined motion information candidate, and can improve the encoding efficiency. Here, the recessive combined motion information candidate is deleted from the combined motion information candidate list. However, the recessive combined motion information candidate may be treated as invalid and the combined motion information selection unit 141 may not select the recessive combined motion information candidate. . That is, a merge index may be assigned to combined motion information candidates other than the recessive combined motion information candidates.

(Modification of Embodiment 4)
The operation of the recessive combined motion information candidate deletion unit 167 in the fourth embodiment is simply to delete the recessive combined motion information candidate registered in the combined motion information candidate list as shown in FIG. 35, but it is a recessive combined motion information candidate. It can also be determined whether or not to delete the recessive combined motion information candidate registered in the combined motion information candidate list by determining whether or not the combined motion information of the prediction block is a duplicate combined motion information candidate.

  Hereinafter, an extended example of the fourth embodiment with determination of overlapping combined motion information candidates will be described. FIG. 36 is a diagram for explaining the operation of the recessive combined motion information candidate deletion unit 167 in the modification of the fourth embodiment. It differs from FIG. 35 in that step S213 is added. It is checked whether the combined motion information of the prediction block including the recessive combined motion information candidate is a redundant combined motion information candidate (step S213). If the combined motion information candidate of the prediction block including the recessive combined motion information candidate is an overlapping combined motion information candidate (Y in step S213), the recessive combined motion information candidate is deleted from the combined motion information candidate list (S212), and processing is performed. Exit. If the combined motion information of the prediction block including the recessive combined motion information candidate is not a redundant combined motion information candidate (N in step S213), the process ends.

  Next, the overlapping combined motion information candidate will be described. FIG. 37 is a diagram for explaining the overlapping combined motion information candidates. FIG. 37 shows an example in which the encoded block is 16 × 16. FIG. 37A is a diagram for explaining a candidate for overlapping joint motion information when the prediction block size type is N × 2N, and the recessive joint motion of the prediction block 1 when the prediction block size type is N × 2N. A candidate block A, a candidate block B, a candidate block C, a candidate block D, and a candidate block E that are spatial combination motion information candidates of the prediction block 0 that is an information candidate are shown. FIG. 37 (b) shows candidate block A, candidate block B, candidate block C, candidate block D, and candidate block E, which are spatially coupled motion information candidates in the case of prediction block 0 having a prediction block size type of 2N × 2N. Show. In FIG. 37A and FIG. 37B, it can be seen that the positions of candidate block A, candidate block D, and candidate block E overlap. A candidate block that overlaps with a position of a candidate block of a larger predicted block size type in a certain predicted block size type is set as a redundant combined motion information candidate.

  Here, in the prediction block 0 that is the recessive combined motion information candidate of the prediction block 1 when the prediction block size type is N × 2N, when these overlapping combined motion information candidates are selected as combined motion information candidates, the prediction block Since selecting the overlapped joint motion information candidate in 1 is equivalent to the case of selecting the overlapped joint motion information candidate as 2N × 2N, it always involves a decrease in coding efficiency. On the other hand, in the prediction block 0 that is the recessive combined motion information candidate of the prediction block 1 when the prediction block size type is N × 2N, the candidate block B or candidate block C that is not the overlapped combined motion information candidate is selected as the combined motion information candidate In such a case, since 2N × 2N is not selected, there is a possibility that encoding efficiency may be improved by selecting candidate block B or candidate block C in prediction block 1. FIG. 37 (c) is a diagram showing that the overlapped joint motion information candidates in the case of the prediction block 0 having a prediction block size type of 2N × N are candidate block B, candidate block C, and candidate block D. FIG. 37 (d) is a diagram showing that the overlapped joint motion information candidates in the case of the prediction block 2 of which the prediction block size type is N × N are candidate block A, candidate block D, and candidate block E. FIG. 37 (e) is a diagram showing that the overlapped joint motion information candidates for the prediction block 1 of which the prediction block size type is N × N are candidate block B, candidate block C, and candidate block D.

  As described above, by determining whether or not the combined motion information of the prediction block that is a recessive combined motion information candidate is a redundant combined motion information candidate, the recessive combined motion information candidate registered in the combined motion information candidate list is obtained. By determining whether or not to delete, encoding efficiency can be improved.

(Modification 2 of Embodiment 4)
The combined motion information candidate list generation unit 140 of the fourth embodiment is applied to the first embodiment as shown in FIG. 33, but can be similarly applied to the second and third embodiments. When applied to the second embodiment, the recessive combined motion information candidate rearranging unit 162 and the recessive combined motion information candidate deleting unit 167 exist after the first combined motion information candidate supplementing unit 164. Here, even if the recessive combined motion information candidate rearrangement unit 162 rearranges the recessive combined motion information candidate and then the recessive combined motion information candidate deletion unit 167 deletes the recessive combined motion information candidate, the recessive combined motion information candidate deletion Even if the recessive combined motion information candidate rearrangement unit 162 rearranges the recessive combined motion information candidate after deleting the recessive combined motion information candidate in the unit 167, the same effect can be obtained. The rearrangement of the recessive combined motion information candidates may be omitted in the unit 162, and the recessive combined motion information candidate deletion unit 167 may simply delete the recessive combined motion information candidates.

[Embodiment 5]
The fifth embodiment will be described below. The configuration and operation of the combined motion information candidate list generation unit 140 are different from those of the first embodiment.

  First, the configuration of the combined motion information candidate list generation unit 140 will be described. FIG. 38 is a diagram for explaining the configuration of the combined motion information candidate list generation unit 140 according to the fifth embodiment. 12 differs from the configuration of the combined motion information candidate list generation unit 140 of Embodiment 1 in that a dominant combined motion information candidate rearrangement unit 168 is installed instead of the recessive combined motion information candidate rearrangement unit 162. .

  Next, the operation of the combined motion information candidate list generation unit 140 will be described. FIG. 39 is a diagram for explaining the operation of the combined motion information candidate list generation unit 140 according to the fifth embodiment. The operation of the combined motion information candidate list generation unit 140 according to the first embodiment of FIG. 13 is that step S103 is replaced with the next step S109.

  The dominant combined motion information candidate rearranging unit 168 causes the dominant combined motion information candidate registered in the combined motion information candidate list supplied from the temporal combined motion information candidate generating unit 161 to be at the head of the combined motion information candidate list. The order is changed (S109).

  Next, the operation of the dominant combined motion information candidate rearranging unit 168 will be described. FIG. 40 is a diagram for explaining the operation of the dominant combined motion information candidate rearranging unit 168. First, it is checked whether the prediction block size type is 2N × N (S220). If the predicted block size type is 2N × N (Y in S220), it is checked whether the predicted block size is 0 (S221). If the predicted block size type is not 2N × N (N in S220), the process ends. If it is the prediction block 0 (Y of S221), the motion information of the candidate block B is moved to the head of the combined motion information candidate list as a dominant combined motion information candidate. Whether the prediction block is 0 is checked (S221). If it is not the prediction block 0 (N of S221), a process will be complete | finished.

  The effect of moving the dominant combined motion information candidate to the top of the combined motion information candidate list will be described. FIG. 41 is a diagram for explaining a dominant combined motion information candidate. In this example, candidate block B is a dominant combined motion information candidate. FIG. 41 shows the position of a spatially combined motion information candidate block whose encoded block is 16 × 16 and whose prediction block size type is 2N × N. Here, when the prediction block sizes of the candidate block B and the candidate block A are both 4 × 4, the length of the tangent line to the prediction block 0 (the part of the side in contact) is the same between the prediction block A and the prediction block B. . On the other hand, when the candidate block B is a part of the prediction block B ′ having a prediction block size of 16 × 16 and the candidate block A is a part of the prediction block A ′ having a prediction block size of 16 × 16, the prediction block The tangent to 0 (the part of the tangent side) is longer in the prediction block B ′ than in the prediction block A ′, and the correlation between the prediction block 0 and the candidate block B is the correlation between the prediction block 0 and the candidate block A. Higher than sex. In addition, when the prediction block size type is 2N × N and the prediction block is 0, the candidate block B is located outside the coding floc and is therefore not a recessive combined motion information candidate.

  In this way, the dominant combined motion information candidate rearranging unit 168 performs coding based on the prediction block size type (division type) of the prediction block to be encoded and the position of the prediction block to be encoded in the encoding block. The combined motion information candidate of the adjacent block whose side part in contact with the prediction block to be converted is longer than the other candidates is selected as the dominant combined motion information candidate.

  As described above, based on the prediction block size type and the position of the prediction block, the candidate block whose tangent to the prediction block is long is moved to the head of the combined motion information candidate list as a dominant combined motion information candidate, and the code length Encoding efficiency can be improved by assigning a short merge index.

(Explanation of effect)
The effects of the moving picture coding apparatus and moving picture decoding apparatus according to Embodiment 5 of the present invention will be described. FIG. 42 is a diagram for explaining the effect of the fifth embodiment of the present invention. Here, it is assumed that candidate block A, candidate block B, candidate block C, candidate block E, and candidate block T are registered in the combined motion information candidate list. The selection probabilities of merge index 0 (candidate block A), merge index 1 (candidate block B), merge index 2 (candidate block C), merge index 3 (candidate block E), and merge index 4 (candidate block T) are respectively Suppose that they are 25%, 35%, 14%, 12%, and 12%. In this case, although the calculation formula is omitted, the expected code length of the merge index before rearrangement is 2.33 bits. On the other hand, the expected code length of the merge index after rearrangement of the dominant combined motion information candidates according to the fifth embodiment is 2.23 bits. Therefore, it can be seen that the expected value of the code length of the merge index after the rearrangement is shortened by 0.1 bit, and the encoding efficiency is improved.

  The moving image encoded stream output from the moving image encoding apparatus of the embodiment described above has a specific data format so that it can be decoded according to the encoding method used in the embodiment. Therefore, the moving picture decoding apparatus corresponding to the moving picture encoding apparatus can decode the encoded stream of this specific data format.

  When a wired or wireless network is used to exchange an encoded stream between a moving image encoding device and a moving image decoding device, the encoded stream is converted into a data format suitable for the transmission form of the communication path. It may be transmitted. In that case, a video transmission apparatus that converts the encoded stream output from the video encoding apparatus into encoded data in a data format suitable for the transmission form of the communication channel and transmits the encoded data to the network, and receives the encoded data from the network Then, a moving image receiving apparatus that restores the encoded stream and supplies the encoded stream to the moving image decoding apparatus is provided.

  The moving image transmitting apparatus is a memory that buffers the encoded stream output from the moving image encoding apparatus, a packet processing unit that packetizes the encoded stream, and transmission that transmits the packetized encoded data via the network. Part. The moving image receiving apparatus generates a coded stream by packetizing the received data, a receiving unit that receives the packetized coded data via a network, a memory that buffers the received coded data, and packet processing. And a packet processing unit provided to the video decoding device.

  The above processing relating to encoding and decoding can be realized as a transmission, storage, and reception device using hardware, and is stored in a ROM (Read Only Memory), a flash memory, or the like. It can also be realized by firmware or software such as a computer. The firmware program and software program can be recorded on a computer-readable recording medium, provided from a server through a wired or wireless network, or provided as a data broadcast of terrestrial or satellite digital broadcasting Is also possible.

  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 moving image encoder, 101 prediction block image acquisition part, 102 subtraction part, 103 prediction error encoding part, 104 code stream production | generation part, 105 prediction error decoding part, 106 motion compensation part, 107 addition part, 108 motion vector detection , 109 motion information generation unit, 110 frame memory, 111 motion information memory, 120 prediction vector mode determination unit, 121 merge mode determination unit, 122 prediction coding mode determination unit, 130 prediction vector candidate list generation unit, 131 prediction vector determination 140 combined motion information candidate list generation unit, 141 combined motion information selection unit, 160 spatial combined motion information candidate generation unit, 161 time combined motion information candidate generation unit, 162 recessive combined motion information candidate rearrangement unit, 163 redundant combined motion Information candidate deletion 164 first combined motion information candidate supplementing unit, 165 second combined motion information candidate supplementing unit, 166 recessive combined motion information candidate adding unit, 167 recessive combined motion information candidate deleting unit, 168 dominant combined motion information candidate rearranging unit, 200 Moving picture decoding apparatus, 201 code string analyzing unit, 202 prediction error decoding unit, 203 adding unit, 204 motion information reproducing unit, 205 motion compensating unit, 206 frame memory, 207 motion information memory, 210 encoding mode determining unit, 211 motion A vector playback unit, 212 combined motion information playback unit, 230 combined motion information candidate list generation unit, 231 combined motion information selection unit.

Claims (1)

A video decoding device that divides a decoded block into one or a plurality of prediction blocks based on a division type, performs motion compensation prediction, and decodes an encoded stream of the video,
A decoded prediction block in which motion information including motion vector information and reference image information is valid is selected from a plurality of prediction blocks adjacent to the decoding target prediction block, and the selected decoded prediction block A candidate derivation unit that uses motion information as a selection candidate;
Based on the division type of the prediction block to be decoded and the position in the decoding block of the prediction block to be decoded, the motion information of a specific prediction block in a plurality of prediction blocks adjacent to the prediction block to be decoded And an invalidation unit that does not make the selection candidate,
A candidate list generating unit for generating a candidate list including the selection candidates;
An adding unit for adding motion information having a predetermined motion vector determined in advance to the candidate list as a new selection candidate;
A code string analysis unit for decoding information indicating a position in the candidate list;
A moving picture decoding apparatus comprising: a selection unit configured to select motion information used for motion compensation prediction of the decoding target prediction block from the candidate list based on the decoded information indicating the position.

Images