`
`CROSS REFERENCE TO RELATED APPLICATIONS
`
`This application is a U.S. continuation application of PCT International
`
`Patent Application Number PCT/JP2019/008089 filed on January 29, 2019,
`
`claiming the benefit of priority of U.S. Provisional Patent Application Number
`
`62/628822 filed on January 30, 2018, the entire contents of which are hereby
`
`incorporated by reference.
`
`1. Technical Field
`
`BACKGROUND
`
`The present disclosure relates to an encoder that encodes a video
`
`including a plurality of pictures, and related technologies.
`
`2. Description of the Related Art
`
`Conventionally, H.265 has been known as standards for encoding
`
`moving pictures. H.265 is also referred to as High-Efficiency Video Coding
`
`(HEVC)
`
`(H.265 USO/TEC 28008-2 HEVOC)/HEVC (High Efficiency Video
`
`Coding)) (see, for example, Non-patent Literature (NPL) 1).
`
`20
`
`SUMMARY
`
`An encoder according to one aspect of the present disclosure includes
`
`circuitry and memory coupled to the circuitry. The circuitry in operation:
`
`determines whether a shape of a current chroma block to be spht in an image
`
`25
`
`satisfies a first condition: generates one or more second candidates for a block
`
`partitioning method by eliminating one or more predetermined candidates from
`
`a plurality offirst candidates for a block partitioning method whenthe shape of
`
`1
`
`
`
`the current chroma block satisfies the first condition;
`
`selects a block
`
`partitioning method from among the one or more second candidates; and splits
`
`the current chromablock according to the block partitioning method selected.
`
`It should be noted that
`
`these generic or specific aspects may be
`
`implemented using a system, a method, an integrated circuit, a computer
`
`program, or a non-transitory computerreadable recording medium such as a
`
`compact dise read only memory (CD-ROM), and mayalso be implemented by
`
`any combination of systems, methods, integrated circuits, computer programs,
`
`and nomtransitory computer-readable recording media.
`
`BRIEF DESCRIPTION OF DRAWINGS
`
`These and other objects, advantages and features of the disclosure will
`
`become apparent from the following description thereof taken in conjunction
`
`with the accompanying drawings that illustrate a specific embodiment of the
`
`present disclosure.
`
`FIG. 1 is a block diagram illustrating a functional configuration of an
`
`encoder according to Embodiment 1;
`
`FIG.
`
`2
`
`illustrates one example of block splitting according to
`
`Embodiment 1
`
`20
`
`FIG.
`
`8 is a chart
`
`indicating transform basis functions for each
`
`transform type:
`
`FIG. 4A illustrates one example of a filter shape used in ALF;
`
`FIG. 4B illustrates another example of a filter shape used in ALF:
`
`FIG. 4C ulustrates another example of a filter shape used in ALF;
`
`25
`
`FIG. 5A illustrates 67 intra prediction modes used in intra prediction:
`
`FIG. 5B is a flow chart for idustrating an outhne of a prediction image
`
`correction process performed via OBMCprocessing:
`
`2a
`
`
`
`FIG. 5C is a conceptual diagram for
`
`illustrating an outline of a
`
`prediction image correction process performed via OBMC processing:
`
`FIG. 5D illustrates one example of FRUC;}
`
`FIG. 6 is for illustrating pattern matching (bilateral matching) between
`
`two blocks along a motion trajectory;
`
`FIG. 7 is for illustrating pattern matching (template matching) between
`
`a template in the current picture and a block in a reference picture:
`
`FIG. 8 1s for illustrating a model assuming uniform hnear motion
`
`FIG. 9Ais for illustrating deriving a motion vector of each sub-block
`
`based on motion vectors of neighboring blocks:
`
`FIG. 9B is for illustrating an outline of a process for deriving a motion
`
`vector via merge mode:
`
`FIG. 9C is a conceptual diagram for illustrating an outhne of DMVR
`
`processing;
`
`FIG. 9D is for illustrating an outhne of a prediction image generation
`
`method using a himinance correction process performed via LIC processing:
`
`FIG. 10 is a block diagram Ulustrating a functional configuration of a
`
`decoder according to Embodiment 1:
`
`FIG.
`
`11 is a flowchart
`
`illustrating a process of selecting a block
`
`20
`
`partitioning method candidate, which is performed by a block splitter in an
`
`encoder according to a first aspect:
`
`FIG. 12 is a diagram ulustrating examples of a block partitioning
`
`method:
`
`FIG. 13 is a diagram illustrating an example of a syntax tree indicating
`
`25
`
`information on a blockpartitioning method;
`
`FIG. 14 1s a flowchart
`
`illustrating a process of selecting a block
`
`partitioning method candidate, which is performed by a block splitter in an
`
`
`
`encoder according tc concrete example 1 of the first aspect:
`
`FIG. 15 is a chart Ulustrating block partitioning methods and a limiting
`
`condition when splitting is performed, according to concrete example 1 of the
`
`first aspect:
`
`FIG. 16 is a flowchart
`
`illustrating a process of selecting a block
`
`partitioning method candidate whena current block to be split is a horizontally
`
`elongated rectangle, according to concrete example 1 of the first aspect:
`
`FIG. 17 is a flowchart
`
`illustrating a process of selecting a block
`
`partitioning method candidate, which is performed by the block splitter in the
`
`encoder according to the first aspect;
`
`FIG. 18 is a chart ilustrating block partitioning methods and a hmiting
`
`condition when splitting is performed, according to concrete example 2 of the
`
`first aspect;
`
`FIG. 19 is a flowchart
`
`illustrating a process of selecting a block
`
`partitioning method candidate, which is performed by a block splitter in an
`
`encoder according to concrete example 3 of the first aspect:
`
`FIG. 20 is a chart illustrating block partitioning methods and a limiting
`
`condition when splitting is performed, according to concrete example 3 of the
`
`first aspect:
`
`20
`
`FIG. 21 is a flowchart ulustrating a process of selecting a block
`
`partitioning method candidate, which is performed by a block splitter in an
`
`encoder according to concrete example 4 of the first aspect:
`
`FIG. 22 is a chart Wlustrating block partitioning methods and a limiting
`
`condition when splitting is performed, according to concrete example 4 of the
`
`25
`
`first aspect:
`
`FIG. 28 is a flowchart
`
`illustrating a process of selecting a block
`
`partitioning method candidate, which is performed by a block splitter in an
`
`4
`
`
`
`encoder according tc concrete example 5 of the first aspect:
`
`FIG. 24 is a chart Ulustrating block partitioning methods and a limiting
`
`condition when splitting is performed, according to concrete example 5 of the
`
`first aspect:
`
`FIG. 25 is a flowchart
`
`illustrating a process of selecting a block
`
`partitioning method candidate, which is performed by a block splitter in an
`
`encoder according to a second aspect;
`
`FIG. 26 is a flowchart
`
`illustrating a process of selecting a block
`
`partitioning method candidate, which is performed by a block splitter in an
`
`encoder according to concrete example 1 of the second aspect:
`
`FIG. 27 1s a flowchart
`
`illustrating a process of selecting a block
`
`partitioning method candidate, which is performed by a block splitter in an
`
`encoder according to concrete example 2 of the second aspect:
`
`FIG. 28 is a flowchart
`
`illustrating a process of selecting a block
`
`partitioning method candidate, which is performed by a block splitter in an
`
`encoder according to concrete example 3 of the second aspect:
`
`FIG. 29 is a flowchart ilustrating a process of referring to block
`
`partitioning information and performing block partitioning, which is performed
`
`by a decoder according to the second aspect:
`
`20
`
`FIG. 30 is a flowchart
`
`illustrating a process of referring to block
`
`partitioning information and performing block partitioning, which is performe
`
`by a decoder according to concrete example 1 of the second aspect:
`
`FIG. 81 is a flowchart ulustrating a process of referring to block
`
`partitioning information and performing block partitioning, whichis performed
`
`25
`
`by a decoder according to concrete example 2 of the second aspect:
`
`FIG. 32 is a flowchart
`
`illustrating a process of referring to block
`
`partitioning information and performing block partitioning, which is performed
`
`
`
`by a decoder according to concrete example 3 of the second aspect,
`
`FIG. 38 is a block diagram illustrating an example of implementation of
`
`encoder 100;
`
`FIG. 84 is a flowchart
`
`illustrating an example of an operation
`
`performed by encoder 100;
`
`FIG. 35 is a block diagram illustrating an example of implementation of
`
`decoder 200;
`
`FIG. 36 is a flowchart
`
`illustrating an example of an operation
`
`performed by decoder 200;
`
`FIG. 87 Ulustrates an overall configuration of a content providing
`
`system for implementing a content distribution service:
`
`FIG. 38 illustrates one example of an encoding structure in scalable
`
`encoding:
`
`FIG. 39 illustrates one example of an encoding structure in scalable
`
`encoding:
`
`FIG. 40 illustrates an example of a display screen of a web page:
`
`FIG. 41 ulustrates an example ofa display screen of a web page:
`
`FIG. 42 illustrates one example ofa smartphone; and
`
`FIG. 48 is a block diagram illustrating a configuration example of a
`
`20
`
`smartphone.
`
`DETAILED DESCRIPTION OF THE EMBODIMENTS
`
`(Underlying Knowledge Forming Basis of the Present Disclosure!
`
`An encoder that encodes a video including a plurality of pictures
`
`25
`
`includes a block sphtter that partitions each of the pictures into units such as
`
`coding tree units (CTUs) and coding units (CUs)
`
`into which a CTU is
`
`recursively split.
`
`re
`0
`
`
`
`In the process of partitioning a picture into CTUs,
`
`the picture is
`
`partitioned into CTUs, each having a fixed size, which are raster-scanned from
`
`upper left to lower right. The size ofa CTU maybe set te any pixel size among
`
`16x16, 32x32, and G4x64, using any value of 16, 32, and 64 which are multiples
`
`of 16.
`
`In the process of partitioning a CTUinto CUs, the CTUis partitioned
`
`into CUs each having a variable size based on quadtree block splitting that is
`
`recursive, Aquadtree is a tree structure in which each board is split into four
`
`branches. When a CTUis not to be partitioned, the CTU becomes a CU and
`
`the size of the CTU becomes the largest size ofa CU. The size ofa CU maybe
`
`set to any pixel size among 8x8, 16x16, 32x32, and G4x64,
`
`In view of this, an encoder according to one aspect of the present
`
`disclosure inchides circuitry and memory. Using the memory, the circuitry:
`
`determines whether a shape of a current block to be split in an image satisfies a
`
`first condition; generates one or more second candidates for a block partitioning
`
`method by eliminating one or more predetermined candidates from a plurality
`
`of first candidates for a block partitioning method when the shape of the
`
`current block satisfies the first condition: selects a block partitioning method
`
`from arnong the one or more second candidates: and splits the current block
`
`20
`
`according to the block partitioning method selected.
`
`This enables the encoder to generate, under a given condition, new
`
`candidates for a block partitioning method by elminating one or more
`
`candidates from a multzple number of candidates for a block partitioning
`
`method, and to spht a current block into shapes corresponding to shapes
`
`25
`
`obtained by using a block partitioning method selected from among the
`
`generated candidates. Accordingly,
`
`the encoder is capable, under a given
`
`condition, of splitting the current block using the block partitioning method
`
`ry
`$
`
`
`
`efficiently selected.
`
`in addition,
`
`the encoder is capable of prohibiting the
`
`appearance of a block having a shape corresponding to a shape obtained by
`
`using a block partitioning method candidate that has been eliminated.
`
`Therefore, when a coding mode is determined using an optimization method
`
`such as a rate-distortion (R-D) optimization,
`
`the number of variations for
`
`carrying out trial calculations decreases and it is expected that the amount of
`
`processing for encoding decreases while the degradation of coding efficiency is
`
`inhibited. Moreover, with the encoder mtentionally biasing the generation
`
`frequency of information relating to a block partitioning direction, accuracy in
`
`probability estimation in arithmetic coding using a context, such as context
`
`adaptive binary arithmetic coding (CABAC), increases, and the improvement of
`
`coding performance can be expected.
`
`The first condition is, for example, that the shape of the current block is
`
`a rectangle.
`
`With this, when the shape of a current block to be split is a rectangle,
`
`the encoder is capable of generating new candidates for a block partitioning
`
`method by eliminating one or more candidates from a multiple number of
`
`candidates for a block partitioning method, and splitting the current block into
`
`shapes corresponding to shapes obtained by using a block partitioning method
`
`20
`
`selected from among the generated candidates. Accordingly, when the shape
`
`of the current block is a rectangle,
`
`the encoder is capable of splitting the
`
`current block using the block partitioning methodefficiently selected.
`
`The first condition is, for example, that a ratio of a longer side to a
`
`shorter side of the current block is greater than afirst value.
`
`25
`
`With this, when the shape of a current block to be split is a rectangle
`
`more elongated than a predetermined shape, the encoderis capable of splitting
`
`the current block into shapes corresponding to shapes obtained by using a block
`
`8
`
`
`
`partitioning method selected from among candidates for a block partitioning
`
`method generated by eliminating one or more candidates from a multiple
`
`number of candidates for a block partitioning method. Accordingly, when the
`
`shape of
`
`the current block is
`
`a
`
`rectangle more elongated than the
`
`predetermined shape, the encoder is capable of splitting the current block into
`
`the shapes corresponding to the shapes obtained by using the block partitioning
`
`method efficiently selected,
`
`The first value is, for example, 2.
`
`With this, when the shape of a current block to be split is a rectangle
`
`having the longer side that is two times as long as the shorter side, the encoder
`
`is capable of sphtting the current block into shapes corresponding to shapes
`
`obtained by using a block partitioning method selected from among candidates
`
`for a block partitioning method generated by ehminating one or more
`
`candidates from a multiple candidates for a block partitioning method.
`
`Accordingly, when the current block is a rectangle having the longer side that is
`
`two times as long as the shorter side, the encoder is capable of splitting the
`
`current block into the shapes corresponding to the shapes cbtained by using the
`
`block partitioning method efficiently selected.
`
`The first value is, for example, 4.
`
`20
`
`With this, when the shape of a current block to be split is a rectangle
`
`having the longer side that is four times as long as the shorter side, the encoder
`
`is capable of sphiting the current block into shapes corresponding to shapes
`
`obtained by using a block partitioning methodselected from among candidates
`
`for a block partitioning method generated by eliminating one or more
`
`25
`
`candidates from a multiple number of candidates for a block partitioning
`
`method. Accordingly, whenthe current block is a rectangle having the longer
`
`side that is four times as long as the shorter side, the encoder is capable of
`
`9
`
`
`
`splitting the current block into the shapes correspondingto the shapes obtained
`
`by using the block partitioning methodefficiently selected.
`
`The first condition is, for example, that the shape of the current block is
`
`a rectangle and a length of a shorter side of the current block is less than a
`
`Or
`
`second value.
`
`With this, when the shape of a current block to be split is a rectangle
`
`having the shorter side that 1s less than a predetermined value, that is, when
`
`the current block has an elongated shape, the encoderis capable of sphitting the
`
`current block into shapes corresponding to shapes obtaimed by using a block
`
`partitioning method selected from among candidates for a block partitioning
`
`method generated by eliminating one or more candidates from a multiple
`
`number of candidates for a block partitioning method. Accordingly, when the
`
`current block is a rectangle having the shorter side that is less than the
`
`predetermined value, that is, when the current block has an elongated shape,
`
`the encoder
`
`is capable of splitting the current block into the shapes
`
`corresponding to the shapes obtained by using the block partitioning method
`
`efficiently selected.
`
`The second value is, for example, 64 pixels.
`
`With this, when the shape of a current block to be split is a rectangle
`
`20
`
`having the shorter side that is less than 64 pixels, the encoder is capable of
`
`splitting the current block into shapes corresponding to shapes obtained by
`
`using a block partitioning method selected from among candidates for a block
`
`partitioning method generated by eliminating one or more candidates from a
`
`multiple number of candidates for a block partitioning method. Accordingly,
`
`25
`
`when the current block is a rectangle having the shorter side that is less than
`
`64 pixels, the encoderis capable of splitting the current block into the shapes
`
`corresponding to the shapes obtamed by using the block partitioning method
`
`10
`
`
`
`efficiently selected.
`
`The first condition is, for example, that a ratio of a longer side to a
`
`shorter side of a block generated through the splitting of the current block is
`
`greater than a third value.
`
`With this, when a current block is split into rectangles each being more
`
`elongated than a predetermined shape, the encoder is capable of splitting the
`
`current block into shapes corresponding to shapes obtamed by using a block
`
`partitioning method selected from among candidates for a block partitioning
`
`method generated by eliminating one or more candidates from a multiple
`
`number of candidates for a block partitioning method. Accordingly, when the
`
`current block is split into rectangles each being more elongated than the
`
`predetermined shape, the encoder is capable of splitting the current block into
`
`the shapes corresponding to the shapes obtained by using the block partitioning
`
`method efficiently selected.
`
`The third value is, for example, 4.
`
`With this, when a current block is split into rectangles each having the
`
`longer side that is four times as long as the shorter side, the encoder is capable
`
`of splitting the current block into shapes corresponding to shapes obtained by
`
`using a block partitioning method selected from among candidates for a block
`
`20
`
`partitioning method generated by eliminating one or more candidates from a
`
`multiple number of candidates for a block partitioning method. Accordingly,
`
`when the current block is split into rectangles each having the longer side that
`
`is four times as long as the shorter side, the encoder is capable of splitting the
`
`current block into the shapes corresponding to the shapes obtainedby using the
`
`25
`
`The third value is, for example, 8.
`
`Withthis, when a current block is split into rectangles each having the
`
`li
`
`
`
`longer side that is eight times as long as the shorter side, the encoder is capable
`
`of splitting the current block into shapes corresponding to shapes obtained by
`
`using a block partitioning method selected from among candidates for a block
`
`partitioning method generated by eliminating one or more candidates from a
`
`multiple number of candidates for a block partitioning method. Accordingly,
`
`when the current block 1s split into rectangles each having the longer side that
`
`is eight times as long as the shorterside, the encoder is capable of splitting the
`
`current block into the shapes corresponding to the shapes obtained by using the
`
`block partitioning method efficiently selected.
`
`The one or more predetermined candidates include,
`
`for example, a
`
`candidate that splits a block having a shorter side and a longerside in such a
`
`rnanner that a ratio of the longer side to the shorter side further increases.
`
`This enables the encoder to eliminate a block partitioning method
`
`candidate such that a current block is split inte shapes each being more
`
`elongated than before the splitting.
`
`The encoder is therefore capable of
`
`prohibiting the appearance of an extremely elongated block that is expected to
`
`hardly appear in the process of block partitioning. Accordingly, when a coding
`
`mode is determined using an optimization method such as an R-D optimization,
`
`the number of variations for trial calculations decreases and it is expected that
`
`20
`
`the amount of processing for encoding decreases while the degradation of
`
`coding efficiency is inhibited. Moreover, the encoder intentionally biases the
`
`generation frequency of information relating to a block partitioning direction.
`
`Thus, accuracy in probability estimation in arithmetic coding using a context,
`
`such as CABAC, increases, and the improvement of encoding performance can
`
`25
`
`be expected. The encoder is also capable of Imiting the appearance of an
`
`extremely elongated block, and this im turn makes it possible to improve
`
`subjective image quahty.
`
`
`
`The one or more predetermined candidates include,
`
`for example, a
`
`candidate that performs binary sphtting on a block having a shorter side and a
`
`longer side in such a manner that a ratio of the longer side to the shorter side
`
`further increases.
`
`This enables the encoder to eliminate a block partitioning method
`
`candidate such that a current block is spht into two blocks each being more
`
`elongated than before the splitting.
`
`The encoder is therefore capable of
`
`prohibiting the appearance of an extremely elongated block that is predictedto
`
`hardly appearin the process of block partitioning. Accordingly, when a coding
`
`mode is determined using an optimization method such as an R-D optimization,
`
`the number of variations for trial calculations decreases and it is expected that
`
`the amount of processing for encoding decreases while the degradation of
`
`coding efficiency is inhibited. Moreover, the encoder intentionally biases the
`
`generation frequency of information relating to a block partitioning direction.
`
`Thus, accuracy in probability estimation in arithmetic coding using a context,
`
`such as CABAC, increases, and the improvement of encoding performance can
`
`be expected. The encoder is also capable of limiting the appearance of an
`
`extremely elongated block, and this in turn makes it possible to improve
`
`subjective image quality.
`
`20
`
`The one or more predetermined candidates include,
`
`for example, a
`
`candidate that performs ternary splitting on a block having a shorter side anda
`
`longer side in such a manner that a ratio of the longer side to the shorter side
`
`further increases.
`
`This enables the encoder to eliminate a block partitioning method
`
`25
`
`candidate such that a current block is split into three blocks each being more
`
`elongated than before the splitting.
`
`The encoder is therefore capable of
`
`prohibiting the appearance of an extremely elongated block that is predictedto
`
`13
`
`
`
`hardly appear in the process of block partitioning. Accordingly, when a coding
`
`mode is determined using an optimization method such as an R-D optimization,
`
`the numberof variations for trial calculations decreases and it is expected that
`
`the amount of processing for encoding decreases while the degradation of
`
`coding efficiency is inhibited. Moreover, the encoder intentionally biases the
`
`generation frequency of information relating to a block partitioning direction.
`
`Thus, accuracy in probability estimation in arithmetic coding using a context,
`
`such as CABAC, increases, and the improvement of encoding performance can
`
`be expected. The encoder is also capable of Himiting the appearance of an
`
`extremely elongated block, and this in turn makes it possible to improve
`
`subjective image quality.
`
`For example, when the shape of the current block does not satisfy a
`
`second condition, the circuitry encodes block partitioning information relating
`
`to the block partitioning method according to which the current block is split,
`
`and when the shape of the current block satisfies the second condition, the
`
`circuitry skips the encoding ofthe block partitioning information.
`
`This enables the encoder to reduce an encoding load by skipping the
`
`process of encoding block partitioning information and writing a bitstream into
`
`syntax as required by the encoding. Accordingly, the encoder is capable of
`
`20
`
`improving coding efficiency.
`
`The block partitioning information includes, for example, at least one of
`
`a mumber into which the current block is spht or a direction in which the
`
`current block is sphit.
`
`This enables the encoder to include, into block partitioning information,
`
`25
`
`information for uniquely determining a block partition shape.
`
`The second conditionis, for example, that the block partitioning method
`
`involves a partitioning direction and the shape of the current block is a
`
`14
`
`
`
`rectangle.
`
`This enables the encoder to skip the encoding of block partitioning
`
`information when the block partitioning method according to which a current
`
`block is split involves a direction and the shape of the current block is a
`
`rectangle. Accordingly, the encoder is capable of improving coding efficiency.
`
`The second condition is, for example, that the block partitioning method
`
`is binarysplitting and the shape of the current block is a rectangle.
`
`This enables the encoder to skip the encoding of block partitioning
`
`information when the block partitioning method according to which a current
`
`block is split is binary splitting and the shape of the current block is a rectangle.
`
`Accordingly, the encoderis capable of improving coding efficiency.
`
`The second condition is, for example, that the block partitioning method
`
`is binary splitting and a ratio of a longer side to a shorter side of the current
`
`block is greater than a predetermined value.
`
`This enables the encoder to skip the encoding of block partitioning
`
`information when the block partitioning method according to which a current
`
`block is split is binary splitting and aratio of the longer side to the shorter side
`
`of the current block is greater than a predetermined value. Accordingly, the
`
`encoder is capable of improving coding efficiency.
`
`20
`
`The second condition is, for example, that the block partitioning method
`
`is ternary splitting and the shape of the current block is a rectangle.
`
`This enables the encoder to skip the encoding of block partitioning
`
`information when the block partitioning method according to which a current
`
`block is split is ternary splitting and the shape of the current block is a
`
`25
`
`rectangle. Accordingly, the encoder is capable of improving coding efficiency.
`
`The second conditionis, for example, that the block partitioning method
`
`is ternary splitting and a ratio of a longer side to a shorter side of the current
`
`15
`
`
`
`block is greater than a predetermined value.
`
`This enables the encoder to skip the encoding of block partitioning
`
`information when the block partitioning method according to which a current
`
`block is split is ternary splitting and a ratio of the longer side to the shorter side
`
`of the current block is greater than a predetermined value. Accordingly, the
`
`encoder is capable of improving coding efficiency.
`
`For example, the circuitry writes the first condition into syntax that isa
`
`sequence layer, a picture layer, or a slice layer.
`
`This enables the encoderto transmit, to a decoder, informationrelating
`
`to a block partitioning method candidate to be ehmimated. Accordingly, the
`
`decoder is capable of improving decodingefficiency.
`
`For example, the circuitry writes the first condition into a sequence
`
`parameter set (SPS).
`
`This enables the encoder to transmit, to a decoder, information relating
`
`to a block partitioning method candidate to be elimimated. Accordingly, the
`
`decoderis capable of improving decodingefficiency.
`
`For example, a decoder according to one aspect of the present disclosure
`
`includes circuitry and memory. Using the memory, the circuitry: parses, from
`
`an encoded bitstream generated by encoding an image, block partitioning
`
`20
`
`information relating to a block partitioning method according to which a
`
`current block to be split im the image is spht, and splits the current block based
`
`on the block partitioning information parsed. When a shape of the current
`
`block satisfies a first condition, the block partitioning information is generated
`
`by G@generating one or more second candidates for a block partitioning method
`
`25
`
`by eliminating one or more predetermined candidates from a plurality offirst
`
`candidates for a block partitioning method and Gi
`
`selecting a block
`
`partitioning method from among the one or more second candidates.
`
`16
`
`
`
`This enables the decoder
`
`to spht a current block into shapes
`
`corresponding to shapes obtained by using ablock partitioning method selected
`
`from among candidates for a block partitioning method generated under a
`
`given condition by eliminating one or more candidates from a multiple number
`
`of candidates for a block partitioning method. Accordingly,
`
`the decoder is
`
`capable of splitting the current block into the shapes corresponding to the
`
`shapes obtained by using the block partitioning method efficiently selected.
`
`In
`
`addition, the decoder is capable, under a given condition, of prohibiting the
`
`appearance of a block having a shape corresponding to a shape obtained by
`
`using a block partitioning method that has been eliminated. Therefore, when
`
`a decoding mode is determined using an optimization method such as an R-D
`
`optimization, the number of variations for trial calculations decreases andit is
`
`expected that the amount of processing for decoding decreases while the
`
`degradation of decoding efficiency is
`
`inhibited. Moreover,
`
`the decoder
`
`intentionally biases the generation frequency of the information relating to a
`
`block partitioning direction. This increases accuracy in probability estimation
`
`in arithmetic decoding using a context, such as CABAD, and the improvement
`
`of decoding performance can be expected.
`
`For example, when the shape of the current block does not satisfy a
`
`20
`
`second condition, the circuitry performs a decoding process by G) parsing the
`
`block partitioning information relating to the block partitioning method
`
`according to which the current block is split and Gi) splitting the current block.
`
`Whenthe shape of the current block satisfies the second condition, the circuitry
`
`performs a decoding process by splitting the current block without parsing the
`
`25
`
`block partitioning information.
`
`This enables the decoder to reduce the amount of decoding by not
`
`decoding informationrelating to encoding and a bitstreamgenerated by writing
`
`fennk, ~4
`
`
`
`block partitioning information into syntax. Accordingly, the decoder is capable
`
`of improving decoding efficiency.
`
`The block partitioning information relates to, for example, at least one
`
`of a number into which the current block is split or a direction in whichthe
`
`Or
`
`current block is split.
`
`This enables the decoder
`
`to inchide,
`
`into the block partitioning
`
`information, informationfor uniquely determining a block partitioning method.
`
`The second conditionis, for example, that the block partitioning method
`
`is uniquely determined based on the shape of the current block.
`
`With this, when a block partitioning method is uniquely determined
`
`based on the shape of a current block to be split, the decoder is capable of
`
`splitting the current block into shapes corresponding to shapes obtained by
`
`using a block partitioning method selected from among candidates for a block
`
`partitioning method generated by eliminating one or more candidates from a
`
`multiple number of candidates for a block partitioning method. Accordingly,
`
`whena block partitioning method is uniquely determined based on the shape of
`
`the current block, the decoderis capable of splitting the current block into the
`
`shapes corresponding to the shapes obtained by using the block partitioning
`
`method efficiently selected.
`
`20
`
`The first condition is, for example, that the shape of the current block is
`
`arectangle,
`
`With this, when the shape of the current block is a rectangle,
`
`the
`
`decoder is capable of splitting the current block into shapes corresponding to
`
`shapes obtained by using a block partitioning method selected from among
`
`25
`
`candidates for a block partitioning method generated by ehmimating one or
`
`more candidates from a multiple numberof candidates for a block partitioning
`
`method. Accordingly, when the shape of the current block is a rectangle, the
`
`18
`
`
`
`decoder is capable of splitting the current block into the shapes corresponding
`
`to the shap