Apparatus and method for encoding or decoding video using adaptive motion vector resolution

A video encoding apparatus for encoding a sequence of one or more pictures in the units of blocks which are split from each picture, the apparatus comprising one or more processors configured to:

• encode, into a header of the sequence in a bitstream, first identification information indicating whether adaptive motion vector difference resolution is allowed in the unit of blocks which belong to the sequence of one or more pictures,
• encode, into the header of the sequence, size information for restricting block splitting in a tree structure which includes a quad-tree and a binary-tree rooted from a leaf node of the quad-tree, the size information including a first syntax element for representing minimum block size allowed for a leaf node of the quad-tree and a second syntax element for representing maximum block size allowed for a root node of the binary-tree,
• split a coding tree block in the tree structure under block size restriction defined by the size information to determine a current block to be encoded, and encode split information related to the splitting,
• determine a current motion vector of the current block among the blocks which belong to the sequence of one or more pictures,
• generate and encode a current motion vector difference of the current block by using the current motion vector of the current block and a motion vector corresponding to a neighboring block which has been already encoded prior to the current block and is adjacent to the current block,
• when the first identification information indicates that the adaptive motion vector difference resolution allowed and the current motion vector difference is non-zero, encode information on motion vector resolution dedicated to the current block among the blocks which belong to the sequence, wherein the information on motion vector resolution is used for representing a resolution at which the current motion vector difference is encoded among three or more motion vector resolutions including a ¼ pixel precision, and
• predict and encode the current block using the current motion vector of the current block,
• wherein the information on motion vector resolution is encoded by:
• encoding a motion vector (mv) resolution flag dedicated to the current block, the my resolution flag indicating whether the resolution of the current motion vector difference of the current block is equal to the ¼ pixel precision, and
• encoding alternative motion vector resolution information for representing a motion vector resolution which is dedicated to the current block among motion vector resolutions other than the ¼ pixel precision, only when the mv resolution flag indicates that the resolution of the current motion vector difference of the current block is not the ¼ pixel precision,
• wherein the alternative motion vector resolution information to be encoded is a log scale value of a ratio between the resolution of the current motion vector difference of the current block and a resolution of a motion vector difference of a block encoded prior to the current block,
• wherein, when the current motion vector difference of the current block is zero, the information on motion vector resolution is not encoded such that the ¼ pixel precision is used as the resolution of the current motion vector difference of the current block, and
• wherein, when the first identification information indicates that the adaptive motion vector difference resolution is not allowed, the resolution of the current motion vector difference of the current block is encoded at the ¼ pixel precision without the information on motion vector resolution encoded.

A method for transmitting a bitstream containing encoded video data, the method comprising:

• generating the bitstream by encoding a sequence of one or more pictures in the units of blocks which are split from each picture;
• and
• transmitting the bitstream to a video decoding apparatus,
• wherein generating the bitstream comprises:
• encoding, into a header of the sequence in a bitstream, first identification information indicating whether adaptive motion vector difference resolution is allowed in the unit of blocks which belong to the sequence of one or more pictures,
• encoding, into the header of the sequence, size information for restricting block splitting in a tree structure which includes a quad-tree and a binary-tree rooted from a leaf node of the quad-tree, the size information including a first syntax element for representing minimum block size allowed for a leaf node of the quad-tree and a second syntax element for representing maximum block size allowed for a root node of the binary-tree,
• splitting a coding tree block in the tree structure under block size restriction defined by the size information to determine a current block to be encoded, and encoding split information related to the splitting,
• determining a current motion vector of the current block among the blocks which belong to the sequence of one or more pictures,
• generating and encoding a current motion vector difference of the current block by using the current motion vector of the current block and a motion vector corresponding to a neighboring block which has been already encoded prior to the current block and is adjacent to the current block,
• when the first identification information indicates that the adaptive motion vector difference resolution is allowed and the current motion vector difference is non-zero, encoding information on motion vector resolution dedicated to the current block among the blocks which belong to the sequence, wherein the information on motion vector resolution is used for representing a resolution at which the current motion vector difference is encoded among three or more motion vector resolutions including a ¼ pixel precision, and
• predicting and encoding the current block using the current motion vector of the current block,
• wherein the information on motion vector resolution is encoded by:
• encoding a motion vector (mv) resolution flag dedicated to the current block, the my resolution flag indicating whether the resolution of the current motion vector difference of the current block is equal to the ¼ pixel precision, and
• encoding alternative motion vector resolution information for representing a motion vector resolution which is dedicated to the current block among motion vector resolutions other than the ¼ pixel precision, only when the mv resolution flag indicates that the resolution of the current motion vector difference of the current block is not the ¼ pixel precision,
• wherein the alternative motion vector resolution information is a log scale value of a ratio between the resolution of the current motion vector difference of the current block and a resolution of a motion vector difference of a block encoded prior to the current block,
• wherein, when the current motion vector difference of the current block is zero, the information on motion vector resolution is not encoded such that the ¼ pixel precision is used as the resolution of the current motion vector difference of the current block, and
• wherein, when the first identification information indicates that the adaptive motion vector difference resolution is not allowed, the resolution of the current motion vector difference of the current block is encoded at the ¼ pixel precision without the information on motion vector resolution encoded.