| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To improve decoding efficiency of a motion vector. <P>SOLUTION: A candidate list generating section selects a plurality of blocks having one or two pieces of motion information including at least information on the motion vector and information on a reference image from a plurality of decoded blocks adjacent to a block being a decoding object, and generates a candidate list including candidates of the motion information used in motion compensation prediction from the motion information of the selected block. A first motion information acquiring section acquires motion information of a first prediction list from a first candidate included in the candidates. A second motion information acquiring section acquires motion information on a second prediction list from a second candidate included in the candidates. A selected candidate generating section combines the motion information of the first prediction list which is acquired by the first motion information acquiring section, and motion information of the second prediction list which is acquired by the second motion information acquiring section, to generate a new candidate of the motion information. <P>COPYRIGHT: (C)2013,JPO&INPIT</p> |
