Motion vector coding apparatus, method and program for coding motion vector, motion vector decoding apparatus, and method and program for decoding motion vector

A method for determining a predicted motion vector for decoding at least one blocks included in a picture of a video from a bit stream which is hierarchical coded using a plurality of layers, the method comprising:

• obtaining, for generation of the predicted motion vector in a target block of a current picture on one of a plurality of layers, at least one of
• (i) a first motion vector for a block at a position corresponding to the target block to be decoded of a previous picture before the current picture including the target block,
• (ii) a second motion vector for a block, at a position corresponding to the target block to be decoded, of a picture in another layer of the plurality of layers, and
• (iii) a third motion vector for one of neighboring blocks of the target block to be decoded of the current picture on one of the plurality of layers;
• generating a candidate group of motion vectors used for prediction of a motion vector for the target block, the candidate group having the second motion vector and the third motion vector excluding the first motion vector if the second motion vector is in the candidate group, and the candidate group having the first motion vector and the third motion vector if the second motion vector is not in the candidate group;
• determining the predicted motion vector from the generated candidate group for the target block in accordance with information about a predicted motion vector for the target block received from a bit stream;
• decoding a quantization coefficient for the target block to be decoded;
• and
• decoding the target block on the one of the plurality of layers using the determined predicted motion vector and the decoded quantization coefficient.