| 摘要 |
PURPOSE: A method for decoding an inter prediction-coded moving picture is provided to encode and decode a motion vector of a current prediction unit efficiently by using a motion vector of a prediction unit adjacent to a current prediction unit and a motion vector at a predetermined position in a temporarily different picture. CONSTITUTION: A moving picture decoding device acquires a reference picture index and a differential motion vector of a current prediction unit from a received bit stream, and acquires left and upper spatial motion vector candidates. After temporal motion vector candidates are acquired, the device removes a subordinated motion vector candidate from a constructed motion vector list according to the existence of duplicate motion vector candidates. If the number of motion vector candidates on the list is smaller than a predetermined number, the device adds a zero vector. [Reference numerals] (AA) Current prediction unit |
