METHOD FOR MULTI-VIEW VIDEO ENCODING/DECODING AND APPARATUS THEREOF

摘要

PURPOSE: A multiview video coding/decoding method and an apparatus thereof are provided to improve the motion vector prediction efficiency and the motion vector transmission efficiency by performing the motion information prediction based on the time difference information between the points and the motion information between the points. CONSTITUTION: A decoder receives an encoded prediction MV(Motion Vector) index and an encoded MV difference, and entropy decodes the prediction MV index and the MV difference(S910). The decoder extracts at least one of the prediction MV candidates and produces a prediction MV candidate group based on the extracted prediction MV candidate(S920). The decoder extracts an MV of current block based on the MV index, the MV difference, and the prediction MV candidate group(S930). [Reference numerals] (S910) Receive a prediction motion vector index and motion vector difference(or merge index) and decode; (S920) Generate a prediction motion vector candidate group(or merge candidate group); (S930) Extract motion information