| 摘要 |
PURPOSE: An inter prediction encoding method is provided to reduce an encoding amount of motion information of a current prediction unit by using spatial and temporal motion vector candidates. CONSTITUTION: A video encoding/decoding apparatus obtains differential motion vectors and a reference picture index of a current prediction unit (S310). The apparatus obtains a spatial motion vector predictor (S320). The apparatus obtains a temporal motion vector candidate (S330). The apparatus generates a motion vector candidate list (S340). The apparatus obtains a motion vector predictor of the current prediction unit (S350). The apparatus restores a motion vector of the current prediction unit (S360). [Reference numerals] (AA) Start; (BB) End; (S310) Drive differential motion vectors and a reference picture index; (S320) Drive a spatial motion vector predictor; (S330) Drive a temporal motion vector candidate; (S340) Generate a motion vector candidate list; (S350) Drive a motion vector predictor of the current prediction unit; (S360) Restore the motion vector of the current prediction unit; (S370) Generate a prediction block using motion information; (S380) Generate a remaining-difference block; (S390) Generate a restoration block |
