摘要 |
Disclosed is a video-encoding method using prediction units based on encoding units determined in accordance with a tree structure, wherein the video encoding method involves: dividing an image of a video into one or more maximum encoding units; encoding, for each maximum encoding unit, the image on the basis of encoding units for each coded depth, which are divided hierarchically in accordance with coded depths, and on the basis of partition types determined on the basis of the coded depths of the encoding units for each coded depth to determine encoding units in accordance with a tree structure; and outputting data encoded on the basis of encoding units determined in accordance with a tree structure and on the basis of partition types, information on coded depths and encoding modes, and information on an encoding unit structure, which indicates encoding unit sizes and variable coded depths. |