VIDEO STREAM CODING METHOD ACCORDING TO PREDICTION STRUCTURE FOR MULTI-VIEW VIDEO AND DEVICE THEREFOR, AND VIDEO STREAM DECODING METHOD ACCORDING TO PREDICTION STRUCTURE FOR MULTI-VIEW VIDEO AND DEVICE THEREFOR

摘要

Provided is a video stream decoding method comprising: obtaining, from a video stream of images coded into multiple layers, prediction information on an advanced motion vector prediction (AMVP) mode, which has been generated using a candidate list comprising two candidates, and a motion vector prediction flag, which indicates the candidate used for generating the prediction information on the AMVP mode between the two candidates comprised in the candidate list, for coding data on a current view image; and generating data on a motion-compensated current view image by using data on the current view image and/or data on a different view image on the basis of the obtained information, thereby decoding the current view image.