Video encoding method and apparatus, and video decoding method and apparatus

Abstract: A method for encoding a video block using reference blocks comprises assigning the video block to one of first and second prediction groups, and encoding the video block according to a motion compensated prediction encoding mode, using the reference blocks depending on the one of the first and second prediction groups to which the video block is assigned, one of the reference blocks being a decoded block, wherein a first prediction group is obtained by a prediction using the reference blocks belonging to a first prediction group, and a second prediction group is obtained by a prediction using the reference blocks belonging to at least one of the second prediction group and the first prediction group. (end of abstract)