Pixel retrieval for frame reconstruction

A video decoder receives data representing an encoded image frame partitioned into one or more partitions. The video decoder stores the motion vector associated with each partition. For partitions of size (k*S), k being an integer, and S being the smallest allowed size for a partition, the video decoder stores (k) separate entries of a same value as the value of the motion vector of the corresponding partition. When performing motion compensation to reconstruct the encoded macro-block, such a manner of storing enables the video decoder to retrieve pixel values of best-match partitions associated with each of the corresponding ones of the one or more partitions without having to process partition information specifying the type of the partitions.