Encoding and decoding code sequences and frames

摘要

A method for encoding a picture based on an intra-picture coding and an inter-picture prediction coding having an absolute value of a maximum range of motion compensation between continuous pictures in terms of time being H pixels (H: a positive integer) in a first direction and V pixels (V: a positive integer) in a second direction, the method includes the steps of: receiving a picture and judging whether the picture is directly/indirectly referred to by a further picture among the continuous pictures in terms of time or not; determining a maximum distance L (L: a positive integer) between the picture and the further picture directly/indirectly referring to the picture; dividing the picture into a first to an Nth sub-pictures (N: an integer equal to or larger than 2), a size of the first to the Nth sub-pictures being larger by at least LxV pixels in the first direction or at least LxH pixels in the second direction as compared with a region of a case where the picture is equally divided into N regions which do not overlap each other; generating a first to an Nth code sequences corresponding to the respective first to the Nth sub-pictures by encoding the first to the Nth sub-pictures based on the intra-picture coding and the inter-picture prediction coding; and adding information indicating positions of the first to the Nth code sequences in the picture.