Method and device for encoding and decoding.

Method for encoding video image sequences in which images of the image sequence are encoded escaladamente such that the video data resulting contain information that ensure representation of the images in multiple steps of different resolution for each image, coming defined the corresponding resolution level for the number of image points per image representation of the images, performing the block-based encoding and generating a block structure which can subdivide the image in the corresponding resolution level into blocks, partial blocks and subblocks, this subdivision describing a possible motion contained in the image sequence and subdividing the block structure starting from a block in partial block and partial block part (MB1_QCIF..MB4_QCIF, MB1_CIF..MB4_CIF, MB1_4CIF .. MB4..4CI) in successive sub-blocks div ission finer having the following steps: a) temporarily generated for a first resolution level a first block structure (MV_QCIF; MV_CIF) and for a second resolution level a second block structure (MV_CIF; MV_4CIF), wherein the first resolution level a number of points lower than the second resolution level, b) the second block structure image (MV_CIF ; MV_4CIF) is compared with the first block structure (MV_QCIF; MV_CIF) such that differences in the subdivision of the block structures are detected as differences in structure, c) based on the characteristics of differences in structure, check which part of the second block structure is to be taken in a second structure of modified blocks (MV '' _ CIF; MV '' _ 4CIF) and the second structure of modified blocks is generated on the basis of the verification, d) the second modified block structure (MV '' _ CIF; MV '' _ 4CIF) is taken as a basis for encoding the image sequence.