CODING PROCESS

摘要

With the coding of picture signals, motion vectors are often determined for blocks of picture elements. A method is described in EP 0 414 113 which explains how a zoom and pan vector can be generated for always a complete picture from an array of motion vectors for blocks of picture elements. With a format conversion of picture signals which are sub-sampled at least vertically prior to the coding, motion parameters valied for always a complete picture are used in order to gain a maximum local resolution after the subsampling. Only the remaining, incompletely motioncompensated parts of the picture are detected by a simple alteration detector and vertically filtered in the field. Fig. 1