| 摘要 |
<p>In the encoding of video signals, motion vectors are frequently determined for picture element blocks. In EP 0 414 113, a method is described how a zoom and pan vector can be generated for in each case one frame from a field of motion vectors for picture element blocks. During a format conversion of video signals which are undersampled at least vertically before the encoding, motion parameters applicable in each case to a frame are used in order to receive maximum spatial resolution after the undersampling. Only the remaining incompletely motion-compensated picture parts are detected by means of a simple change vector and vertically filtered in the field. <IMAGE></p> |
