METHOD AND APPARATUS FOR ADAPTIVE PREDICTIVE ENCODING/DECODING OF MULTI-LEVEL PICTURE SIGNALS

In method and apparatus for adaptive predictive encoding and decoding a multi-level television signal, an optimum prediction function of the present picture element time is predicted by taking into consideration the spatial distribution states of optimum prediction functions corresponding to previously encoded reference picture elements. In a modification, for a motion-compensated interframe encoding and decoding method and apparatus, the same principle is applied to a television picture including scene change and motions exceeding a detection range.