| 摘要 |
PURPOSE:To perform high-efficiency compression encoding without paying attention to variation in the distribution of the generation probability of an error in forecasting, by finding the forecasting error signal of a multilevel picture signal being inputted by using a picture signal having been inputted, and grouping the signal by mode signal values. CONSTITUTION:A picture signal having been inputted is stored in a shift register 1 for picture signal storage, a forecasting signal generator 2 generates a forecasting signal X, and a subtracting circuit 3 obtains a forecasting error signal (e) regarding a multilevel picture signal X being inputted. A mode signal M representing the probability of correct forecasting is generated by a mode signal generator 4 and applied to a forecasting error signal encoding circuit 5 together with the error signal from the circuit 3. Then, the picture signal inputted already is used to generate a mode signal for the forecasting error signal, and the mode signal is used to compress and encode the error signal, thus achieving high- efficiency compression encoding without paying attention to the variation of the distribution of the generation probability of a forecasting error. |
