| 摘要 |
<p>PURPOSE:To minimize all the generated information quantity, and to obtain a high encoding efficiency, by using information for representing each forecasting function and forecasting error information by each forecasting function, for detecting a dynamic vector. CONSTITUTION:A video signal from an input line IN is A/D-converted 11, and 1 frame is divided into blocks of plural picture elements by a scanning and converting circuit 12. This signal is supplied to both a delaying circuit 13 and a dynamic vector outputting circuit 14. Subsequently, a dynamic vector of the circuit 14 is sent to a code compressing circuit 20 and a variable delaying circuit 16, and the circuit 16 generates a forecasting signal from the dynamic vector and a picture signal supplied from a frame memory 15. This forecasting signal is supplied to a subtracting circuit 17 and an adding circuit 18, and the circuit 17 executes subtraction between the circuit 13 and the forecasting signal, and outputs a forecasting error signal of its difference. This forecasting error signal is electronized 19, its picture element and line are converted to samples by a gate circuit 21, and it is added 18 to the forecasting signal and also code-compressed 20.</p> |
