| 摘要 |
<p>PURPOSE:To simplify the device and to improve the coding efficiency by decomposing a picture signal into a high level and a low level signal designating a level of the picture signal roughly and finally, using the high level signal to discriminate a direction of contour and selecting adaptively a reference signal for predicting the low level signal. CONSTITUTION:A picture signal decomposing device 1 decomposes a picture signal (x) into a high level (y) and a low level signal (z) designating a level of the picture signal roughly and finally. A reference signal decision device 3 uses the signal (y) to detect a direction of a contour of the picture and decides the reference signal at coding of the signal (z) and adds a direction code S and reference signals yi, yj, yk to a low level signal coder 4. A code synthesizer 5 selects codes in the order of Cy, Cz and sends the selected code to a decoder side via a transmission line 6. A code decomposing device 7 decomposes the compressed code C' into codes Cy', Cz' corresponding to the high level and the low level signal and high level and low level signal decoders 8, 10 are used to apply inverse processing to the coders 2, 4 and a picture signal synthesizer 11 synthesizes the codes Cy', Cz' to obtain a picture signal x'. Thus, the coding efficiency is improved without making the device complicated.</p> |
