| 摘要 |
PURPOSE:To improve the degradation of the picture quality by reducing the quantization error brought about in processes of quantization and inverse quantization used for video signal compression. CONSTITUTION:At the time of recording, 8X8 unit blocks subjected to DCT operation are quantized by first and second quantizers 55 and 57 and are compressed with a variable length code. At the time of reproducing, the signal decoded by a variable length decoder 62 is subjected to inverse operation of quantization by an inverse quantizer 63. Simultaneously, a quantization discriminator 64 determines whether correction is necessary or not in accordance with a quantization condition (class information, color information, quantization step Brn, and area information) supplied from the inverse quantizer 63 and outputs a control signal to a corrector 65. The corrector 65 performs correction by addition/subtraction of a certain value to/from the inversely quantized value based on this control signal. The output signal of the corrector 65 is subjected to IDCT operation and is outputted to an output terminal 68. |
