| 摘要 |
PURPOSE:To perform quantization suitable for image signal processing by executing A/D conversion of lowering quantization accuracy each time the input analog signal level exceeds a prescribed value and by executing subtraction of a prescribed number and a prescribed shifting in accordance with the level of the conversion output level. CONSTITUTION:An input analog signal 19 is converted in a folded line form into digital signal with lowering in the quantization accuracy at a prescribed ratio each time it exceeds a prescribed value. The converted output 21 is supplied to a level decision circuit 22 for decision of the domain of the quantized level while also being supplied to a bit shift circuit 24. The shift quantity of the shift output 26 is arithmetically changed in accordance with the level domain where decision is made. Also, an output 27 of a constant generating circuit 25 is output corresponding to the level domain. The output 27 is subtracted from the output 26 of the subtractor 28 so that the input analog signal level is raised. Although the quantization accuracy is lowered, linear digital output is obtained. In such a manner, the number of quantization bits is suppressed to make possible quantization suitable for image signal processing.
|
