Removing the quantization error by repeated observation [image processing]

An image processing method, which can reduce the quantization error by averaging a set of measures performed on a relatively stable signal, is introduced. The use of noise that may be contaminating the signal or the addition of sawtooth wave function previous to quantization are the keys to this method. Also, averaging a signal corrupted by 'natural' noise may decrease not only the error due to the noise but the quantization error as well. This method can be used to achieve high precision results which could not be achieved by A/D converters with more bits. The results of computer simulated experiments for this method are discussed. >