The two level quantization of gray level picture signal

The methods for the quantization of gray level picture signal into two levels are discussed. With ordinary facsimile quantizer, input signal is applied to a comparator, where it is compared with a fixed threshold. But since this quantizer will hard limit the signal, the coding characteristics are very poor for half tone pictures and documents with tint background. Quantizers with fixed threshold, moving average threshold, dither threshold and their combinations were examined. Through simulations and experiments on actual facsimile equipment, it was determined that the combined system of moving average and dither threshold gives the best reproduction for photographic images of human face.