Gradual and random binarization of gray-scale holograms.

A new method called gradual and random binarization to binarize gray-scale holograms, based on an iterative algorithm, is proposed. The binarization process is performed gradually, and the pixels to be binarized are chosen randomly. Errors caused by this operation are spatially diffused. A comparison with other established methods based on error diffusion, direct binary search, and iterative stepwise quantization shows that the gradual and random binarization method achieves a very good compromise between computational complexity and reconstruction quality. Optical reconstructions are presented.