SIGNAL WINDOW MINIMUM AVERAGE ERROR ALGORITHM FOR COMPUTER-GENERATED HOLOGRAMS

We present an explanation of the minimum-average-error (MAE-) based er ror diffusion algorithm for computer-generated hologram (CGH) calculat ion. This leads to a direct and straightforward link between the CGH r econstruction plane signal windows and the MAE diffusion weights: the diffusion weights should be the Fourier transform of the signal window function. A MAE algorithm based on these results is described and use d to calculate test CGH's, whose computer-simulated and experimental r econstructions confirm our analysis by generating true, low-error sign al windows akin to those obtained with iterative algorithms. Compariso ns made with an iterative algorithm show that the new algorithm is a p owerful, low-computation-load, CGH binarization tool and that, when co mbined with random or image-independent diffusers, it makes possible t he calculation of acceptable-performance, high-space-bandwidth-product CGH's whose calculation would be unfeasible with iterative algorithms . (C) 1998 Optical Society of America.