Figure 3.

Processed images for the computer-generated holograms. (a) UniPD Logo bitmap format with pure black and white pixel, (b) two intersecting ‘H’ with a grayscale 8 bit/channel and (c) wolf portrait characterized by a 8 bit/channel grayscale with finer details. (d) Schematic representation of the iterative Fourier transform algorithm. After the signal input of the ith iteration step in the object domain enters the loop, the inverse fast Fourier transform (FFT⁻¹) allows the transition to the hologram plane with the hologram function H _i (1), before normalization with respect to the incident field U ⁱ, hence the quantization operator Q is applied for both direct partial quantization of phase and amplitude elimination (2). The discretized hologram pattern H _i ′ is multiplied by the incident field U ⁱ and the fast Fourier transform (FFT) is performed (3), obtaining the corresponding reconstructed field hi′ in the object domain. Then the final new signal h _i+1 is obtained with the proper replacement of the output signal amplitude with the desired image amplitude within the signal window (4). The loop is repeated (5) for N iterations, until convergence. The University of Padova logo is © University of Padova and used with permission.