Amplitude-only feedback iterative optimization algorithm for phase-only holograms

Iterative algorithms based on the fast Fourier transform (FFT) are widely used in the calculation of phase-only holograms due to their advantages of easy implementation and fast computation. However, existing iterative algorithms continuously change the target phase during the iteration process, and the varying noise caused by the changing phase imposes limitations on the optimization upper limit of these algorithms. Here, we propose an amplitude-only feedback (AOF) iterative optimization algorithm. The AOF iterative algorithm locks in the noise caused by irregular phases and maximally offsets such noise through amplitude compensation. With a small number of iterations, the AOF iterative algorithm can further significantly improve the reconstruction quality of existing phase holograms: it increases the peak signal-to-noise ratio (PSNR) of hologram reconstruction images generated by the Gerchberg-Saxton (GS), Fienup, and Fienup with don't-care (Fidoc) algorithms by 16.34, 17.50, and 12.52, respectively. The high speed and broad compatibility of AOF iteration make it promising to improve the reconstruction quality of various phase hologram generation algorithms at low cost and greatly expand their application prospects.