Adaptive weighted Gerchberg-Saxton algorithm for generation of phase-only hologram with artifacts suppression

In the conventional weighted Gerchberg-Saxton (GS) algorithm, feedback is used to accelerate convergence. However, it will lead iteration divergence. To solve this issue, an adaptive GS algorithm proposed in paper. By replacing with our designed feedback, convergence can be ensured method. Compared traditional method, method improves peak signal-noise ratio of reconstructed image 4.8 dB on average. Moreover, approximate quadratic phase suppress artifacts optical reconstruction. Therefore, a high-quality without Argument Reality device. Both numerical simulations and experiments have validated effectiveness