Analysis and reduction of errors caused by Poisson noise for phase diversity technique.

An effective method for reducing the sensitivity of phase diversity (PD) technique to Poisson noise is proposed. The denoising algorithm based on blocking-matching and 3D filtering is first introduced in the wavefront sensing field as a preprocessing stage. Then, the PD technique is applied to the denoised images. Results of the numerical simulations and experiments demonstrate that our approach is better than the traditional PD technique in terms of both the root-mean-square error (RMSE) of phase estimates and the structural similarity index metrics (SSIM). The RMSEs of phase estimates on synthetic data are decreased by approximately 40% across noise levels within the range of 58.7-18.8 dB in terms of peak signal-to-noise ratio (PSNR). Meanwhile, the overall decline range of SSIM is significantly decreased from 49% to 9%. The experiment and simulation results are in good agreement. The approach may be widely used in various domains, such as the measurements of intrinsic aberrations in optical systems and compensations for atmospheric turbulence.