High noise-resistant slope and curvature signal extraction algorithm in frequency domain for the Shack-Hartmann wavefront sensor

Conventional Shark-Hartmann wavefront sensors (SHWS) only get the slope information of the wavefront in each subaperture. In this paper, Fourier demodulation technology and finite difference method are used to process the Light spot column plot of the SHWS, and the slope and curvature information of the wavefront in each subaperture can be obtained. With the slope and curvature information of the wavefront obtained by the algorithm as input, the distorted wavefront of the input is reconstructed by using the slope and curvature hybrid wavefront reconstruction technology. Simulation results show that the relative reconstruction error of wavefront is between 1%-5% under ideal conditions. In the noisy condition, the relative reconstruction errors of the wavefront are also between 1%-5%, and the wavefront can be recovered well. Compared with the traditional spatial centroid detection algorithm, this algorithm is simple and convenient to process information in the frequency domain, and has strong anti-noise ability.