Wavefront Reconstruction from Noisy Slope or Difference Data Using the Discrete Fourier Transform

A general algorithm is presented for reconstructing a two-dimensional wavefront optical path difference (OPD) map from noisy slope or difference measurements by means of a least squares fit using complex exponentials. This form of modal estimation can be described as a filtering operation in the spatial frequency domain. Thus fast Fourier transform (FFT) algorithms can be used for rapid reconstruction. The reconstruction is unbiased also in the case of finite data arrays. The error propagation from the noisy measurement data to the integrated wavefront is minimal in a least squares sense. It is believed that this reconstruction algorithm can be implemented in an adaptive optical system by using commercially available array processor hardware, thus reducing the total system cost and the need for specialized hardware.