A maximum likelihood approach to wavefront sensing

A novel approach to two-dimensional wavefront sensing based on a maximum likelihood algorithm is introduced. The data used is the same as the Misell type algorithm i.e. two or more intensity patterns near the focus. The algorithm developed is iterative but requires only a forward transformation not an inverse transformation. This means that the method has the advantage of not being restricted to low numerical aperture cases. The feasibility of the method is demonstrated through numerical simulations. These demonstrate that compared with the Misell algorithm, the maximum likelihood method is relatively insensitive to noise and works well for a wider range of chosen defocus settings. This latter feature is particularly important for a practical application of the technique for lab-based testing of lenses. Results of applying the method to experimental data are also presented.