Comparison of cubic B-spline and Zernike-fitting techniques in complex wavefront reconstruction.

We analyze an alternative to classical Zernike fitting based on the cubic B-spline model, and compare the strengths and weaknesses of each representation over a set of different wavefronts that cover a wide range of shape complexity. The results obtained show that a Zernike low-degree polynomial expansion or a cubic B-spline with a low number of breakpoints are the best choices for fitting simple wavefronts, whereas the cubic B-spline approach performs much better when more complex wavefronts are involved. The effect of noise level in the fit quality for the different wavefronts is also studied.