Efficient iterative algorithm for phase-shifting interferometry

This paper presents an efficient iterative least-squares algorithm for phase-shifting interferometry. If the phase shifts contain errors, the residuals (i.e. the differences between the recorded intensities and their recalculated values) of least-squares algorithm are dependent on these errors. Based on this fact, an approach that allows estimating the phase shift errors from these residuals is derived. Using the results, the values of phase shifts are corrected. By iterating the two steps: phase evaluation and phase shift correction, the phase distribution can be accurately reconstructed. Its validity and performances have been investigated by both the numerical simulation and experiment results. The convergence condition with this algorithm is also discussed.