Phase shifting interferometry using a robust parameter estimation method

The paper proposes a maximum-likelihood (ML) method based on spectral estimation theory for the estimation of phase distribution in interferometry in the presence of nonsinusoidal waveforms, noise, and the miscalibration of the piezoelectric device. The proposed method also allows the use of arbitrary phase steps. ML estimators are asymptotically efficient for large number of data samples. The method is complemented well by the incorporation of global search algorithm known as Probabilistic Global Search Lausanne for minimizing the ML function. The performance of the proposed method is studied in the presence of noise.