Abstract
A phase shift technique is proposed to analyze closed-fringe patterns. This technique requires two fringe patterns to calculate wrapped phase data, where there is a π/2 phase shift between the two patterns. Fourier transform is applied to transform each of the fringe patterns into frequency domain. Then a filter is designed to remove low frequency and high frequency components that correspond to unwanted background and noise of the fringe patterns. After filtering, two frequency spectra are inverse transformed to spatial domain and a wrapped phase map is calculated from two real parts of the inverse Fourier transform. Finally, a continuous phase profile is retrieved from the wrapped phase map. This paper represents the principle of the proposed method, discusses different filters' effects on the recovered phase data and tests the validity of the proposed method on real interferograms obtained by Electronic Speckle Pattern Interferometry as well as computer-simulated fringes.
