An improved Stoilov algorithm based on probability and statistics

Abstract

In phase measuring profilometry (PMP), Stoilov is a popular algorithm with satisfied effect. But Stoilov algorithm relies too much on gray level information of fringes. Meanwhile, in the phase calculation, the light intensity subtraction, division or roots extraction will cause some singular phenomena, such as making the numerator or denominator zero in some positions or making roots plural, etc. This would lead to meaningless or large error in phase unwrapping, which might cause the reconstruction distorted or even unable to reconstruct. This paper proposes an improved Stoilov algorithm based on probability and statistics. Suppose that the probability of real data is the highest, we choose the cosine square matrix value of phase shift for processing. After ruling out the singular points of large error, we get the closest value to true phase shift using the method of probability and statistics. Experimental results show that this method can effectively restrain singular phenomenon, and can ensure the accuracy of PMP.