Moving sine average–based surface recovery method with enhanced noise resistance in white light interferometry

White light interferometry (WLI) is widely used in the measurement of surface profiles with the advantage of high-precision and nondestructive. Because of the mechanical vibration and the instability of the light source in the process of signal acquiring, the undesired signals including linear or nonlinear baseline drift and intensity noise will be introduced into the collected interference signal, which decreases the quality of the extracted envelope and the positioning accuracy of the envelope peak. Thus, we propose a moving sine average-based filtering method in the time domain to significantly reduce the negative impact of the undesired signal. Besides, the position of zero optical path difference is accurately located by position compensation of the envelope peak with its phase on the basis of half-wave loss analysis in the Mirau objective. Experiments have been performed on Mirau-type WLI, and a standard trapezoidal step sample is measured. Both the relative error of 0.049% (compared with the nominal height of 11.97±0.050 μm) and the standard deviation of 0.002 μm demonstrate the accuracy and the repeatability of the proposed recovery method.