Noise reduction method in both time and frequency domains for envelope peak position extraction of pulse-train interferometry measuring distance between interference fringes

Abstract

Multi-pulse train interferometry is used to measure distance for meter-scale/nanometer-resolution geometric measurements. The core of multi-pulse train interferometry is object positioning accuracy, which is limited by the ability to position the interference fringe envelope peak. The Fourier fringe analysis method, optical Hilbert transform method, and other methods are used in the envelope demodulation of multi-pulse train interferometry. The matched filter, which provides good noise reduction capability in the time domain, is widely used in radar signal processing. In this study, a matched filter is combined with the Fourier fringe analysis method and used to demodulate the envelopes of a fringe in a multi-pulse train interferometry because the proposed combined method performs noise reduction in both the time and frequency domains for fringe envelope signal processing. Compared with two conventional methods, the proposed combined method was proven to be feasible for envelope demodulation of multi-pulse train interferometry.