Abstract:As an important measuring technique, white light interferometry can realize fast and non-contact measurement, thus it is now widely used in the field of ultra-precision engineering. However, the traditional recovery algorithms of surface topographies have flaws and limits. In this paper, we propose a new algorithm to solve these problems. It is a combination of Fourier transform and improved polynomial fitting method. Because the white light interference signal is usually expressed as a cosine signal whose amplitude is modulated by a Gaussian function, its fringe visibility is not constant and varies with different scanning positions. The interference signal is processed first by Fourier transform, then the positive frequency part is selected and moved back to the center of the amplitude-frequency curve. In order to restore the surface morphology, a polynomial fitting method is used to fit the amplitude curve after inverse Fourier transform and obtain the corresponding topography information. The new method is then compared to the traditional algorithms. It is proved that the aforementioned drawbacks can be effectively overcome. The relative error is less than 0.8%.

An Accurate Surface Topography Restoration Algorithm for White Light Interferometry