Processing of Interferometric Data

Stefan Trittler, Fred A. Hamprecht

Aim

More and more parts manufactured today are subject to very tight tolerances. Quality control requires fast and precise measurement strategies in order to test these parts.

High-precision 3D measurements of technical surfaces can be done using interferometry. For rough surfaces white light interferometry and multiple wavelength interferometry can be used. These techniques are used in laboratories, but in many cases they are too slow for in-line inspection. Industrial application of interferometric systems requires robust and fast measurement strategies and algorithms.

Methods

Interferometric systems can be improved by using optimized sampling strategies, gaining as much information as possible while keeping the measurement time as short as possible. This reduces measurement time, but it can lead to data that is not uniformly sampled and might be more sensitive to errors due to vibrations.

In order to interpret this kind of data we are developing new algorithms for data analysis. Our research focuses on accurate frequency estimation for non-equispaced samples of harmonic processes.

A 3D surface measured by white light interferometry (height color-coded) and an image from a sequence of images taken during a WLI measurement (klick to download the video).