Laser interferometry-based characterization of surfaces by datadependent systems

Document Type

Conference Proceeding

Publication Date

12-10-1993

Abstract

© 1993 SPIE. All rights reserved. In the process of surface recovery using interferometric methods, a mixture of artifacts, noise and errors affect the phase map stage of the process. Based on a methodology called data dependent systems (DDS), a stochastic approach is developed which will discern the phase signal from noisy data points and use it to reconstruct a clean phase map. Such a phase map is then easily converted to a height distribution. Stepwise, the procedure first identifies the presence of discontinuities, then applies an adaptive thresholding criteria to return an outlier free set of residuals, and finally uses them to reconstruct a clean phase map.The methodology is robust and can handle very complex phase maps. It reduces reliance on complicated and expensive experimental setups and replaces them with relatively inexpensive but reliable computation. Phase unwrapping is shown to be accomplished in a natural way, but the method also provides other benefits such as diagnostic identification of surface defects and other trouble spots.

Publication Title

Proceedings of SPIE - The International Society for Optical Engineering

Link to Full Text

Share

COinS