A model of confocal microscopy measurements combining empirical and physical properties

Abstract

Virtual measuring instruments are useful tools in surface topography measurement and can be applied for the prediction of measuring results and the determination of measurement uncertainty. For optical measuring instruments, often computationally expensive ray-tracing algorithms are applied. Other models provide an approximation of the transfer behavior based on techniques from signal processing or physical considerations of the instrument setup. For confocal microscopy, we suggest the application of a hybrid model which combines a linear filter model with limitations caused by physical properties of the measuring instrument. The application of the comprehensive model and the estimation of measurement uncertainty based on the instrument characteristics are presented. The results are in good compliance with experimental data and the model requires very few computing effort. An easy-to-use approach to implement the corresponding models for a specific measuring instrument is presented.

Three-dimensional confocal reflectance microscopy for surface metrology

“…The fitting process enables a more accurate estimation of the peak position, that is, sample height. Since the peak detection is essential in confocal microscopy, various approaches are constantly being proposed to enhance the efficiency and accuracy of calculation, such as a modeling of confocal microscopy considering physical properties [49], mean-shift vector [50], and Monte Carlo analysis [51]. With peak detection algorithms, the axial resolution, or repeatability of confocal reflectance microscopy is much better than the FWHM of the axial response curve or the depth sampling interval [37,46].…”