SEM AUTOMATION USING AUTOCORRELATION TECHNIQUES

Abstract

This project uses autocorrelation to investigate the noise behaviour of Scanning Electron Microscope images. Autocorrelation techniques have also been used to study the variation of parameters like focusing and magnification on the images. Based on these results an autofocusing algorithm for the Scanning Electron Microscope has been developed. Two dimensional autocorrelation plots of images reveal image noise which is confined to the vicinity of the origin. These plots also give relevant information on low-pass filtered images, and help to choose suitable analog time constants for the system such that it improves image noise, with minimum blurring. The information collected has been put into an algorithm which detects features in a sample and brings it to a fine focus. This saves the operator's effort especially in sparse samples and defocused images. The algorithm developed was found to be reliable on a variety of samples.