Signal processing of high-resolution transmission electron microscope images using Fourier transforms

Abstract

Interpretation of images obtained by high-resolution transmission electron microscopy (HRTEM) can be aided by the use of processed signals. Images are recorded digitally or are digitized from a photograph, and then these data are Fourier transformed and treated. The low- and high-frequency signals are removed, and a variety of circular and elliptical (anisotropic) apertures or screens are applied to explore and highlight features of special interest. A minicomputer can be used to perform such image processing rapidly, interactively, and with high precision. Elliptical filters are of special interest for the examination of linear or planar features such as the distribution of stacking faults or the presence and distribution of superstructures. Some superstructures themselves contain defects, and these can similarly be highlighted and examined. The uses of processing for examining HRTEM images of layering in a complex sheet silicate mineral and dislocation cores in grossular garnet are illustrated.