Analysis, Filtering, and Decomposition of Electron Microscope Images Based on Orthogonal Transformations
B. N. Grudin, E. G. Kislenok, V. S. Plotnikov, and V. K. Fishchenko
Far East State University, Vladivostok, Russia, E-mail: grudin@lemoi.phys.dvgu.ru
Pages: 18-27
Subsection: ANALYSIS AND SYNTHESIS OF SIGNALS AND IMAGES
Abstract
Methods for estimating the integral characteristics of electron microscope image periodograms, based on Hartley transform, Q-transform, and cosine transform, are proposed. It is shown that the cosine transform is more appropriate for filtering and improvement of psychophysical image perception on its basis. Various approaches to implementing matched filtering via cosine transform are considered. A procedure of electron microscope image decomposition with the use of various orthogonal transforms and filters based on Fourier patterns of Gauss and Morle wavelets is proposed and studied. It is shown that electron microscope image structure analysis using scalograms of orthogonal image transforms is promising
|
