Digital processing of lattice images from a diffraction spot selected in diffractograms of electron micrographs

A method of noise removal using the inverse transform of an image autocorrelation function

“…where G,k) represent equidistant grid points in the digital image, and MxN is the total number of picture elements. In this work, the digital image of an electron micrograph was obtained by the scanning densitometer system with a microprocessor (Kanaya et al, 1982a). In practical computation of the image autocorrelation, the FFT (Fast Fourier Transform) algorithm is used because it makes the computation much faster than the direct evaluation of equation 1 (Frank, 1980).…”

“…For a case of periodic objects, the filtering method, masking all Fourier spectra except those lying on certain spectrum regions such as diffraction spots and layer lines, etc. (see DeRosier and Klug, 1968;Frank, 1973;Hawkes, 1980;Baumeister and Vogell, 1981;Kanaya et al, 1982aKanaya et al, ,b, 1983Kanaya et al, , 1984, is useful, and it results in a considerable improvement of the signalto-noise ratio. The photographic superposition technique (Markham method), or the periodic signal averaging method (Kirkland et al, 19811, is also effective, and it may be further utilized for a case of objects with rotational symmetries (Crewe and Ohtsuki, 1982).…”

Section IV: Digital processing methods for structural analysis in high resolution electron microscopy

“…The optical transfer theory on the optical diffractograms of electron micrographs was originally proposed by Thon (1966) and subsequently developed by Klug and DeRosier (1966). Later, digital image processing techniques were established (Baumeister and Vogell, 1980;Hawkes, 1980;Frank, 1973;Kanaya et al, 1981Kanaya et al, , 1982aKanaya et al, , 1982b.…”