IMAGE-PROCESSING AND PATTERN RECOGNITION OF SCANNING TUNNELING MICROSCOPE DATA.

Scanning Tunneling Microscopy (STM) gives information on the topography, chemical composition and electronic structure of metal and semiconductor surfaces down to the atomic scale. The experimental data can be filtered using a Wiener or least-square filter to eliminate blurring, to suppress noise, and, in addition, to correct spatial distortions by correlating typical picture elements. The application of mask filters in Fourier space, however, can easily falsify processed images.