To the Problem of Recognition the Results of Electron Spectroscopy of Nanostructures with the Use of Discrete Wavelet-Transformation

摘要

The article presents application of discrete wavelet transform (Haar wavelet) for the spectra of recognition of two samples - silicon and silicon compound. The method presented by the authors uses the data obtained as a result of the experiment on a scanning electron microscope ZEISS SIGMA SEM HD. The visual comparison of the wavelet coefficients obtained is laborious, so the problem arises of automating the process of comparison and recognition, thus the results of the discrete wavelet transformation are the statistical characteristics-the arithmetic mean, the variance and the standard deviation.