Ellrich, F.F.EllrichMolter, D.D.MolterJonuscheit, J.J.JonuscheitFreymann, G. vonG. vonFreymannBeigang, R.R.BeigangPlatte, F.F.PlatteNalpantidis, K.K.NalpantidisSprenger, T.T.SprengerHübsch, D.D.Hübsch2022-03-122022-03-122014https://publica.fraunhofer.de/handle/publica/3868722-s2.0-849068717712-s2.0-84944691731Combining different chemometric tools we demonstrate the reliable identification of substances out of real-world terahertz-spectra. Systematic evaluation of the spectra allows for the distinction between spectral properties arising from samples geometry versus true material absorption.enpattern recognitionterahertz spectroscopyterahertz wave spectra621Multi-step pattern-recognition: A powerful tool for substance identification based on real-world terahertz-spectraconference paper