Investigation of material identification with terahertz pulsed imaging

Abstract

Terahertz pulsed imaging provides a wealth of information about the terahertz response of the material under test. We investigate the identification and classification of explosives and common materials based on their terahertz absorption spectra obtained by this spectroscopic imaging technique. In the classification experiment, an accuracy of 96.6% is achieved by using Fisher's linear discriminant analysis for dimensionality reduction of the original spectral data set and support vector machine classifier for classification.