Di erential Re ection Spectroscopy : A Novel Method for Explosive Detection

Di erential Re ection Spectroscopy: A Novel Method for Explosive Detection S.E. Yuksel, T. Dubroca, R.E. Hummel and P.D. Gader Department of Materials Science and Engineering, University of Florida, Gainesville FL, USA Department of Computer and Information Science and Engineering, University of Florida, Gainesville, USA In the aftermath of the recent terrorist attacks, there has been an increasing need for automated, high-speed detection technologies that can detect trace amounts of explosives without human intervention. Our group at the University of Florida has developed di erential re ection spectroscopy which can detect explosive residue on surfaces such as parcel, cargo and luggage. In this di erential re ection device, explosives show spectral ngerprints at speci c wavelengths, for example, the spectrum of 2,4,6, trinitrotoluene shows an absorption edge at 420 nm. Additionally, we have developed a support vector machine based computer software to classify the explosives and non-explosive materials. In this study we will (i) describe this system and give an insight into the operation of our prototype, (ii) demonstrate our software for the detection of the spectral nger-prints, and (iii) discuss the normalization of the data which signi cantly increases classi cation rates and decreases the number of parameters. DOI: 10.12693/APhysPolA.123.263