Multidimensional Data Processing Methods for Material Discrimination Using an Ideal X-ray Spectrometric Photon Counting Detector
Résumé
Development at LETI of pixelated X-ray photon counting detectors based on a CdTe/CZT architecture has shown many improvements in counting abilities, photon energy discrimination, and energy resolution for fast digital imaging modalities. In this context, this study aims at quantifying contribution of photon counting technologies associated with new data processing methods for radiographic material recognition in homeland security. An ideal spectrometric detector was initially simulated to investigate its ability to identify millimetric thicknesses of three homogeneous plastics with similar chemical properties. Data processing methods presented in this study provide spectral data condensed into a set of N coefficients, where N is adjustable by the user. Counting technology performances are compared to an ideal pixelated integrating detector with scintillator architecture. Counting detectors, with data condensed into two coefficients (N = 2) show material identification improvements of more than 50 % compared to integrating detectors. For the counting detector, increasing the number of coefficients implies enhancing material identification. Using N = 90 (one coefficient per channel of 1 keV width) ensures a gain of 80% with respect to scintillator performances. A simulation was also conducted to analyze the effects of the non-ideal response function of the detector on its identification abilities.