Field tests of a NaI(T1)-based vehicle portal monitor at border crossings

Noise-Adjusted Principal Component Analysis for Buried Radioactive Target Detection and Classification

“…T HE detection and discrimination of radioactive objects has many important applications, such as illicit cargo detection at border crossings [1], [2], buried target detection within battlefields [3], and nuclear threat discrimination from benign sources [4]. Several approaches have been developed [4]- [6].…”

The use of energy windowing to discriminate SNM from NORM in radiation portal monitors

“…Moreover, the total energy deposited in PVT is not necessarily directly proportional to the original gamma-ray energy, but is spread over a broad energy range (Compton continuum). This is opposed to detector materials for which photo-absorption is significant, such as thallium-doped sodium iodide [NaI(Tl)] crystals, where all of the energy of the incident gamma ray is often deposited as an easily discernable full-energy peak [11,12]. It is this full-energy peak that is directly proportional to the incident energy and, along with the information in the full spectrum, can be used for spectroscopy.…”

Smoothing Low Resolution Gamma Spectra

“…For example, deployed sensors at locations that screen for illicit nuclear material rely on isotope identification to resolve innocent nuisance alarms arising from naturally occurring radioactive material [1].…”

“…Studies comparing the energy spectra measured by low, medium, and high resolution detectors suggest that relatively low resolution detectors, such as Sodium Iodide (NaI) detectors that typically count the number of gamma photons in 512 or 1024 calibrated energy channels will continue to play a key role in isotope identification [1], [2]. However, RIID performance from the raw counts collected over short time intervals is strongly impacted by the intrinsic variations from Poisson count statistics.…”

“…In that case, it is not clear that a long (1 hour or more) count can be regarded as the "true" count rate as measured by the detector [1], [2]. Smoothing such noisy data is still a viable option, but performance assessment would have to be modified compared to that presented below where we assume we know the true count rate in each case.…”