Non-invasive material discrimination using spectral x-ray radiography

Current radiographic methods are limited in their ability to determine the presence of nuclear materials in containers or composite objects. A central problem is the inability to distinguish the attenuation pattern of high-density metals from those with a greater thickness of a less dense material. Here, we show that spectrally sensitive detectors can be used to discriminate plutonium from multiple layers of other materials using a single-view radiograph. An inverse algorithm with adaptive regularization is used. The algorithm can determine the presence of plutonium in simulated radiographs with a mass resolution per unit area of at least 0.07 g cm−2.