Hyperspectral Imaging based Characterization and Identification of Sintered UO2 Fuel Pellets

Hyperspectral Imaging (HSI) is a well-established technology able to capture the same spatial scene or image in hundreds of different wavelengths across the electromagnetic spectrum, covering not only the visible but also the shortwave infrared range (1000-2500nm), potentially revealing information otherwise invisible to the human eye. In this work, we explored whether HSI technology can be used for the fast-non-invasive characterization and identification of sintered UO 2 fuel pellets. Preliminary experiments included the imaging of the pellets, revealing their spectral responses. These show promising features that could be used for their identification, where two different groups of pellets, pure and doped, seem to be easily recognized based on their spectral response. The experiments included a pixel-wise classification map generated via the Spectral Angle Mapper (SAM) technique in which the pure and doped pellets are successfully identified.