Comparison of material discrimination curves obtained from Log-Log, R-T, r-θ, and α2-α1 algorithms in a dual-energy radiography system

Dual-energy X-ray radiography is an important inspection technology in imaging large cargos for finding illegal materials such as explosives, narcotics, weapons, etc. A useful aspect of dual-energy X-ray radiography is the possibility of discriminating and identifying materials with different atomic numbers which can be obtained by imaging large cargos at two different X-ray energies (normally above 3 MeV). In the present work, first a dual-energy X-ray radiography system including X-ray sources, collimator, and detector array was simulated using MCNPX code. Then, the obtained data from this radiography system for four step wedges of graphite, aluminum, steel, and lead materials were gathered and material discrimination curves were obtained using four algorithms including log-log, R-T, r-θ, and α2-α1. The results show that among the proposed algorithms for material discrimination, R-T, r-θ, and α2-α1 algorithms have better performance and due to the less overlap of material discrimination curves for low sample thicknesses, R-T algorithm can be a better choice for obtaining material discrimination curves in dual-energy radiography systems.