Modeling of high-Z materials detection in assessing brightness/density ratios and their impact on detection accuracy

摘要

Smuggling of radioactive materials and special nuclear materials (SNM) present a concern to national and global nuclear security. Due to weak signals, nuclear materials are not easily detected and numerous techniques were developed addressing issues related to illicit trafficking across the borders. In this paper, we present a numerical model for high-Z radiographic-based detection in cargo scanning for a search of hidden weak-signaling nuclear materials in assessing the effect of brightness ratios on accuracy of this method. This numerical model is attempted to present a new direction in improving the accuracy of cargo inspections by distinguishing the high-Z materials such as shielding materials (always the high-Z), or SNM, from all other materials (that could as well be in the group of high-Z) present in an examined volume of a cargo container. Gamma/X-ray radiography is widely established method to scan the interiors of cargo containers. However, processing of such radiographic images employs only 2-D radiography images. This presents a challenging limitation for an accurate discrimination between high-Z to medium-Z to low-Z materials. To address these shortcomings, we developed a new approach (demonstrated using numerical mirroring of the real scanning of cargo containers) based on two orthogonal radiography images in providing brightness and thickness of an observed volume. An empirical formula using these two variables is then derived to estimate what is a density of a scanned material in the interior of a cargo container. In addition to this empirical formula, a discrimination threshold for high-density materials is suggested in possibly providing new aspects of improvements in real scanning systems.