Improving Security X-Ray Screening: Jig-Based Analysis for Enhanced Material Differentiation and Pseudocoloring

In security X-ray systems, traditional dual-energy X-ray absorptiometry methods are hindered by color distortion and inadequate material boundary detection due to a lack of thickness information. Meanwhile, AI-based pseudocoloring methods encounter computational challenges and issues with misclassification. We propose a novel approach that emphasizes the estimation of material thickness in conjunction with property classification. By constructing a property-based database using custom-fabricated jigs and calculating attenuation coefficients (R-values) from high- and low-energy X-rays, our method assigns pseudocolors with brightness gradations that accurately reflect material thickness. Experimental results obtained from samples compliant with ASTM-F792 and ECAC STP standards demonstrate the effectiveness of the proposed method in differentiating material properties and enhancing the visualization of high-density materials. These findings highlight the potential for improving the efficiency and reliability of real-time cargo screening systems.