Identification of Materials in Fragments of Large-Sized Objects in Containers by the Dual-Energy Method

Abstract

The mathematical and simulation models of image formation in a system for the recognition of materials of internal fragments of containers by the dual-energy method are developed. A series of computational experiments were carried out on the recognition of materials inside shipping containers without and with compensation for the influence of the factor being analyzed on the quality of identification. The thickness of the prefilter, the ADC digit capacity, the ratio of the numbers of low- and high-energy bremsstrahlung pulses, and the dimensions of the averaging filter window were varied. An algorithm has been developed to compensate for the influence of the container wall thickness on the identification quality. The algorithm is based on the statistical processing of radiographic images of the object. Its effectiveness has been experimentally proved.