Inspection of cargo using dual-energy X-ray radiography: A review

Abstract

Cargo inspection is essential for material discrimination and security aims such as threat detection/identification. In this paper, dual-energy X-ray imaging for cargo inspection is studied. Dual-energy imaging challenges such as improving image quality (i.e. noise reduction), automatically removing image defects and errors, and developing algorithms for image detection and identification are investigated. Selecting the right method for image formation, selecting the right source, choosing appropriate X-ray energies, and the appropriate detectors are also discussed. Cargo inspection using dual-energy X-ray shows that not many techniques have been presented for denoising, and there is no comparative comparison between them. Material discrimination using dual-energy X-ray has had significant success in cargo, especially when convolutional neural network (CNN) is integrated with traditional techniques. However, it will have limitations on performance due to the existence of challenges such as the need for large data and the possibility of noise presence in the extracted features. It seems that future studies should cover these issues.