Image reconstruction for X-ray computed tomography in security screening applications
Abstract
X-ray tomographic systems have increasingly widespread use in security screening applications. For example, most major airports now utilize X-ray CT systems for efficient screening of baggage and cargo. While decades of research has benefited CT for medical diagnostics, a number of practical differences in the security application present a new set of challenges for the reconstruction problem. For example, the size and composition of the scan subjects, the throughput requirements, and the measure of image quality are all factors that lead to a different set of design considerations. This thesis investigates the application of model-based iterative reconstruction (MBIR) methods for X-ray CT systems in the security context. This reconstruction approach is demonstrated to produce high quality images that are much less susceptible to image distortion compared to direct reconstructions using direct Fourier methods or filtered back projection. Presented first is a mapping of a fully 3D MBIR algorithm to a multi-slice helical-scan CT system certified for baggage screening. Model enhancements for the system are discussed, and MBIR reconstructions are presented alongside direct reconstructions for a set of real baggage scans. Also investigated is the performance of model-based methods for reconstruction on sparse-view angle security CT systems. Advances in the forward and prior modeling are demonstrated to improve several aspects of the reconstructed images, including resolution, artifact suppression, and CT accuracy.
Degree
Ph.D.
Advisors
Bouman, Purdue University.
Subject Area
Electrical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.