Automated explosive detection systems based upon CT technology

摘要

CT has established itself as superior technology over conventional X-rays initially in medical imaging and now for explosive detection system (EDS) applications. What is true for anatomical structures of the human body is also true for baggage: cross-sectional images improve greatly the detection probability of specific structures. The challenges of transitioning CT from medical to EDS applications include the development of a new market as well as technical, environmental and operational issues. This paper explains automatic detection, which has not been done for medical applications yet. The approach taken went through feasibility study, supporting the definition of EDS standards, prototype development, and extensive airport trials. Threat detection required new algorithms to be invented which make use of the density and other topological information provided by CT. Today, operator intervention has been reduced to a small percentage of the luggage to make a decision on a threat. More automation is desirable in principle, and the current technical limits to detection and false alarm rates are discussed.