| It is estimated that over 10 million cargo containers and 100 million items of baggage enter the United States each year; inspecting these bags and containers for explosives and other weapons of mass destruction is an enormous challenge. To meet this challenge, the inspection process needs to be optimized so that items are inspected in a cost effective manner while maintaining a satisfactory detection rate. The following research will be presented: 1) A technique for optimizing an inspection process via genetic algorithm. 2) A technique for optimizing an inspection process capable of detecting two types of contraband. 3) A novel improvement to current inspection processes based on building a level of "belief" that an object may or may not hold contraband. This improvement has produced inspection processes which operate at lower costs while having the same detection rate than previously published work. |
