Detecting bombs in hold baggage: 3D imaging is better than 2D imaging when image quality is the same

Abstract

Three-dimensional (3D) imaging is increasingly utilized in hold baggage screening (HBS) at airports. It enables screeners to rotate images and view cross-sectional layers of the bag (slicing). Previous studies on multi-target search in cabin baggage screening indicate that novices benefit from 3D-rotatable images when image quality is the same in 2D and 3D. We investigated the effects of 3D imaging for professional screeners in HBS, where the screening task is to detect bombs in pre-alarmed bags flagged by automated explosives detection systems. In our study, 51 professional screeners completed a highly realistic HBS task involving two levels of bag complexity in 2D and 3D imaging conditions with the same image quality. We found that 3D imaging resulted in higher detection performance in terms of hit rate and sensitivity (d’ and d_a, two sensitivity measures of signal detection theory). Although better bomb detection was accompanied by increased response times, the practical implication of our results is that the transition to 3D HBS substantially enhances human-machine system performance in terms of both effectiveness and efficiency.