The impact of color composition on X-ray image interpretation in aviation security screening

Abstract

In order to improve aviation security, many airports
apply Threat Image Projection (TIP) and computer-based X-ray
image interpretation training (CBT). One difference between TIP and CBT X-ray images is the algorithm used to merge virtual threat items into X-ray images of passenger bags, resulting in different color nuances. In this study, we tested the influence of merging algorithms on threat object detection performance, reaction time and confidence rating of 12 airport security screeners. The image merging algorithms of the individually
adaptive CBT X-Ray Tutor were used. We have found that the
hit rate was higher for TIP images compared to CBT images.
Accordingly, the mean of confidence ratings was increased for TIP images. The reaction times tended to be shorter for TIP images. The results of our study indicate that the CBT merging algorithm used in this study is more realistic than the tested TIP image merging algorithm.