On the Impact of Object and Sub-component Level Segmentation Strategies for Supervised Anomaly Detection within X-ray Security Imagery

TL;DR

This paper evaluates how different segmentation strategies at object and sub-component levels affect supervised anomaly detection in cluttered X-ray security images, finding sub-component segmentation slightly improves detection accuracy.

Contribution

It introduces and compares segmentation strategies for anomaly detection in X-ray imagery using CNNs, highlighting the benefits of sub-component level segmentation.

Findings

Sub-component segmentation yields ~98% true positive rate.

False positive rate is approximately 3%.

Sub-component segmentation performs marginally better than object-level segmentation.

Abstract

X-ray security screening is in widespread use to maintain transportation security against a wide range of potential threat profiles. Of particular interest is the recent focus on the use of automated screening approaches, including the potential anomaly detection as a methodology for concealment detection within complex electronic items. Here we address this problem considering varying segmentation strategies to enable the use of both object level and sub-component level anomaly detection via the use of secondary convolutional neural network (CNN) architectures. Relative performance is evaluated over an extensive dataset of exemplar cluttered X-ray imagery, with a focus on consumer electronics items. We find that sub-component level segmentation produces marginally superior performance in the secondary anomaly detection via classification stage, with true positive of ~98% of anomalies, with a ~3% false positive.