Research Article
PIDNet: Prohibited Items Detection Network and Fine-Coarse Encoder Module
@INPROCEEDINGS{10.1007/978-3-031-65123-6_20, author={Yu Yao and Boliang Zhang and H. K. Kan and Chan Tong Lam}, title={PIDNet: Prohibited Items Detection Network and Fine-Coarse Encoder Module}, proceedings={Quality, Reliability, Security and Robustness in Heterogeneous Systems. 19th EAI International Conference, QShine 2023, Shenzhen, China, October 8 -- 9, 2023, Proceedings, Part II}, proceedings_a={QSHINE PART 2}, year={2024}, month={8}, keywords={Prohibited Items Environmental Security Security Inspection X-ray Images}, doi={10.1007/978-3-031-65123-6_20} }- Yu Yao
Boliang Zhang
H. K. Kan
Chan Tong Lam
Year: 2024
PIDNet: Prohibited Items Detection Network and Fine-Coarse Encoder Module
QSHINE PART 2
Springer
DOI: 10.1007/978-3-031-65123-6_20
Abstract
Security inspection using X-rays are absolutely familiar in everyday life and have an essential function in protecting public safety. However, it is not straightforward to perceive the presence of prohibited items, and the key challenge is that any prohibited items in X-ray images may exhibit color-monotonous and luster-insufficient, mainly due to the characteristics of X-ray imaging mechanisms. In this paper, to address this problem, we constructed a fresh prohibited items detection dataset (PIDD) and proposed a prohibited items detection network (PIDNet), which searches enrichment fine-grained and coarse-grained features for powerful prohibited items detection with a novel Fine-Coarse Encoder (FCE) module. Extensive experiment demonstrates that our proposed method achieves significantly superior contraband detection results on the PIDD test set compared to progressive methods for prohibited items detection, effectively proving the practicability of the method proposed in this paper.
