Machine Learning Based Prohibited Item Detection in Airport X-Ray Images Using YOLOv11

Detecting and identifying prohibited items in X-ray images at airports is essential for ensuring the safety of both the airport and its travelers. Therefore, it is important to have an accurate detection system that can identify these objects easily, utilizing X-ray images. The aim of this paper is to leverage the YOLOv11 machine learning model to accurately detect and classify prohibited items in X-ray images. The model was trained on an augmented dataset that includes four main object categories: blade, gun, knife, and shuriken. To evaluate the model’s effectiveness, metrics like precision, recall, mean average precision (MAP), and confusion matrices were employed. The findings reveal strong detection performance, achieving a mean average precision of 0.863 across all categories. These findings suggest the proposed YOLOv11-based approach is effective for real-time security screening applications.