X-ray security screening stations are used in various fields. It takes a lot of time and training for security guards to detect storage devices exported from some manufacturing industries (semiconductors, displays, secondary batteries, automobiles, re...
X-ray security screening stations are used in various fields. It takes a lot of time and training for security guards to detect storage devices exported from some manufacturing industries (semiconductors, displays, secondary batteries, automobiles, research labs). In this paper, we aim to prevent the leakage of key information in major industries and contribute to national security through accurate detection of storage devices. The first step is to build a dataset specialized for storage device detection using the storage medium of the hazardous goods image set and sample data from the client. To find a suitable model for detecting storage devices, various deep learning-based object detection models were investigated The dataset was divided into an 8:2 ratio and applied for learning and testing. Finally, the mAP of Faster R-CNN, Xception and YOLOv5 was 84.33%, 91.84%, 95.94%, respectively, keyword : X-ray, Faster R-CNN, Xception, YOLOv5 * A thesis for the degree of Master in February 2024.