Automated Threat Detection In X-Ray Imagery For Advanced Security Applications
Amit Upreti, Bothra Rajat · HAL (Le Centre pour la Communication Scientifique Directe) · 2021
In areas of high security, like airports, etc, X-Ray machines are used to scan baggage to look for hazardous objects such as guns, knives, razor blades. But, in a manual scan, it is easy to miss some details. Also manual scan is a tedious and time-consuming process. We investigate and compare several algorithms for detection of harmful/hazardous objects like razor blades/handguns etc in X-Ray images of travellers' baggage. Deep Convolutional neural networks such as RCNN, Detectron, RetinaNet and Yolo has shown great results in object detection and recognition. We plan to use the object detection techniques and apply them to improvise upon the already existing Automatic baggage screening methods.