We present a method for the automated detection of firearms in cargo x-ray images using RetinaNet. RetinaNetis a recently proposed powerful object detection framework that is shown to surpass the detection performanceof state-of-art two-stage R-CNN family object detectors while matching the speed of one-stage object detectionalgorithms. We trained our models from scratch by generating training data with threat image projection (TIP)that alleviates the class imbalance problem inherent to the x-ray security inspection and eliminates the need forcostly and tedious staged data collection. The method is tested on unseen weapons that are also injected intounseen cargo images using TIP. Variations in cargo content and background clutter is considered in trainingand testing datasets. We demonstrated RetinaNet-based firearm detection model matches the detectionaccuracies of traditional sliding-windows convolutional neural net firearm detectors while offering moreprecise object localization, and significantly faster detection speed.
展开▼
