Threat object detection and analysis for explosive ordnance disposal robot

Explosive Ordnance Disposal (EOD) robots are useful in military applications like the safe disposal of explosives. However, many of these robots do not have the capability to identify threat objects using their onboard vision system due to data unavailability for training an improvised explosive device (IED) detector. As a solution, this study used image processing and object detection algorithms to detect and analyze threat objects inside the baggage. A threat object detector was developed and composed of two separate modules such as baggage detection and IED detection and analysis modules. The experiments showed that baggage detection achieved 22.82% mean average precision (mAP) using Single Shot Detector (SSD) in the Microsoft Common Objects in Context (COCO) dataset, while IED detection achieved 77.59% mAP using Faster R-CNN in the X-ray dataset. The threat objects from the X-ray image were also analyzed using image processing techniques to get the dimension of the object and the distance from a reference object. Also, the baggage detection module was successfully deployed in Jetson TX2, which runs at a frame rate of 12 frames per second (FPS).