A Statistical FDFD Simulator for the Generation of Labeled Training Data Sets in the Context of Humanitarian Demining using GPR

Due to years of warfare in Colombia, humanitarian demining is more important than ever. However, practice has shown that the detection of Improvised Explosive Devices is challenging and represents a tedious task, due to the lack of reliable sensors. However, developments in the field of Ground Penetrating Radar and machine learning are opening new possibilities. On the downside, supervised machine learning classifiers require large amounts of training data that are not available until today. Therefore, we introduce a 2D-simulator for the generation of randomly behaving ground scenes, which is capable of generating training data sets in a short amount of time.