Landmine detection using fuzzy sets with GPR images

This paper describes a fuzzy set based approach to the detection of landmines using a novel ground penetrating radar (GPR) imaging system. The GPR produces a three-dimensional array of intensity values, representing a volume below the surface of the ground. Multiple prototypes are generated from fuzzy clustering of gradient features on training data, and a fuzzy confidence is then constructed for the test data from the “object” prototypes. This confidence plane is used to automatically detect objects, which are then scored by the ground truth information. Results on the training and testing with the DARPA backgrounds data set (open fields) and mine lanes (roads) are analyzed