Detection of perturbations in thermal IR signatures: an inverse problem for buried land mine detection

Abstract

The analysis of perturbations on the thermal signature of the soil is a powerful tool for the detection of the presence of buried objects on the soil from measured infrared images but, by itself, gives little insight in the nature of the detected targets. In this paper, we will present a method for the detection of surface and shallowly buried land mines in infrared images based on a 3D thermal model of the soil. This model will be used to detect perturbations on the expected behavior that will lead to the assumption of the presence of unknown buried objects. Next, we will outline a procedure that makes use of the theory on inverse problems in order to extract information of the nature of the detected targets and to infer whether they actually correspond to land mines or not.