Localisation and Detection of Buried Objects like Mines in Diverse Soils with Softcomputing Methods

Originally, mines were detectable because of their more or less high metal content. Unfortunately, in the fast developing mine technologies, new materials that may last for many years, have made it possible to produce varieties of mines that are practically undetectable with the existing methods because of very low metal contents. As we showed, this fact can be caught with special pre-processing and classification structures like hybrid neural nets also if non destructive systems as common metal detectors are used. Also the existence of large numbers of landmines poses a several problems. Here also adaptive learning techniques, including neural nets can be considered as a possible solution, as they extract features from the ground bounce-removed responses and input this feature set automatically and can be empowered in field for new classification concepts. The hybrid classification structure we developed and tested consists of self organizing memories (SOMs) and Backpropagation nets combined with the interneural WHU-Structures, whereby special pre-processing methods like DLS and fitting-methods are used to extract the feature specific items.