Novel Principle for Data Processing from Hand-Held Ground Penetrating Radars

摘要

The goal of this project was to develop a new algorithm of inverse problems to image locations and electrical properties of land mines using GPR data. These parameters, in turn, might be used on a next stage as an input for the classification procedure. Such a method has been developed and tested on models. Accurate images of mine-like targets are demonstrated. Comparison with the best competing algorithm in the field favors the algorithm of this project. The main innovation, as compared with conventional algorithms is the differential rather than conventional integral form of the resulting system. This enables one to obtain solutions in a much shorter time, as compared with the conventional techniques.