Microwave detection of unknown targets by using Support Vector Machines

Microwave imaging techniques are becoming rather common methods for the detection of unknown targets, since they are in general able to directly provide information on the dielectric properties of the objects under test. However, in several applications only few parameters have to be detected (e.g., position, size, etc.) In this paper, a new approach for the determination of such parameters is presented. The proposed method is based on the use of support vector machines for approximating the unknown relations between the quantities of interest and the measured scattered fields. In this paper, the mathematical formulation of the approach is described and some preliminary numerical results are reported.