A Combined Iterative Data-Driven Inversion Strategy for Microwave Imaging

An iterative inversion strategy for microwave imaging is presented in this paper. The method, which combines a data-driven approach and a Newton-based technique, aims at improving reconstruction results by exploiting an approximate prior knowledge of some features of the region under test. In particular, an additional regularization term is introduced inside the solving procedure. This term includes a surrogate model, which is constructed by considering a set of numerical simulations based on the available prior information. The behavior and the performance of the proposed inversion method is assessed for the reconstruction of dielectric targets in a numerical test case, involving a host structure with an inclusion.