Microwave imaging within the interval analysis framework

Abstract

An approach based on the use of the arithmetic of intervals and Interval Analysis for the solution of inverse scattering problems is presented and assessed. By exploiting the property of the Interval Analysis to find the global minimum of a functional in a n-dimensional space, the proposed approach adopts a branch and bound process to discard the regions of the solutions space not containing the global solution, while keeping those where a feasible solution is expected until a suitable converge criterion is reached. A representative set of results concerned with the reconstruction of circular dielectric objects within the first-order Born approximation are reported and discussed to show potentialities and current limitations of the proposed approach.