Three-dimensional inverse scattering approach using analytical singular value decomposition method

Abstract

In this paper, a 3-D electromagnetic inverse scattering algorithm based on analytical singular value decomposition (SVD) of the scattering integral operator is proposed. The spectral expansion of the dyadic Green's function is employed to derive continuous SVD of scattering integral operator. The analytical formulation of singular values and vectors is straightforward without considerable computation cost. Influence of medium properties and receivers polarization on singular values are studied. By virtue of analytical SVD two inversion strategies are proposed. Numerical results are presented to show the fidelity and high efficiency of the proposed algorithm for 3-D real-time imaging. The proposed analytical approach significantly accelerates imaging speed. In addition, better noise control is achievable since small singular values can be easily filtered out.