Subspace-Rytov Approximation Inversion Method for Inverse Scattering Problems
Tiantian Yin, Li Pan, Xudong Chen
Abstract
In this article, we propose a novel non-iterative method, viz., the subspace-Rytov approximation (SRA) method, to solve inverse scattering problems. This method improves the inversion results of the Rytov approximation (RA) by retaining the integral term neglected therein. The evaluation of this integral involves two approximations. The integrand is approximated with its major part that results from the major induced current and is calculated by the subspace method. The integration on the whole space is approximated by an integration on a finite domain by truncation. Tests with both synthetic data and measured data show that the SRA method outperforms the RA inversion method and the Born approximation (BA) inversion method, and that the SRA outperforms the modified BA (MBA) method in accuracy for relatively large scatterers.