A model based recursive inverse scattering algorithm for microwave imaging

An iterative inverse scattering method is introduced. This approach decouples the nonlinear effects due to diffraction by using an integral equation formulation to predict the measured scattered field and a differential equation formulation to describe the equivalent source field. The equivalent sources are evaluated by the finite element model to achieve computational efficiency. A Bayesian estimation strategy combined with an orthogonal transform of the object function is applied to solve the ill-posed inverse problem. The orthogonal transform provides a natural constraint on the solution space. Consequently, the a priori information required in Bayesian estimation can be determined appropriately. The computational simulation demonstrates the efficiency of this method for very limited measurements.