Second-order Born approximation imaging of dielectric scatterers via a Markov random field model

A new approach to microwave imaging of 2D inhomogeneous dielectric scatterers is presented. The method is developed in the spatial domain. It uses the second-order Born approximation in order to simplify the original problem. A priori knowledge is used for modeling the dielectric features in the scattering region trough Markov random fields. By using the Bayes theorem, the solution to the equations of the EM scattering is reduced to a search for the minimum of an energy function by means of simulated annealing. Numerical simulations confirming the capabilities and effectiveness of the proposed method are reported and discussed. Results show that solutions are quite accurate, and a notable improvement with respect to a first-order approximation can be obtained