Fast Multiparametric Electromagnetic Full-Wave Inversion via Solving Contracting Scattering Data Equations Optimized by the 3-D MRF Model

Abstract

This article presents a novel hybrid electromagnetic full-wave inversion that combines the traditional inversion method, i.e., the variational Born iterative method (VBIM), with the 3-D Markov random field (MRF) model. In each iteration, VBIM first reconstructs the model parameters of all discretized cells in the inversion domain by solving the discretized data equations. Then, MRF is adopted to classify the cells according to the reconstructed parameter values, i.e. to determine which homogenous scatterer a certain cell is belonging to. Finally, partial cells classified as “background” are removed, and partial cells classified as “scatterer” are merged. Consequently, the discretized data equations gradually contract, and the unknowns in the following VBIM iterations are reduced. Numerical experiments show that compared with the traditional VBIM, the proposed hybrid VBIM-MRF model can achieve higher reconstruction accuracy for multiple model parameters. Furthermore, the computational cost is also significantly reduced.