Electromagnetic Imaging of Closely Spaced Objects Using Matching Pursuit Based Approaches

Abstract

In this letter, sparse signal recovery framework is applied for the reconstruction of two closely placed point-like objects from measured scattered electromagnetic field. Greedy matching pursuit-based algorithms are investigated as the sparsity promoting method. The performances of the matching pursuit algorithms are compared against convex relaxation methods. Orthogonal matching pursuit algorithm implemented with a flexible tree structure is shown to improve the resolution for the inverse scattering problem.