A Multi-Frequency Quantitative Inversion Technique for Imaging through Obscuring Media

We develop a multifrequency, nonlinear quantitative inversion technique with applications to radar imaging. In particular, we perform a novel extension of a contrast source inversion algorithm, called subspace based optimization method (SOM), to its multi-frequency counterpart, called multi-frequency SOM (MFSOM). With a small angular aperture, MFSOM can range-resolve objects by introducing frequency diversity into SOM. We show that MFSOM is able to reconstruct a scatterer with improved accuracy when compared to the filtered backprojection approach. A limitation of typical radar imaging algorithms, in applications such as imaging through obscuring media, is that the range focusing of obscured scatterers is inaccurate owing to range ambiguities induced by the unknown dielectric properties of the media. Our new technique, MFSOM, is able to solve the range ambiguity problem of obscured targets by using an accurate estimate of the relative permittivity, to find the true range from the radar to a scatterer in the scene.