Reconstruction algorithm for object determination from ultra-wideband field data

An inverse obstacle scattering method is presented for reconstructing the shape and effective constituent material parameters of a buried dielectric object from measured scattered field data collected above the surface of an air/ground interface. A forward model is developed based upon the linearised Born approximation and the dyadic Green's function for a single homogeneous medium. The forward model is filtered for noise using the Tikhonov regularization operator and then inverted and back propagated. Experimental results are provided to illustrate the steps involved and the performance of the inverse scheme to extract constituent material parameters.