An inverse electromagnetic algorithm for aspect-limited data

A polarization/depolarization approach is taken to improve the algorithms for extracting shape information from electromagnetic radiation scattered of a body. A numerical method is presented which determines the shape from a sparse data set as a function of aspect. The technique is based on the fact that principal radii of curvature local to specular points of a perfectly conducting convex body have measurable radar signatures. It is shown that a variational solution of the Stoker's differential equation approach to inverse scattering is not constrained to full aspect data, thereby making the shapes of convex portions of a body identifiable by radar scans. Tests have yet to be performed on real world objects, although the simulations indicate the technique is valid.