Quantitative methods aim at giving the full characterization of the targets, both in term of morphology and electromagnetic properties. They are usually grouped into two classes, linear and nonlinear optimization methods. The former methods exploit approximations of the scattering phenomenon to linearize the data-to-unknown relationship and are typically capable of providing only a rough description of the target’s morphology when exploited outside of the range of validity of the underlying approximation. Conversely, nonlinear methods, that for instance include modified gradient method and the contrast source inversion method, tackle the inverse scattering problem in its full nonlinearity. All these approaches seek for the problem’s solution by means of a local iterative optimization scheme, as the large number of unknowns makes global optimization generally not viable, due to the exponential growth of the computational complexity.

Click on the pertaining box to find a list of the innovative linear and nonlinear methods that the LEMMA’s researchers have introduced.

non linear linear