Radar-diffraction tomography using the modified quasi-linear approximation

The authors propose a new approach for solving the scattering problem for tomographic imaging based on the modified quasi-linear approximation (MQLA). Rather than using the incident field to approximate the total field inside the scatterer as the Born approximation (BA) does, the authors utilize the incident field plus a secondary field that equals the incident field multiplied by a scattering coefficient, which is referred to as the quasi-linear approximation (QLA). Generally, this coefficient is source dependent and hence, this approximation cannot be used to solve mutiple-source problems. By invoking the localized approximation, they derive a source-independent scattering coefficient and then obtain the modified quasi-linear approximation (MQLA). Based on this approximation, they developed a new diffraction-tomography algorithm and applied it to crosshole radar tomography. This approach is rapid, since iteration techniques are not involved in the algorithm. Numerical simulations indicate that this tomography algorithm can reconstruct images with better imaging quality than the conventional BA algorithm.