Electromagnetic inverse scattering is a problem including of detecting, locating, shape finding, and imaging of the unknown objects by using scattered fields and Maxwell’s equations. There are different methods for solving inverse scattering problems which one may group them under qualitative and quantitative titles. In this thesis, modified suace optimization method (SOM) with sequential quadratic programming (SQP) algorithm, called SQP-SOM, for inverse scattering of dielectric, conducting and mixed objects which are located in different mediums such as free space, inhomogeneous background and stratified media, is proposed as an accurate and fast quantitative reconstruction method, and the effect of nonradiating (NR) current induced on the scatterer, onto renovation process is evaluated. For this, the direct scattering problem for each case is formulated by a proper numerical method such as coupled dipole method (CDM), method of moment (MoM), finite elements-boundary integral (FE-BI) and electric field integral equations (EFIE), and the corresponding external mapping operator has been spectral analyzed by singular value decomposition (SVD) technique to determine the radiating and NR singular vectors. For expanding the induced current space, the coefficients of the NR singular vectors have been dependent on the unknown components of the scatterers with least squares procedure. This analysis is presented for mixture of dielectric and conducting scatterers, and the related equations have been extracted for the first time. For reconstructing the unknown scatterers, the objective function has been constructed based on induced current residue and data scattering mismatch and minimized by SQP. Also, to evaluate the effect of NR current for imaging, the radiating objective function is defined by eliminating of NR part of the induced current. The reconstructed results from SQP-SOM are successful for detecting, locating, shape finding and imaging levels. Comparison with the results of radiating objective function indicates that NR current reconstruction is necessary for accurate imaging of scatterers. Keywords Inverse Scattering, nonradiating (NR) current, SQP-SOM, singular value decomposition (SVD), FE-BI, stratified media, radiating objective function
