Quantitative Microwave Imaging Methods for Dielectric Profiling

Abstract

newlineMicrowave Imaging (MWI), is an emerging modality that is being used as a diagnostic newlinetool in various applications such as ground-penetrating radar, synthetic aperture radar, newlinetarget identification, medical imaging etc. due to its unique features. Some of these newlineapplications require only qualitative information about the object under test. In newlinecontrast, many of applications needs a quantitative description of the scatterer, such as newlineits permittivity and conductivity profiles. The process of retrieving such features from newlinethe measured field values is known as Inverse Scattering Problem (ISP). The inverse newlinescattering problem (ISP) of microwave imaging is highly non-linear and ill-posed and newlineis liable to get stuck in local minima. The iterative techniques employed for solving newlinethese non-linear and ill-posed problems are computationally expensive. An attempt is newlinemade in this thesis to reconstruct the image cross-section of the object by determining newlinethe distribution of dielectric permittivities, from the scattered data measured around the newlineobject. Development of these methods begins with the formulation of the ISP by using newlineElectric Field Integral Equations (EFIE). The above modeled system of equations are newlinediscretized using Method of Moments (MoM) and a cost function is formulated in newlineterms of the field values. This cost function is solved in an iterative manner using Born newlineIterative Method (BIM) or Distorted Born Iterative Method (DBIM).