Microwave imaging of high-contrast objects

This thesis describes a novel approach to microwave imaging. The proposed methods are based on the spectral representation of the object. Regularization schemes for the solution of electromagnetic inverse problems are addressed. An iterative algorithm for the solution of the nonlinear least-squares problem arising in iterative image reconstruction is developed. The application of this algorithm to one-dimensional and three-dimensional imaging of high-contrast lossy objects is examined. The properties of the imaging systems designed to implement the algorithms are verified by simulation and actual measurements. The algorithms proposed in this work are computationally more expensive but accuracy and stability are superior compared to other microwave imaging methods.