Spatially Varying Regularization for Microwave Imaging

This paper presents a new spatial-based regularization technique to efficiently reconstruct unknown scatterers using microwave imaging. In general, a constant regularization parameter is used to obtain an inverse solution. However, the major drawback of using this parameter is that it produces ghost images, often known as artifacts. To overcome this problem, a spatially varying regularization parameter is developed in this work. The main contribution of this approach is that it removes unwanted artifacts and provides uniform reconstruction over the entire investigation domain. This results in improved reconstruction quality. The effectiveness of this approach is demonstrated using several numerical examples utilizing both synthetic and experimental data. In all the cases, the proposed method is able to provide an optimal solution free of speckle artifacts. Consequently, better reconstruction results are obtained compared to a standard Tikhonov method with constant regularization parameter.