L1-L2 Spatial Adaptive Regularization Method for Electrical Tomography

Regularization is an effective method for the ill-posed problem of inverse problems in electrical tomography(ET). Traditional regularization methods utilize fixed regularization terms which neglect the spatial characteristics of the field. In this paper, a novel regularization method was proposed which combines selection of the regularization terms with spatial information such as electrical parameters that obtained from iterative result of each step. L1 or L2 norm is chosen as the regularization term according to the distribution of electrical properties. The simulation results have verified that this proposed method can effectively improve the imaging resolution and enhance the noise immunity of the inverse problem compared with the traditional regularization methods such as L2 regularization and L1 regularization.