Two Dimention Numerical Reconstruction of Electrical Impedance Tomography using Tikhonov Regularization Algorithms with a-posteriori parameter choice rule

Abstract

Electrical Impedance Tomography, as an Inverse Problem, is calculation of the resistivity distribution due to given boundary potential and current density distribution. Most of the inverse problem are ill-posed, since the measurement data are limited and imperfect. This paper describes a regularization technique for solving the ill-posed problem appeared in the inverse EIT. In this regularization technique, a smoothing function with a regularization parameter, is penalizing the objective function in order to obtain a regularized resistivity update equation. The regularization parameter can be chosen from a-posteriori information. We made comparison of 3 methods, the rst method can be thought of as a discrepancy principle, where we select an initial value of the regularization parameter by trial and error technique. The second and third methods are methods adopted from Linear ill-posed problem, with a posteriori information characters. We presents numerically the reconstruction using arti cially generated data.