An Iterative Electrical Capacitance Tomography Image Reconstruction Algorithm

Abstract

Conventional electrical capacitance tomography (ECT) image reconstruction relies on the sensitivity matrix of ECT sensor. The linear back-projection algorithm using sensitivity matrix, (LBP_SM for short) and the Landweber Iterative algorithm using sensitivity matrix (I_SM for short) are the most widely used algorithms. An iterative ECT reconstruction algorithm is researched in this paper. Its ECT forward model is obtained by using multivariate linear regression, and its solution of the inverse problem is obtained by iteration, so we call this algorithm I_MLR for short. Capacitance values of a 16-electrode ECT sensor under 10 permittivity distributions are calculated using ECT finite element analysis software. Reconstruction images and reconstruction errors generated by above three algorithms are compared. Comparison results show that I_MLR algorithm can produce much better images than can the LBP_SM algorithm; better images than can the I_SM algorithm in most cases.