Image Reconstruction Method Based on Iterative Linear Back Projection and Logistic Forward Solver for Electrical Capacitance Tomography Measurement System

Abstract

Image reconstruction is one of the important tasks in the application of electrical capacitance tomography (ECT) systems. Though numerous algorithms have been implemented, it is often challenging to obtain satisfactory images in all imaging regions by the use of a single algorithm due to the soft-field nature of ECT systems. The preferred iterative reconstruction algorithms are highly computationally inefficient. A new iterative reconstruction algorithm is proposed which combine iterative Linear Back Projection and Logistic Regression. In this method, the solution to the ECT forward problem is implemented using logistic regression, whereas the ECT inverse problem is solved using the algebraic reconstruction technique. By doing so, it is possible to obtain high quality images at relatively efficient computational cost. The simulated experimental results shows that the proposed algorithm outperforms the Projected Landweber and Iterative Linear Back Projection in terms of spatial similarity accuracy, quality of reconstruction images, and computational efficiency. There are improvements of 29 % spatial similarity accuracy and 58 % computational cost relative to Iterative Linear Back Projection algorithm. This is significant improvement toward using ECT system for online industrial operations.