An ECT image reconstruction algorithm based on object-and-background adaptive regularization

Abstract

Electrical capacitance tomography (ECT) image reconstruction is an ill-posed inverse problem. Regularization methods are generally employed to solve the ill-posed problem, and the reconstructed image quality is seriously influenced by the selection of the regularization parameter. At present, a same regularization parameter is generally adopted for the whole reconstruction region, but the object region and the background region have different requirements on the regularization parameter, so it is difficult to obtain high-quality reconstruction images. In order to solve this problem, an image reconstruction algorithm based on object-and-background adaptive regularization, the OABAR algorithm for short, is proposed in this paper. The basic idea of the proposed OABAR algorithm is to dynamically divide the reconstructed region into object and background regions according to the reconstructed gray values. This is followed by iteratively reducing the regularization parameter values corresponding to the object region on the one hand to improve the detail reconstruction ability of the object region. On the other hand, a fixed regularization parameter value larger than the initial regularization parameter is provided for the background region to improve the smoothing ability of the background region. The iterative process is terminated when the change trend of capacitance residual norm first reverses or the number of iterations reaches the preset maximum value. If the change trend of capacitance residual norm does not reverse within the preset maximum number of iterations, the reconstructed result with the minimum capacitance residual norm is taken as the final result. Simulation and experimental tests were carried out and the results verify the effectiveness of the proposed algorithm on improving ECT image reconstruction quality.