An image reconstruction algorithm for high-contrast dielectrics in ECT

Yunjie Yang, Lihui Peng

Department of Automation, Tsinghua University, Beijing, 100084, China

Abstract

Image quality improvement has long been a topic for electrical capacitance tomography (ECT) image reconstruction, especially for applications with high-contrast dielectrics. In such cases, the reconstructions might suffer from serious distortion and blurring, due to the severe nonlinearity. In this paper, we investigated the nonlinearity of different capacitance groups and proposed an enhanced linear model for high-contrast dielectrics image reconstruction. A weighting strategy and error compensation strategy were introduced in the enhanced linear model. Numerical and experimental results with the proposed algorithm were evaluated and compared with those employing the LBP algorithm and the projected Landweber iteration (PLI), which validated the advantage and showed the potential of the proposed algorithm.

Keywords: Electrical capacitance tomography; image reconstruction; high-contrast dielectrics