Image reconstruction of electrical capacitance tomography using the dynamic mesh and changing physical properties methods

Abstract

In this research, two new algorithms are introduced for image reconstruction of electrical capacitance tomography (ECT): the dynamic mesh (DM) and changing physical properties (CPP) algorithms. The presented algorithms estimate the boundary of an unknown object inside a circular domain. To capture the complex geometries, an unstructured mesh in conjunction with a finite volume method is employed. Several test cases are investigated to show the capability and reliability of these methods. The results show that the introduced methods need fewer measurements for accurate image reconstruction in comparison with previous methods. The number of iterations required for shape estimation and the accuracy of the DM method are better than the CPP method. Furthermore, the sensitivity analysis results show that by increasing the uncertainty of the measurements, the estimated shape has a slight deviation from the exact shape. The accuracy and speed of the introduced algorithms represent the new applicable algorithms for ECT image reconstruction.