Abstract
Purpose - A sensitivity matrix is typically a crucial part of a deterministic image reconstruction process in three-dimensional capacitance tomography (3D ECT) and strictly decides about a final image quality. Commonly used sensitivity matrix computation methods mostly provide acceptable results and additionally allow to perform a recalculation of sensitivity maps according to the changing permittivity distribution. In this paper a significant modification of the sensitivity maps calculation process is introduced using electric field distribution analysis. Design/methodology/approach - The new "tunnel based" algorithm is proposed which traces the surfaces constructed along the electric field lines. The new solution is developed and tested using experimental data.Findings - To fully validate new technique both: linear and non-linear image reconstruction process was performed and criteria of image error estimation were discussed. This paper discusses some preliminary results of the image reconstruction process using the new proposed algorithm. As a result of this research an increased accuracy of the new method is proved.Originality/value - The presented results of image reconstruction with new sensitivity matrix in comparison with the classic matrix proved that the new solution is able to improve the convergence and stability of image reconstruction process for 3D ECT imaging.
Purpose - A sensitivity matrix is typically a crucial part of a deterministic image reconstruction process in three-dimensional capacitance tomography (3D ECT) and strictly decides about a final image quality. Commonly used sensitivity matrix computation methods mostly provide acceptable results and additionally allow to perform a recalculation of sensitivity maps according to the changing permittivity distribution. In this paper a significant modification of the sensitivity maps calculation process is introduced using electric field distribution analysis. Design/methodology/approach - The new "tunnel based" algorithm is proposed which traces the surfaces constructed along the electric field lines. The new solution is developed and tested using experimental data.Findings - To fully validate new technique both: linear and non-linear image reconstruction process was performed and criteria of image error estimation were discussed. This paper discusses some preliminary results of the image reconstruction process using the new proposed algorithm. As a result of this research an increased accuracy of the new method is proved.Originality/value - The presented results of image reconstruction with new sensitivity matrix in comparison with the classic matrix proved that the new solution is able to improve the convergence and stability of image reconstruction process for 3D ECT imaging.