Multiphase flow imaging using an adaptive multi-threshold technique in electrical resistance tomography

Abstract

Electrical resistance tomography has been employed as an alternative technique to visualize multiphase flows, because it has the high temporal resolution for monitoring fast transient processes. However, the non-linearity and ill-posedness of the inverse problem cause the poor spatial resolution of the reconstructed images. In this paper, the inverse problem is solved with an adaptive multi-thresholding method (multi-Otsu's method) to improve the reconstruction accuracy and decrease the ill-possedness. A multi-threshold technique is employed to separate the background from the target regions in the resistivity profile estimated on each iteration. The iterative Gauss–Newton method with an adaptive multi-threshold method is applied to the multiphase flow imaging. Numerical simulations and experiments have been carried out to illustrate the reconstruction performance of the proposed method.