Sensitivity map based sub-pixel image reconstruction by using simulation data for capacitive wire-mesh tomography

Weimin Lao 1*, Yi Li 2 and Lihui Peng 1

1 Tsinghua National Laboratory for Information Science and Technology, Department of Automation, Tsinghua University (Beijing, China)

2 Graduate School at Shenzhen, Tsinghua University (Shenzhen, China)

*laowm14@mails.tsinghua.edu.cn; liyi@sz.tsinghua.edu.cn; lihuipeng@tsinghua.edu.cn

ABSTRACT

Wire-mesh tomography is a visualization and measurement technique for multiphase flow by using direct imaging. There are basically two types of wire-mesh sensors, which are conductivity sensor and capacitance sensor. Both sensors have similar structure and measuring principle. Conductivity sensor is used for two-phase flows that have quite different conductivity, while capacitance sensor is used for flows different in permittivity. Recent investigation shows that it is possible to use wire-mesh sensor to measure oil-gas-water three-phase flow.

As an instantaneous tomography method, wire-mesh tomography has an advantage in speed but has less image resolution because classic wire-mesh tomography sensor only provides the same amount of pixels as measurement number. In order to increase image resolution, an image reconstruction method based on sensitivity map is investigated, which is of providing more pixels (sub pixels) by solving inverse problem with capacitive wire-mesh tomography image reconstruction. The sensitivity map of capacitive wire-mesh sensor is calculated by using the finite element method, which will be used for image reconstruction according to the requirement with different image resolutions. The mathematical model based on sensitivity matrix for capacitive wire-mesh tomography image reconstruction, which has a similar form to that used in electrical capacitance tomography, is discussed. The image reconstruction algorithms, including the linear back projection (LBP), the Landweber iteration, and the total variation based regularization, are conducted and the results are compared with each other and with those from the traditional wire-mesh direct image reconstruction algorithm. Simulation results show that the proposed sensitivity map based image reconstruction method is of providing better resolution and helpful to improve the reconstruction quality of capacitive wire-mesh tomography.

Keywords Wire-mesh tomography, capacitive wire-mesh sensor, sensitivity map, image reconstruction, sub- pixel