Image reconstruction of Electrical capacitance tomography with non-uniform electrode distribution

  • Jingqian Hong

Student thesis: Master of Philosophy

Abstract

Electrical capacitance tomography (ECT) is a process imaging technology based on capacitance-sensitive mechanisms. Its advantages of non-radiation, non-invasive fast response, simple structure, low cost, wide application range, and excellent safety performance have become the current process tomography technology. The main stream of development and research hotspots have broad industrial application prospects. However, the development of ECT is still at the laboratory stage and is rarely used in the industrial environment. Applying ECT in the industry is necessary to improve the technology. One of the tasks is to study image reconstruction with non-uniform electrode distribution. For this purpose, an ECT sensor with 8 electrodes but non-uniformly distributed is studied. The sensitive maps of the traditional 8-electrode ECT sensor are used to solve the inverse problem of the ECT sensor with 8 electrodes non-uniformly distributed. The influence of non-uniform electrode distribution on the quality of the reconstructed image is analysed and evaluated by various performance indicators. In addition, the tactile detection of manipulators has been continuously developed in the past ten years, and the object detection of manipulators’ tactility has become one of the key research topics today. To combine the ECT system with manipulator tactile detection, according to the shape of pipe grasped by fingers, the experiment simulates the distribution of multiple capacitive electrodes mounted on a robot hand’s fingers. It also reconstructs images through these capacitive sensors and assesses image quality. Besides, in practical applications, not every permittivity of the measured substances is known. It is necessary to measure the multiphase flow distribution of materials with unknown permittivity. In order to reconstruct the image of the material distribution with unknown permittivity, this thesis will use the sensitivity map measured by other known permittivities for image reconstruction in the case of unknown permittivity. COMSOL software is used to model the ECT sensor. MatLab software is used to solve the forward and inverse problems of the ECT system. Linear-Back projection algorithm and Landweber iterative algorithm with 100 iterations are used for image reconstruction.10 Simulation results show that the image quality of the ECT sensor with non-uniform electrode distribution is affected by the lack of electrodes, and the quality of the re constructed image is related to the sensitive maps used. If the reconstruction effect of the sensitive map which use for solving inverse problem can be optimised, the quality of reconstruction images in the same ECT system will be improved.
Date of Award1 Aug 2021
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorWuqiang Yang (Supervisor) & Wuliang Yin (Supervisor)

Keywords

  • COMSOL
  • Electrical capacitance tomography
  • Image reconstruction
  • ECT Sensor

Cite this

'