ECT Image Reconstruction Method Based on Multi-Exponential Feature Extraction

Electrical capacitance tomography (ECT) provides an effective way for solving the problem of two-phase flow measurement. However, there is a problem of underdetermination, which leads to the low accuracy of ECT image reconstruction. In this article, a new ECT image reconstruction method based on multi-exponential feature extraction is proposed to improve the accuracy of ECT image reconstruction. First, it can be observed that the grayscale of two-dimensional images is a typical Dirac pulse sequence, and the grayscale vector can be modeled as a discrete finite rate of innovation (FRI) signal. Then, the FRI sampling system is used to extract the feature information after filtering by exponential reproducing kernel. Finally, by means of zero filling and random recombination, a comprehensive observation equation can be constructed by using feature information. The original image signal is recovered by solving a L0 norm optimization problem. Simulation results have shown that the images reconstructed by the method in this article outperform those reconstructed by existing algorithms.

I. Introduction

Process tomography has an attractive development prospect in industrial measurement and medical monitoring. With the development of industry, there is an increasing demand for the visual detection of two-phase flow parameters in pipelines. Electrical resistance tomography (ERT) and electrical capacitance tomography (ECT) are two of the most mature process tomography technologies. ERT measures the resistivity distribution of the imaging area, while ECT measures the permittivity distribution by detecting the capacitance values [1], [2]. ERT and ECT have the following advantages over other process tomography techniques, such as X-ray tomography, gamma ray tomography, ultrasound tomography: low cost, noninvasive, simple structure and safety, and so on [3]–[5]. ERT already has many applications. Ren et al. [6] investigated the feasibility of using ERT to visualize the surface deposit during a cleaning-in-place (CIP) process. Wang et al. [7] demonstrated that ERT is a reliable imaging tool for the real-time measurement of complex two-phase flows in the packed bubble column reactors.