An Image Reconstruction Method of Capacitively Coupled Electrical Impedance Tomography (CCEIT) Based on DBSCAN and Image Fusion

Based on linear back projection (LBP) algorithm, density-based spatial clustering of applications with noise (DBSCAN) algorithm and image fusion technique, a new image reconstruction method of capacitively coupled electrical impedance tomography (CCEIT) is proposed in this work. LBP is used to obtain the original image of the real part and the original image of the imaginary part. The gray level thresholds of the two original images are determined by DBSCAN clustering algorithm and the initial image of the real part and the initial image of the imaginary part are obtained by gray level threshold filtering, respectively. Finally, the two initial images are fused and the final image is obtained. Image reconstruction experiments are carried out with a 12-electrode CCEIT system prototype. The experimental results verify the effectiveness of the proposed image reconstruction method. Experimental results show that the quality of the final image is satisfactory. Compared with the conventional electrical tomography (ET) image reconstruction methods, the new proposed image reconstruction method needs less prior knowledge or manual intervention and realizes more effective usage of the impedance information. The research results also indicate that the DBSCAN clustering algorithm is an effective way to determine the gray level thresholds of the original images and the mean square error (MSE) is a reasonable criterion for image fusion. Meanwhile, LBP + DBSCAN is an effective algorithm for image reconstruction and image fusion technique is an effective way to utilize the whole impedance information of the conductive gas–liquid two-phase flow.

I. Introduction

In the research field of industrial process tomography (PT), capacitively coupled electrical impedance tomography (CCEIT), which refers to the contactless detection idea of capacitively coupled contactless conductivity detection technique () [1]–[12], is a new electrical tomography (ET) technique. Compared with conventional contact electrical resistance tomography (ERT) or electrical impedance tomography (EIT), CCEIT can realize contactless measurement and avoid the electrode polarization and electrochemical corrosion effect [6], [10]. Furthermore, unlike the conventional ERT systems or EIT systems, CCEIT regards the conductive gas–liquid two-phase flow as an impedance and uses the whole impedance information (including the real part and the imaginary part) to implement image reconstruction, which pays sufficient attention to the importance of the imaginary part [3]–[7], [10]. For conventional ERT or EIT systems, the usage of the impedance information is not sufficient [6]–[11], [15]–[21]. The conventional ERT systems and some EIT systems usually use the real part of the impedance to implement image reconstruction [6]–[10]. Some EIT systems use the amplitude of the impedance to implement image reconstruction [6], [10], [15]–[21]. Although the amplitude is a way to use the real part and the imaginary part simultaneously, it is a simple and direct information usage approach and the role of the imaginary part has not received sufficient attention [5]–[7], [10]. Therefore, the CCEIT system has better potential in phase distribution reconstruction and parameter measurement of the gas–liquid two-phase flow [6], [10]. However, as a developing technique, the research works on the image reconstruction method of CCEIT are relatively limited.