Multiple Measurement Vector-Based Complex-Valued Multifrequency ECT
Maomao Zhang, Liying Zhu, Haokun Wang, Yi Li, Manuchehr Soleimani, Yunjie Yang
Abstract
Complex-valued, multifrequency electrical capacitance tomography (CVMF-ECT) is a recently developed tomographic concept which is capable of simultaneously reconstructing spectral permittivity and conductivity properties of target objects within the region of interest. To date, this concept has been limited to simulation and another key issue restricting its wide adoption lies in its poor image quality. This letter reports a CVMF-ECT system to verify its practical feasibility and further proposes a novel image reconstruction framework to effectively and efficiently reconstruct multifrequency images using complex-valued capacitance data. The image reconstruction framework utilizes the inherent spatial correlations of the multifrequency images as a priori information and encodes it by using multiple measurement vector (MMV) model. Alternating direction method of multipliers was introduced to solve the MMV problem. Real-world experiments validate the feasibility of CVMF-ECT, and MMV-based CVMF-ECT method demonstrates superior performance compared with conventional ECT approaches.