Deep Learning-Based Image Reconstruction for Electrical Capacitance Tomography

Abstract Electrical capacitance tomography (ECT) is a non-invasive measurement technique widely used for two-phase flow imaging and parameter measurement. Image reconstruction of ECT analyzes the capacitance measurements from the ECT sensor and reconstructs the permittivity distribution in the sensing domain through certain algorithms. Due to its ill-posedness, image reconstruction has always been a hotspot and a challenge in ECT research. Over the past decade, the blooming of deep learning has introduced promising avenues for addressing this challenge. Numerous deep learning-based models and algorithms have been developed for ECT image reconstruction, and remarkable achievements have been made. This paper comprehensively summarizes the state-of-the-art deep learning approaches for ECT image reconstruction. In addition, the challenges and future directions of deep learning-based ECT image reconstruction are also discussed in perspective.