Objective: To develop, and demonstrate the feasibility of, a novel image reconstruction method for absolute Electrical Impedance Tomography (a-EIT) that pairs deep learning techniques with real-time robust D-bar methods. Approach: A D-bar method is paired with a trained Convolutional Neural Network (CNN) as a post-processing step. Training data is simulated for the network using no knowledge of the boundary shape by using an associated nonphysical Beltrami equation rather than simulating the traditional current and voltage data specific to a given domain. This allows the training data to be boundary shape independent. The method is tested on experimental data from two EIT systems (ACT4 and KIT4). Main Results: Post processing the D-bar images with a CNN produces significant improvements in image quality measured by Structural SIMilarity indices (SSIMs) as well as relative 2 and 1 image errors. Significance: This work demonstrates that more general networks can be trained without being specific about boundary shape, a key challenge in EIT image reconstruction. The work is promising for future studies involving databases of anatomical atlases. S. J. Hamilton is with the

In many fields of process industry, it is essential to be able to accurately measure the volumetric flow rates and mass flows of different materials flowing in the process pipes. To estimate the volumetric flow rate of a certain phase, the volumetric fraction and the flow velocity field of the phase in a cross-section of the process pipe need to be known. Electrical tomography (ET), such as electrical resistance tomography, can be used to measure the cross-sectional volumetric fractions in two-phase flows. For velocity field metering, electromagnetic flow tomography (EMFT) techniques have recently been developed. By combining the multimodal information from the ET and EMFT modalities, volumetric flow rates in process pipes can be estimated. This paper reports the results from experiments applying a multimodal imaging approach utilizing an earlier developed EMFT system and a commercial ET system, combined for measuring the volumetric flow rates in a laboratory flow loop. The studies undertaken in the paper demonstrate that the proposed multimodal imaging approach can produce reliable volumetric flow rate and mass flow rate estimates in laboratory environment in the cases of oil–water and solids–water flows. The relative errors of the estimated volumetric flow rates in the case of oil–water flows, with the average flow velocity between 0.79 ms−1 and 1.17 ms−1, were found to be in the range of 7.8%–9.5%, the average being 8.2%.

Electromagnetic flow meters have succesfully been used in many industries to measure the mean flow velocity of conductive liquids. This technology works reliably in single phase flows with axisymmetric flow profiles but can be inaccurate with asymmetric flows, which are encountered, for example, in multiphase flows, pipe elbows and T-junctions. Some computational techniques and measurement devices with multiple excitation coils and measurement electrodes have recently been proposed to be used in cases of asymmetric flows. In earlier studies, we proposed a computational approach for electromagnetic flow tomography (EMFT) for estimating velocity fields utilizing several excitation coils and a set of measurement electrodes attached to the surface of the pipe. This approach has been shown to work well with simulated data but has not been tested extensively with real measurements. In this paper, an EMFT system with four excitation coils and 16 measurement electrodes is introduced. The system is capable of using both square wave and sinusoidal coil current excitations and all the coils can be excited individually, also enabling parallel excitations with multiple frequencies. The studies undertaken in the paper demonstrate that the proposed EMFT system, together with the earlier introduced velocity field reconstruction approach, is capable of producing reliable velocify field estimates in a laboratory environment with both axisymmetric and asymmetric single phase flows.

No abstract

No abstract