Development of 3-Dimensional Electrical Capacitance Tomography Based on Neural Network Multi-criterion Optimization Image Reconstruction

W Warsito and L-S Fan

Department of Chemical Engineering, The Ohio State University, Columbus, OH 43210, USA Email: fan.1@osu.edu

ABSTRACT

A 3-dimensional image reconstruction technique based on neural network multi-criterion optimization (NN-MOIRT) for electrical capacitance tomography (ECT) is developed for the first time. The 3-D ECT is accomplished by introducing a 3-D sensitivity matrix based on a twin-plane capacitance sensor into the NN-MOIRT algorithm. The optimization problem in the image reconstruction is solved using the Hopfield network model by minimizing the four objective functions: 1) negative entropy function, 2) least square errors, 3) smoothness and small peakedness function, and 4) 3-to-2D matching function. The technique has been tested on a capacitance data set obtained from simulated measurements, showing the feasibility of the 3-D image reconstruction.

Keywords 3-D ECT, NN-MOIRT, image voxel, multi-criterion optimization