A Self-Generated Local Structural Prior for Resolving Target Merging in Electrical Impedance Tomography (EIT)

Abstract

Electrical impedance tomography (EIT) is an ill-posed inverse imaging technique with limited and spatially nonuniform resolution. Conductivity targets located in regions with degraded spatial sensitivity may be reconstructed as a single connected region, resulting in the loss of spatial separability between neighboring structures. In this work, we propose a self-generated local structural prior for suppressing target merging in EIT reconstruction. First, an initial conductivity image is obtained from the measured boundary data using an existing EIT reconstruction method. This reconstruction is subsequently used as a reference image to construct a Gaussian similarity matrix according to the intensity similarity between image pixels. The adaptively chosen eigenvector of the covariance matrix is then employed to extract the principal structural information of the reference image. A localized structural mask is constructed and incorporated into a subsequent EIT reconstruction to preserve the spatial separability of neighboring targets. Numerical and water tank experimental results demonstrate that the proposed method effectively separates neighboring conductivity targets that are merged by conventional EIT reconstruction methods.