Neural Network-Based Conductivity Imaging Using Time-Domain Reflectometry
Xinying Li, Georgiana Dima, Jiaruo Yan, Christopher John Stevens, LASZLO SOLYMAR, Ekaterina Shamonina · 2025
We present a neural network-based approach for sensing conductive objects using time-domain reflectometry of magnetoinductive (MI) waves in metamaterials with inter-element coupling. Our method significantly improves previously reported accuracy and expands the detection range, enabling simultaneous localisation of multiple conductive objects and paving the way for real-time, contactless imaging of inhomogeneous conductive media, with potential applications ranging from 3D printing quality control to medical imaging.