Shape reconstruction of an inhomogeneous impedance cylinder via potential and neural network approach

In this article, a hybrid method is presented to reconstruct two-dimensional shape of an inhomogeneous impedance cylinder from the noisy far field data in the case of single electromagnetic plane wave illumination. The proposed reconstruction method can also be applied for complex-shaped smooth objects having different type of boundary conditions. For the solution of the direct problem, a potential (layer) approach is chosen that leads to a boundary integral equation. Applying Nyström method to the equation, an exponential convergence is obtained. The inverse scattering problem is solved via neural network approach, which is a stable technique under considerable level of noise. The applicability and the effectiveness of the proposed method are supported by numerical examples, and the proposed approach provides reconstructions which are in good agreement with the actual shapes.