Research on the Super Resolution of Electromagnetic Inverse Scattering

Yahui Ding,Yongpin Chen,Jun Hu
DOI: https://doi.org/10.1109/piers62282.2024.10617986
2024-01-01
Abstract:Electromagnetic inverse scattering (EIS) has a great performance on its resolution as an imaging system. The super-resolution is usually achieved in the numerical simulation and practical experiments of EIS. However, the causation of super-resolution is unclarified, which limits the practical application and further development of EIS. A Classic explanation is the multiple scattering (nonlinearity) leads to the super-resolution. But this explanation is controversial in the field. In this paper, we believe that the reason for super-resolution is composite rather than singular. First, the influence of the size of the scatterers on resolution is often ignored. It is easy to confirm the resolution would be improved with the increase of the size of the scatterers by using the Sparrow criterion. Second, since the Born approximate (BA) could be regarded as the low-pass filter. The iteration methods based on BA should be regarded as rearranging the filter coefficients. With the iteration process, the coefficients of the high-frequency part will increase. Due to anti-apodization, the PSF becomes steeper, thereby improving resolution. Third, the point spread function (PSF) could be regarded as the distortion of the shape of scatterers in EIS. The position of the scatterers would also be distorted in EIS by the nonlinearity. When the distance of two scatterers is increased in the reconstruction image, it can be distinguished even if the distance is much smaller than the wavelength. At last, the iteration method uses the solution of the last step as the initial value in the new step would lead to extending the optical transfer function (OTF) which includes the information outside the Ewald sphere and leads to resolution improved. The new explication reveals the reason for super-resolution in EIS.
What problem does this paper attempt to address?