Generative Adversarial Network-based Enhancement for Super-Resolution Reconstruction in Division of Focal Plane Images

Advancements in technology have refined polarization imaging systems for realtime, multi-directional imaging. However, their super-pixel design leads to instantaneous field of view (IFoV) issues. Addressing this, a super-resolution method using the Super-Resolution Generative Adversarial Network (SRGAN) has been introduced. This method efficiently recovers high-quality details from low-resolution polarimetric images. Using PSNR and SSIM metrics, this method demonstrates enhanced performance over existing techniques.