Fig. 4
From: Self-supervised model-informed deep learning for low-SNR SS-OCT domain transformation
Pre-processing for super-resolution task, make size of low-resolution image same as the size of high-resolution image with missed A-scans filled by noise.
