A novel robust fractional-time anisotropic diffusion for multi-frame image super-resolution

In this paper, we propose an image multi-frame Super Resolution (SR) method based on fractional-time Caputo derivative combined with Weickert-type diffusion process idea. We provide the existence and uniqueness results with a detailed discretization using the finite difference scheme. Our approach is based on anisotropic diffusion behavior with coherence enhancing diffusion tensor together with the fractional-time derivative to benefit from its memory effect potential and to control the smoothing process near strong edges and flat regions while avoiding tiny corners destruction. The experimental results confirm the effectiveness of fractional-time derivative and the robustness of the proposed Partial Differential Equation (PDE) compared with some competitive super-resolution methods.