Multi-frame Super-Resolution from Noisy Data

Obtaining high resolution images from low data with clipped noise is algorithmically challenging due to the ill-posed nature of problem. So far such problems have hardly been tackled, and few existing approaches use simplistic regularisers. We show usefulness two adaptive regularisers based on anisotropic diffusion ideas: Apart evaluating classical edge-enhancing regulariser, we introduce a novel non-local one one-sided differences superior performance. It termed sector diffusion. combine it all six variants super-resolution observational model that arise permutations its three operators for warping, blurring, downsampling. Surprisingly, evaluation in practically relevant noisy scenario produces different ranking than noise-free setting our previous work (SSVM 2017).