Graph Spatio-Spectral Total Variation Model for Hyperspectral Image Denoising

The spatio-spectral total variation (SSTV) model has been widely used as an effective regularization of hyperspectral images (HSI) for various applications such mixed noise removal. However, since SSTV computes local spatial differences uniformly, it is difficult to remove while preserving complex structures with fine edges and textures, especially in situations high intensity. To solve this problem, we propose a new TV-type called Graph-SSTV (GSSTV), which generates graph explicitly reflecting the structure target HSI from noisy HSIs incorporates weighted difference operator designed based on graph. Furthermore, formulate removal problem convex optimization involving GSSTV develop efficient algorithm primal-dual splitting method problem. Finally, demonstrate effectiveness compared existing models through experiments source code will be available at https://www.mdi.c.titech.ac.jp/publications/gsstv.