Generalized PDE Framework for Image Super Resolution

In this paper, we present a generalized partial differential equation (PDE) method with a generalized regularization term for super-resolution. The generalized regularization term based on fractional order derivatives which can be seem as a generalization of the total variation (TV) regularization term and two order regularization term, which it is possible to obtain better super resolution results in terms of both the similarity of the images and smoothness of the high resolution image. Experiments show the proposed method improves the objective quality as well as the subjective visual quality of the high-resolution images.