Image Super-Resolution reconstruction based on adaptive gradient field sharpening

The Super-Resolution (SR) image reconstruction is a computational technique that improves the resolution of an image system. In this paper we propose a new method for multi-image SR reconstruction, which is capable of producing sharper images with clearer image details. A two-step method is used. The first step is conventional SR image reconstruction. From this preliminary result, an image gradient is determined and transformed to increase the sharpness, and then it is included into the second reconstruction as a prior to get the final result. The Gradient Profile prior [4] is used in the gradient transform for sharpness enhancement. Results demonstrate the effectiveness of the proposed method to produce sharp and clear images from low-resolution images.