Joint Dictionary-Based Method for Single Image Super-Resolution

Image super-resolution technique mainly aims at restoring high-resolution image with satisfactory novel details. In recent years, sparsity-based super-resolution has attracted great interests for its impressive results. By using learning dictionaries, sparsity-based methods try to find some mapping relationships as prior knowledge between low-and high-resolution example images for better reconstruction. In this paper, based on two of the state-of-the-art sparsity-based super-resolution methods, we propose a joint dictionary-based framework to improve the quality of reconstructed high-resolution images. Experimental results illustrate that our method outperforms the other state-of-the-art methods in terms of sharper edges, clearer textures and better novel details.