Single Image Super-Resolution (SISR) through sparse representation has received much attention in the past decade due to significant development in sparse coding algorithms. However, recovering high-frequency textures is a major bottleneck of existing SISR algorithms. Considering this, dictionary learning approaches are to be utilized to extract high-frequency textures which improve SISR performance significantly. In this paper, we have proposed the SISR algorithm through sparse representation which involves learning of Low Resolution (LR) and High Resolution (HR) dictionaries simultaneously from the training set. The idea of training coupled dictionaries preserves correlation between HR and LR patches to enhance the Super-resolved image. To demonstrate the effectiveness of the proposed algorithm, a visual comparison is made with popular SISR algorithms and also quantified through quality metrics. The proposed algorithm outperforms compared to existing SISR algorithms qualitatively and quantitatively as shown in experimental results. Furthermore, the performance of our algorithm is remarkable for a smaller training set which involves lesser computational complexity. Therefore, the proposed approach is proven to be superior based upon visual comparisons and quality metrics and have noticeable results at reduced computational complexity.

To better extract feature maps from low-resolution (LR) images and recover high-frequency information in the high-resolution (HR) images in image super-resolution (SR), we propose in this paper a new SR algorithm based on a deep convolutional neural network (CNN). The network structure is composed of the feature extraction part and the reconstruction part. The extraction network extracts the feature maps of LR images and uses the sub-pixel convolutional neural network as the up-sampling operator. Skip connection, densely connected neural networks and feature map fusion are used to extract information from hierarchical feature maps at the end of the network, which can effectively reduce the dimension of the feature maps. In the reconstruction network, we add a 3×3 convolution layer based on the original sub-pixel convolution layer, which can allow the reconstruction network to have better nonlinear mapping ability. The experiments show that the algorithm results in a significant improvement in PSNR, SSIM, and human visual effects as compared with some state-of-the-art algorithms based on deep learning.