Research on Novel Single Image Super-resolution Algorithm through Regularization Approach and Joint Learning Theory: Theoretical Analysis and Applications

Abstract

In this research paper, theoretical analysis and applications of a novel single image super-resolution algorithm through regularization approach and joint learning is introduced. Digital image during the process of obtaining the optical fuzzy, movement deformation and degradation factors such as random noise, the influence of the resulting often degradation image, sometimes its resolution is difficult to meet the actual demand of engineering or military applications. In this paper, we combine the joint learning theory together with the regularization standard, through parameter selection, error estimation with omission and solution analysis steps. The proposed framework is based on modified super-resolution model and novel error estimation metrics. In the experiment section, we compare our proposed algorithm with other state-of-the-art and popularly adopted methodologies and use the well-known test image databases to conduct the experiment. The experimental result shows the feasibility and effectiveness of the algorithm. In the future, we plan to do more in-depth research on the parameter selection part to modify our method.