A Review of Deep Learning-Based Super-Resolution

Super-resolution technology generally refers to the use of known information to infer more vivid and detailed images. Due to their wide range of applications, they have consistently garnered significant attention and discussion in the field of image vision. This paper aims to give a comprehensive overview of the super-resolution reconstruction technology under the framework of deep learning, divide it into five categories according to the network model and expand the overview. In addition, it also organizes the commonly used data sets and evaluation indicators for training super-resolution algorithms, and summarizes the influence of data sets on these algorithms. Finally, this paper points out the challenges and limitations of the current super-resolution technology, and looks forward to the future.