Abstract:

With the development of computer theory and technology, image super-resolution theory and technology have continuously made new progress, and a series of methods such as interpolation, reconstruction and machine learning have been developed. This paper reports the research progress of image super-resolution. Firstly, it introduces the methods of image super-resolution. Secondly, it summarizes major deep learning super-resolution models and the current development trend of image super-resolution. Finally, the future development and challenges of super-resolution research are proposed.

Key words: super-resolution, image reconstruction, deep learning

摘要：

随着计算机理论与技术的发展，图像超分辨率理论和技术手段不断取得新的进步，发展出插值法、重构法和学习法等一系列方法。报告了图像超分辨率的研究进展，梳理了主要的图像超分辨率方法，阐述了几种较为重要的深度学习超分辨率模型，总结了当前图像超分辨率的发展趋势，对超分辨率的研究提出了展望。

关键词: 超分辨率, 图像重构, 深度学习