超分辨率图像失真分析与质量评价研究

Distortion Analysis and Quality Evaluation of Super-resolution Images

随着视频增强算法取得的效果日益精进,如何设计客观质量评价模型来评价增强视频或图像的视觉感知质量成为目前的技术难点。本文针对超分辨率图像失真特点与类型开展分析,这也是后续研究视频增强质量评价算法的一项基础性工作。同时,本文还从建立超分辨率质量评价数据库、开展无参考客观质量评价模型的研究等方面讨论了超分辨率图像失真质量评价问题的研究思路。

With the increasingly sophisticated video enhancement algorithms,how to design an objective quality evaluation model to evaluate the visual perception quality of enhanced videos or images is still a technical difficulty at present.This paper analyzes the characteristics and classifications of super-resolution image artifacts as the basic work for prospective research on super-resolution image quality evaluation algorithms.Besides,future research of super-resolution image quality evaluation is discussed from the aspects of establishing a quality evaluation database of super-resolution content and developing a blind image quality assessment metric.