全信息图像质量评估研究发展综述

A Review of the Research on Full Reference Image Quality Assessment

随着图像技术的广泛应用,图像的质量评估问题,即信息的可靠性问题,在自动化系统中的地位日益重要。全信息图像质量评估是针对目标原始图像和被测图像都已知时的被测图像评估。从历史发展角度系统地回顾了20世纪60年代以来全信息图像质量评估研究的发展历程和现状,并从算法的构成角度对代表性的全信息图像质量评估算法进行了分类与分析。由于图像质量评估函数的性能测定在其研究中具有重要的地位,对于图像质量评估函数的性能测定方案给予了说明。最后,从研究的深入角度上讨论图像质量评估研究的未来发展趋势。

With the wide appliance of imaging technology in the automation system, image quality assessment （IQA）, namely the reliability of vision information, has IQA is the quality measurement with test and reference enjoyed an increasingly important status. Full reference image are known. This paper systematically reviews the history of full reference image quality research. And a rational category for representative IQAs is given from algorithm aspect. As performance assessment is important in IQA research, the performance evaluation of IQA is also introduced here. In the end of this article, the future trends of full referemce IQA are discussed from research aspects.