Recent advances in image quality assessment

Abstract

Over the last few decades, IQA has been an increasingly popular research topic in the fields of image processing and computer vision. In this chapter we will survey some recent advances in image quality assessment (IQA). This chapter is organized into three sections: subjective IQA, objective IQA, and new directions in IQA. More specifically, the first section will review widely used IQA databases, with emphasis on those recently proposed ones. The second section will review quality metrics in the familiar categorization of full-reference (FR), reduced-reference (RR), and no-reference (NR) ones. The third section introduces some emerging and interesting research directions including comparative IQA, multiply-distorted IQA, contrast-changed IQA, which we believe will be important topics for the future study of perceptual quality assessment.