Retargeted image quality assessment: Current progresses and future trends

Abstract

The diversity and versatility of the display devices have imposed new demands on digital image processing. Variant devices of different resolution screens need to display the same image for human visual experience. The retargeting methods are proposed to adapt the source image into arbitrary sizes and simultaneously keep the salient content of the source signal of high visual quality. Therefore, there is a new challenge of objectively evaluating the retargeted image perceptual quality, where variant resolutions may be presented, the objective shape may be distorted, and some content information may be discarded. In this chapter, recent progresses in quality assessment of retargeted images are reviewed. Firstly, we will review and discuss the recently developed retargeting methods for images. Afterwards, the subjective approaches to assess the retargeted image are reviewed, as well as the constructed subjective databases. Thirdly, some objective quality metrics developed recently are reviewed and compared based on the databases. Finally, future trends are discussed on retargeted image quality assessment in terms of both subjective and objective approaches.