Objective Quality Assessment for Digital Video Coding Based on Human Vision Model

摘要

There has been tremendous effort in the scientific community over the past years to conceive and develop a whole generation of new video communication systems. A valid quality assessment for compressed video is of interest to many, since the testing methodology for digital video coding and transmission systems has not been well formalized. This paper is to review the up-to-date knowledge of important characteristics of human visual system (HVS), which are helpful in optimizing video codec and providing constant picture quality in digital video services. In addition, a brief survey of recent progress in classification of reconstruction artifacts and quantification of coding distortions is presented. Furthermore, we discuss how to incorporate those perceptual properties into vision models for quantitatively measuring video quality, instead of presenting a particular model. This serves as future directions of research to be investigated on the objective visual quality assessment.