Quality controlled artifact reduction

Strongly varying input content to modern flat panel displays makes highly adaptive algorithms mandatory. On the one hand, low quality content is dominated by coding artifacts and therefore strong artifact reduction can increase image quality. On the other hand, too strong processing removes image details in high quality video. These contrary optimization criteria cannot be fulfilled easily by simple processing strategies. This paper proposes a framework utilizing different artifact reduction algorithms while their output is controlled by objective quality measures selecting the most suitable algorithm at run-time. Thus, a dynamic selection of the optimal processing for each input scenario over a wide quality level is possible. Integrating a feedback loop and applying the process again can further enhance the output quality.