Stephen D. Voran

Abstract: Discusses research efforts to derive objective measures of video quality that emulate human perception. The derivation of these metrics involves the following steps: (1) a set of test scenes in selected and distorted, (2) a set of candidate objective measures are extracted, (3) a panel of viewers rates the quality of the same set of test scenes, (3) a simultaneous statistical analysis of the subjective and objective data sets reveals which portion of the objective data is meaningful, and how the objective data should be combined to create an overall metric that emulates human perception. One objective metric that correlates well with subjective quality quantifies the amount of false or extra edges that have been added to the output video. There appears to be some advantages to applying the metric separately to the still and motion portions of the video.

Keywords: digital; performance; models; video; quality; parameters; objective; subjective; correlation; spatial; temporal; compression; conferencing; motion; information; Sobel; assessment; bit; errors; difference; still; false; edges

Disclaimer: Certain commercial equipment, components, and software may be identified in this report to specify adequately the technical aspects of the reported results. In no case does such identification imply recommendation or endorsement by the National Telecommunications and Information Administration, nor does it imply that the equipment or software identified is necessarily the best available for the particular application or uses.

For questions or information on this or any other NTIA scientific publication, contact the ITS Publications Office at ITSinfo@ntia.gov or 303-497-3572.

Back to Search Results