This report details a laboratory study investigating the effects of various scene parameters and network conditions on the ability of viewers to discriminate among objects in recorded video. It is intended to be used both to make minimum bit rate recommendations for encoding of video used in public safety contexts and to compare the encoding performance of the H.265 High-efficiency Video Coding (HEVC) standard to that of H.264 Advanced Video Coding (AVC) using the results of a previous study conducted by the Public Safety Communications Research (PSCR) program. The scene content parameters under study in this test are target size, motion, and lighting. Network conditions include resolution reduction and H.265 HEVC encoding at a variety of bit rates. Recognition rates are calculated and presented as percentages of correctly recognized objects within a particular set of videos, adjusted to account for the probability of correct guesses. As in the previous PSCR study, the task-based subjective tests described here follow the methods outlined in International Telecommunications Union, Telecommunication Standardization Sector (ITU-T) Recommendation P.912.
|Task-Based Video Quality Assessment of High-efficiency Video Coding (2015)||2.27 MB|