Quality monitoring of streaming video based on packet-loss artifacts

This paper investigates quality monitoring of streaming video affected by packet losses. Our objective is to present a methodology that is both reasonably accurate and simple enough to support perceptual quality monitoring of streaming video. Because the ultimate impact of packet-loss on streaming video is the possible decrease of intact playback frames per second of video and the characteristics of video content play an important role in the perceptual quality of the affected video, we develop an approach which can evaluate the perceptual video quality based on the characteristics of video content and packet-loss artifacts. In the approach we firstly have subjective tests to study the impact of frame rate on video quality based on the characteristics of video content and then provide an algorithm to compute the intact playback frames per second of video affected by packet-loss, and finally introduce a video quality metric. The simulations and experiments have demonstrated its feasibility and accuracy.