Reduced-Reference Video Quality Assessment Based on BP- Neural Network Model for Packet Networks

This paper investigates quality monitoring of videos transmitted over packet networks. Our goal is to develop a methodology that is both simple and accurate to support quality assessment for videos over packet networks. For this purpose, this paper focus on the parameters that affect the quality of videos and uses back propagation Neural Networks (BP-NN) to mimic the way that human viewers assess the quality of videos transmitted over packet networks. Because network factors and motion change degree, that many people may ignore, are both important factors for video quality assessment, in this paper we propose a video quality assessment system considering packet loss rate (PLR), the motion change degree of video and subjective ratings as the inputs of back propagation Neural Networks (BP-NN) for training to get more precise video quality. Simulation result shows that our system can be used to measure the subjective video in real time with very good precision. © 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of ESEP 2011