Issues related to slice based modeling of MPEG VBR encoded video

Abstract

The objective of the paper is to explore the possibility of developing a slice based model whose random variable is independent and would fit classical distributions. The authors analyze four MPEG-1 VBR encoded high quality, color video sequences. None of the sequences contain audio. The authors present three types of slice based models and discuss the merits of each. They show the distributions given by each of the models and show their fit to the gamma and Pareto distributions using the QQ plot.