A Novel Modeling of MPEG-4 Video Sequence
Yan Huo, Tao Jing, Yumin Feng · 2006
This paper presents a novel modeling of MPEG video sequence. In this work, the sequence is modeled using mixture-exponential distribution. In order to capture long range dependence and self-similar characteristics of MPEG video sequence, the method of mixture-exponential is employed on the predicted model. Simulation results demonstrate that this model captures important statistical features of original sequence