Simplified Model for Representing Dynamic Textures using Markov Model
Tong Fu1, Di Yin2, Li Xiaoli3, Chen Hui4

1Tong Fu, College of Computer Science and Engineering, Xi‘an University of Technology, Xi‘an, China.
2Di Yin, School of Electrical Engineering, Wuhan University, Wuhan, China.
3Li Xiaoli,School of Computer Science, Jiangxi Normal University, Nanchang, China.
4Chen Hui, College of Information Science and Technology, Chengdu University, China.
Manuscript received on April 04, 2013. | Revised Manuscript received on April 28, 2013. | Manuscript published on May 05, 2013. | PP: 106-109 | Volume-3, Issue-2, May 20 13. | Retrieval Number: B1449053213/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (

Abstract: Dynamic textures are sequences of images of moving scenes that exhibit certain stationary properties in time; these include sea-waves, smoke, foliage, whirlwind etc. In previous works [1,2], dynamic textures are usually modeled as linear models, and parameters of the model are identified in the sense of maximum likelihood or minimum prediction error variance. Once its parameters are learned, a model has predictive power and can be used for extrapolating synthetic sequences. In this work we study a particular type of dynamic textures that can be represented in the form of Markov Models. An aggregation algorithm can then be adopted to reduce its complexity. The resulting low-dimensional models can capture complex visual phenomena with low computation cost.
Keywords: Dynamic Texture, Markov Model, Aggregation, Reduced order model.