Research And Realization Of Dynamic Textures In Natural Scenery Simulation

In this paper, time series analysis and statistic analysis are used to study the statistic characteristics of video information and dynamic characteristics of scenery. The state space model of the scenery is built based on dynamic texturing mechanism. A suboptimization method is used to solve the model parameters whose effects on image sequence are analyzed and used to edit dynamic texture. The dynamic characteristics of the image sequence generated based on MATLAB programming are similar to those of original video. The main contents are as follows.1) Dynamic texture generating mechanism. According to practical requirements of actual effects and computation capacity, time series analysis method is used to analyze how to build the state space model of natural scenery dynamic texture that can facilitate dynamic texture composition.2) Dynamic texture model solution. Based on the established state space model, the characteristics and meaning of the model solution are analyzed. Matrix A, B, and C are solved to construct the state space model, namely dynamic texture model.3) Dynamic texture editing. Under certain conditions, some parameters are changed to adjust matrix A, B, and C of the dynamic texture model so as to interactively edit the simulation effects of dynamic texturing. In experiments, 100 frames of smoke images of 320*220 pixels are used to qualitatively observe and quantitatively analyze the effects of matrix A, B, and C on image fineness, velocity of movement, and noise intensity of dynamic texturing.4) Dynamic texture realization. The aforementioned dynamic texture modeling and editing methods are used to realize the dynamic texturing of fire based on MATLAB programming.5) Dynamic texture applications. The effects of channel selection, fusion of dynamic texture and scene, etc on dynamic texturing are discussed. Several dynamic texturing applications are analyzed that validate the methods and system developed.Finally, the conclusions and future researches are presented at the end of the paper.