动态特效实时建模方法的研究  

Research on Modeling Method of Dynamic Scenery

作　　者：杨会君[1] 彭宇行[2] 景旭[1] 

机构地区：[1]西北农林科技大学信息工程学院 [2]国防科学技术大学计算机学院

出　　处：《计算机应用研究》2007年第5期90-93,共4页Application Research of Computers

基　　金：国家"973"计划资助项目(2202CB312105);西北农林科技大学人才基金资助项目

摘　　要：提出了一种α-体建模方法:对实拍图像的学习生成动态纹理之后,再对纹理进行边界合成得到划分了颜色区域的图像;通过对图像区域边界的扫描勾勒出图像的轮廓线网格,并根据其到中心的距离为每个网格边界赋予标准的颜色值;最后根据网格的颜色值生成三维模型。实验证明这种动态景物的建模方法能够满足虚拟场景实时绘制的要求,而且具有很好的鲁棒性和连续性。This paper proposed a novel a-volume modeling arithmetic ： firstly, it learned inputed image sequence in order to create dynamic texture ,then the color regions of image could be obtained by synthesizing texture＇ s boundary, then drew the outline of image by scaning boundary of region and computed color value for every boundary of gridding according to distance away from the center,at last, created 3D model according to color value of gridding. Experimentation bear out that this modeling arithmetic of motion scenery can satisfy demand of paiting timely and is robust and consecutive.

关 键 词：动态纹理 网格 三维模型 动态景物 基于图像的绘制 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]