一种新的连续性LOD实现算法

A Novel Continuous LOD Algorithm

为进一步提高网格简化速度和实现连续性的层次细节(LOD),文中利用球形包围体对模型顶点进行分割,通过基于视距的误差阈值和最小误差聚类算法综合控制包围球的体积大小并对包围球的顶点进行聚类.实验结果显示,该算法能够较快、高保真地对模型网格进行简化,并解决了传统算法在层次细节实现上非连续性的问题,证明该算法是有效和实用的.

For improving velocity of Mesh Simplification and realizing continuous Level of Details（LOD）,this paper uses spheres to segment vertexes of models,and control the radius of each spheres by a cluster error threshold,then cluster the vertexes in each sphere by a minimum error-based cluster algorithm.The experimental results show that this method can simplify model meshes with a comparatively high speed and high fidelity,and provides a solution to continuous LOD,so it is effective and useful.