基于顶点重要度和层次聚类树的地形网格简化

摘要

To improve the efficiency of simpilification for terrain mesh based on digital elevation model (DEM),on the basis of traditional vertex clustering algorithm,an mesh simplification method based on vertex importance and hierarchical clustering tree was presented.The vertexes of terrain blocks were clustered,and the importance of vertex which consisted of eigenvalue and simplification error was used to choose the representative vertexes of clusters.Coarse meshes were constructed using these repre-sentative vertexs.By carrying on the suture and simplification process iteratively,the whole terrain mesh was simplified.A hie-rarchial clustering tree in internal memory was used to organize the intermediate simplification meshes and to guide the simplifica-tion and tessellation in the view-dependent level-of-detail rendering for terrain.Experimental results show that the algorithm gains good performance in remaining the terrain feature and reduces memory consumption.The efficiency of rendering is im-proved by taking use of the intermediate simplification results.%为提高基于精细DEM数据的地形网格简化效率，在传统顶点聚类算法的基础上，提出一种基于顶点重要度和层次聚类树的地形网格简化算法。对地形分块顶点进行聚类，引入由顶点特征值和简化误差构成的顶点重要度来筛选簇代表点；根据代表点构建粗化网格，迭代缝合与简化过程直至完成整个地形网格简化；采用内存层次聚类树对简化的中间结果进行组织，将之用于地形多分辨率绘制中，指导网格的简化和细分。实验结果表明，该算法较好地保持了原始地形特征，缩减了内存消耗，中间结果的利用有效提升了地形渲染效率。