基于几何信号小波分析的网格编辑  

Mesh Editing Based on Geometric Signal Wavelet Analysis

作　　者：赵向军[1] 夏士雄[1] 

机构地区：[1]中国矿业大学计算机科学与技术学院

出　　处：《中国矿业大学学报》2008年第2期186-191,共6页Journal of China University of Mining & Technology

基　　金：国家自然科学基金项目(60373034)

摘　　要：提出了基于几何信号频谱分析的网格编辑新方法;该方法结合了自由变形(FFD)、层次细节编辑、几何信号处理以及小波变换等方法的优点,用简化模型作为几何代理,可以用较少操作完成编辑;将几何代理上的形变量作为几何信号进行小波分析,可以很好地分离出局部细节信息,方便进行分区信号处理工作;提出了允许修改代理模型拓扑结构的基于一致性参数化的网格编辑方法,以及保频谱自适应采样方法;并将基于小波变换的编辑方法进行了推广,将其应用于球面小波编辑之中,获得了较为满意的效果.A novel scheme of mesh editing based on wavelet analysis is proposed. Our approach has the merits of a number of techniques, such as FFD, LOD, DGP and wavelet analysis. With simplified models from original meshes used as geometric proxies, editing tasks can be completed with fewer operations. Detailed local information can be derived elegantly, while using shape variables as geometric signals on which wavelet analysis is conducted. This provides a convenient method to divisional signal processing. Its crucial novelty lies in the proposal of mesh editing based on consistent parameterization, which allows for modification of the topological structure of the proxy model. We also propose a new adaptive approach for better sampling of the geometric signals of the original model. As an extension of our spectrum editing scheme, we applied an editing principle to mesh editing, based on spherical wavelets with satisfactory results.

关 键 词：网格编辑 球面小波 自适应采样 数字几何处理 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]