一种曲面特征保持的航空叶片点云精简方法

摘要

A novel point cloud simplification method was proposed to avoid insufficient sampling in aviation thin-walled blades’edge regions with high curvature feature and sparse regions with fewer points based on re-clustering strategy.To establish the geometric deviation between the original and simplified point cloud,point-surface distance function was defined to distinguish curvature information based on moving least-square method.The distance function was used to re-cluster the regions with high curvature feature to realize adaptive feature-preserving process.Moreover,the regions with sparse points was distinguished and subdivided to ensure sufficient points in sparse regions.Comparing with uniform sampling method and hierarchical clustering method,the method presented above has higher geometric precision for point-sampled blade surfaces or other complex surface point cloud,such as fan disk.%针对现有点云精简方法在航空薄壁叶片叶缘高曲率特征区域及点云稀疏区域存在取样不足的问题，提出了一种基于重聚类策略的精简方法。通过移动最小二乘法定义点---曲面距离函数，区分高低曲率特征，建立了原始点云和精简点云之间的距离关系，在高曲率区域重聚类，实现叶缘高曲率特征保持；并在重聚类时判断点云疏密程度，对稀疏区域进行重聚类。航空叶片、铸造模具等典型复杂曲面测量点云的精简过程中，该方法相比于均匀采样法、层次聚类法，在高曲率区域可保持较高的几何精度。