基于包围盒法的散乱点云数据的曲率精简

Curvature Estimation of Scattered-point Cloud Data Based on Bounding Box Method

采用非接触式扫描方法测量工件,能够获得高密集度的点云数据,但是过多的点云数据会严重影响曲面重构的光顺性。因此,精简点云数据成为逆向工程中相当重要的一环。提出了直接根据曲率变化精简点云的方法,对邻域搜索、曲率估算和曲率精简原则等进行了研究。对传统的邻域搜索方法进行了改进,采用包围盒法分割曲面,提高了点云精简的效率和精度。

When the work piece with the non-contact scanning method is measured,the high density point-cloud data can be obtained. But the excessive point-cloud data seriously influence the fairness of the surface reconstruction. Therefore,condensing point-cloud data becomes a quite important link in the reverse-engineering. According to the curvature variation, a method to reduce point-cloud is proposed, and it investigates the neighborhood search, curvature estimate and the curvature estimation principle and so on. Especially in the neighborhood search, an improvement is made to the traditional method. The new method divides up the surface with the bounding box method and raises the efficiency and the precision of a point-cloud simplification.