基于散乱点的增量式曲面逼近  

Incremental Surface Approximation from Point Cloud

作　　者：秦绪佳[1] 王青[2] 鲍虎军[2] 

机构地区：[1]浙江工业大学软件学院,杭州310032 [2]浙江大学CAD&CG国家重点实验室,杭州310027

出　　处：《计算机辅助设计与图形学学报》2006年第9期1408-1413,共6页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金(60573152);浙江省自然科学基金(Y104341);浙江大学CAD&CG国家重点实验室开放课题基金

摘　　要：针对用接触式三维点数据获取设备快速输入的物体表面散乱点云数据,提出了增量式B样条曲面快速逼近算法.该算法首先要获得重建曲面的边界数据,以生成初始曲面;然后对输入的散乱数据点云用投影法计算出其参数值;再用模板子块在曲面上移动,反算出模块子块的控制点;最后更新整个曲面的相应控制点,实现边输入、边逼近,即增量式曲面逼近.在输入过程中可看到曲面逐渐逼近目标曲面的过程,在误差大的区域可以增加输入点来改善曲面逼近效果.对于复杂曲面进行多次投影计算散乱数据点参数及曲面逼近,可达到良好效果.A rapid incremental surface approximation algorithm is proposed to establish B-spline surface from the unorganized point cloud. The algorithm is composed of four steps. Firstly, boundary data are obtained in order to set up an initial surface. Secondly, parameterization of raw data is established by projecting. Thirdly, a mask surface is designed to move on the reconstructed surface and its control points are calculated. Finally, the control points of mask surfaces and control points of the whole surface are updated. The reconstructed surface can gradually approximate the target surface by incrementally adding additional sampling points. For a complex surface, can be obtained by computing points＇ parameter and approximating the target surface twice or more a satisfactory result.

关 键 词：曲面重建 逼近 B样条 点云数据 增量 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]