平面上有噪音散乱点集的参数化  

Parameterization of planar unorganized points with noise

作　　者：王平[1,2] 刘轩昊[1] 邓建松[1] 

机构地区：[1]中国科学技术大学数学系,安徽合肥230026 [2]苏州大学数学系,江苏苏州215006

出　　处：《中国科学技术大学学报》2010年第12期1225-1229,1235,共6页JUSTC

基　　金：国家自然科学基金(60873109;61073108);新世纪优秀人才支持计划(NCET-08-0514)资助

摘　　要：散乱点集曲线重建关键步骤是参数化.提出了一种对平面上有噪音散乱点集参数化的新算法.算法中,首先求出点集的Delaunay三角化的最小生成树,继而对最小生成树的最长路径进行参数化.通过把其他点投影到最长路径上,对最长路径以外的点进行参数化.还给出了提高参数化的准确度的方法.以圆盘控制顶点B样条拟合为例展示了该算法的应用.Parameterization is a crucial step in curve reconstruction from unorganized points.An algorithm for parameterizing planar unorganized points with noises was described.In the algorithm,the minimum spanning tree of the Delaunay triangulation of the points was constructed first.Then,the longest path of the minimum spanning tree was parameterized.Finally,the other points which were not on the longest path were parameterized by projecting them onto the longest path.Some procedures were discussed to improve the accuracy of the algorithm.As an application,an example of disk B-spline fitting was provided.

关 键 词：参数化 噪音 散乱点 拟合 

分 类 号：TP391.72[自动化与计算机技术—计算机应用技术]