移动多曲线/曲面逼近权函数的改进  

Improvement of Weight of Moving Multiple Curves/Surfaces Approximation Model

作　　者：冯文月 FENG Wen-yue(School of Statistics and Apllied Mathematics,Anhui University ofFinance and Economics,Bengbu 233030,Anhui,China)

机构地区：[1]安徽财经大学统计与应用数学学院,安徽蚌埠233030

出　　处：《合肥学院学报（综合版）》2019年第2期1-4,15,共5页Journal of Hefei University：Comprehensive ED

基　　金：安徽财经大学校级一般项目"移动多曲线/面逼近权函数的改进"(ACKY1761)资助

摘　　要：移动多曲线/曲面逼近模型可以分离来自多条曲线/曲面的混合点,具有一定的实际应用意义,但这个局部模型中的权函数比较复杂,针对这一具体问题,根据带有高斯噪音的数据点的特征及概率论的理论知识对其权函数进行简化,便于计算,并为权函数的改进提供了一定的理论基础。一系列的数值实验结果表明,简化后的模型分离混合点云数据也是很有效的。Moving multiple curves/surfaces approximation model can be used to separate mixed points.It is significant for practical application.But this local model is quite complicated.For this problem,according to the characteristics of data points with Gaussian noise and theoretical knowledge of probability theory,the weight function in the model is simplified.This is convenient for computation and this improvement is based on the theoretical basis of probability theory.A series of numerical experiments show that the simplified model is quite effective for separating mixed data.

关 键 词：移动多曲线/曲面逼近 权函数 点云数据 概率论 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]