摘要
提出了一种密集采样数据的特征点表示法。该方法首先对密集采样数据进行滤波处理，去掉测试过程中的干扰信号和随机误差；然后，采用预定的特征点表示的Ｂ样条曲线与密集采样数据比较，若超过允差，则增加特征点修改Ｂ样条曲线，直至使特征点表示的Ｂ样条曲线与密集采样点值相符为止。这一方法主要适用于几何量或与之相当的缓变物理量的密集采样数据的数据处理。本文还对该方法用于活塞中凸型线的测量数据处理进行了实验，结果表明该方法实用可靠。
A method is proposed to process high density sample data based on feature points . After the interference signal and random error are filtered out by filter processing, the high density sample data is compared with the B-spline curve based on feature points. If the error exceeds the limit value, the feature point is increased to modify the B-spline curve until the high density sample data is conformed to it. The method is mainly suited to process the high density geometrical sample data.
出处
《计量学报》
CSCD
北大核心
1997年第4期284-288,共5页
Acta Metrologica Sinica
