摘要
提出了一个基于节点消去的B样条曲线的逼近算法。该算法首先从插值于给定数据点的一阶B样条曲线出发,利用广义逆矩阵实现节点消去,并通过升阶、最小二乘逼近和投影修正误差等步骤,得到了与给定数据点的误差在容许范围内的逼近曲线。
Algorithm for approximating a sequence of point data,which is sampled in a curve,is presented.Doing so,we have first generated a one-degree B-spline curve interpolating the given data.Then we have removed removable knots from the knot sequence by means of theory of generalized inverse matrix.After degree lifting,least squares approximation,and projecting point data onto the current approximation to correct error,we have arrived at the desired approximation with given accuracy.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第13期80-83,共4页
Computer Engineering and Applications
关键词
曲线逼近
广义逆矩阵
B样条曲线
节点消去
curve approximation, Moore-Penrose pseudoinverse matrix, B-spline curve, knot removal