用直线分解曲线的最简单算法是什么？
What is the simplest algorithm of decomposition of the curve by straight lines?
我在平面上有一组点 (x, y) 来描述某条曲线。
用预定数量的直线分解曲线最简单的算法是什么？
float x[100], y[100]; // points of curve
int lines = 5; // straight lines
float x_l[lines + 1], y_l[lines + 1]; // required to find 6 points for 5 lines
要求最优分解，使曲线所有点与直线的标准差最小化。
不知道这是不是最简单的...
佩雷斯，胡安-卡洛斯；维达尔，恩里克。数字化曲线的最佳多边形近似。 模式识别字母, 1994, 15.8: 743-750.
Given N ordered points in the plane and a constant M < N, an efficient
algorithm is proposed to find M points, among those given, which
define a polygonal curve that is a globally optimal approximation to
the given points. The algorithm accommodates any properly defined
error measure and the use of the most popular of these measures is
studied in detail to maximize the computational efficiency.
Experiments are reported showing the performance and usefulness of the
proposed method.
我在平面上有一组点 (x, y) 来描述某条曲线。 用预定数量的直线分解曲线最简单的算法是什么？
float x[100], y[100]; // points of curve
int lines = 5; // straight lines
float x_l[lines + 1], y_l[lines + 1]; // required to find 6 points for 5 lines
要求最优分解，使曲线所有点与直线的标准差最小化。
不知道这是不是最简单的...
佩雷斯，胡安-卡洛斯；维达尔，恩里克。数字化曲线的最佳多边形近似。 模式识别字母, 1994, 15.8: 743-750.
Given N ordered points in the plane and a constant M < N, an efficient algorithm is proposed to find M points, among those given, which define a polygonal curve that is a globally optimal approximation to the given points. The algorithm accommodates any properly defined error measure and the use of the most popular of these measures is studied in detail to maximize the computational efficiency. Experiments are reported showing the performance and usefulness of the proposed method.