matlab 里catmull rom,Catmull_Rom算法的实现

Catmull_Rom算法的具体原理在网络上都有，感兴趣的可以自己查找，此处只是它的代码实现，实现环境为OSG：

给定一堆离散点，将该离散点进行插值，绘制的曲线经过每一个离散点。

实现的代码：

//生成Catmull_Rom Spline算法顶点

osg::Vec3 Catmull_Rom_SplinePoint_new(float u, osg::ref_ptrsubconctrl_P)

{

float f0, f1, f2, f3;

osg::Vec3 vert;

f0 = (-0.5)*pow(u, 3) + pow(u, 2) + (-0.5*u);

f1 = 1.5*pow(u, 3) - 2.5*pow(u, 2) + 1;

f2 = (-1.5)*pow(u, 3) + 2.0*pow(u, 2) + 0.5*u;

f3 = 0.5*pow(u, 3) - 0.5*pow(u, 2);

vert = subconctrl_P->at(0)*f0 + subconctrl_P->at(1)*f1 + subconctrl_P->at(2)*f2 + subconctrl_P->at(3)*f3;

return vert;

}

//生成Catmull_Rom Spline样条线顶点,过端点

osg::ref_ptrCatmull_Rom_SplinePoints_new(osg::ref_ptrconctrl_P, int Unum = 10)

{

osg::ref_ptrC_Line_vertices = new osg::Vec3Array;

osg::ref_ptrNew_conctrl_P = new osg::Vec3Array;

auto pointsnum = conctrl_P->size();

auto P0 = conctrl_P->at(0) * 2 - conctrl_P->at(1);

New_conctrl_P->push_back(P0);

auto Pn = conctrl_P->at(pointsnum - 1) * 2 - conctrl_P->at(pointsnum - 2);

for (int i = 0; i < pointsnum; i++)

{

New_conctrl_P->push_back(conctrl_P->at(i));

}

New_conctrl_P->push_back(Pn);

pointsnum += 2;

for (int i = 0; i < pointsnum; i++)

{

if (i < pointsnum - 3)

{

osg::ref_ptrsubcontrl = new osg::Vec3Array;

subcontrl->push_back(New_conctrl_P->at(i));

subcontrl->push_back(New_conctrl_P->at(i + 1));

subcontrl->push_back(New_conctrl_P->at(i + 2));

subcontrl->push_back(New_conctrl_P->at(i + 3));

float deltaU = 1.0 / float(Unum);

for (int j = 0; j < Unum; j++)

{

auto Uu = j*deltaU;

auto tempp = Catmull_Rom_SplinePoint_new(Uu, subcontrl);

C_Line_vertices->push_back(tempp);

}

}

}

return C_Line_vertices;

}

实现效果：

蓝色是生成的曲线，红色的线作为对照，是用于插值的离散点