摘要
三维点云配准是逆向工程中的关键。为提配准确度,提出了一种基于曲率特征点的ICP改进算法。算法在初始配准的基础上,采用二次曲面逼近的方法求得每一点的方向矢量和曲率,利用据曲率确定特征点集,并根据方向矢量调整对应关系,减少了ICP算法的搜索量,提高了ICP算法的效率。针对目标函数,引入Niloy坐标框架,可以根据点云距离调整收敛速度和配准精度。改进后的算法在精确度基本不受影响的情况下提高了配准速度,进行仿真实验。实验验证了配准效果和算法的稳定性。
The registration of 3D point clouds is the key problem in 3D surface reverse.A registration method of 3D point clouds based feature points is put forward.Based on the initial registration,the curvature of each point was estimated according to the point and its neighbor points.Curvature is determined in accordance with the characteristics of point set and correlation is adjusted in accordance with the direction of the vector,which decrease the searching load of ICP and improve the efficiency of the ICP algorithm.The accuracy of improved algorithm will not be affected to improve the matching speed.The effect of the algorithm is verified in the applications.
出处
《计算机仿真》
CSCD
北大核心
2010年第8期235-238,共4页
Computer Simulation
关键词
点云配准
迭代最近点
曲率
四元数
Point cloud alignment
ICP algorithm
Curvature
Quaternion
