摘要
针对目前Li DAR获取的点云数据配准中ICP算法迭代计算效率较低和使用标靶配准时精扫标靶费时费力等问题,提出利用扫描地物所包含的平面特征及点云数据的离散特性,通过拟合平面得到平面的单位法向量进行旋转角的求解。由于点云数据中含有误差,使用拟合平面的法向量不仅避免了对标靶的精细扫描,而且也消弱了点云误差对转换参数的影响。最后通过实例验证了本文方法的可行性与严密性。
In view of the present LiDAR to obtain the point cloud data registration, the ICP algorithm iterative calculation efficiency is low and target with fine scanning consuming much time, put forward take the advantage of plane characteristics of scanning object and the discrete characteristics of point cloud data, the unit normal vector of plane is obtained by fitting plane and through this got the rotation angle. At last, the feasibility and rigor of the presented approach are verified by examples.
出处
《测绘与空间地理信息》
2015年第5期202-205,210,共5页
Geomatics & Spatial Information Technology