摘要
针对存在粗差或异常数据点时,最小二乘定位方法会产生定位错误的情况,本文提出了基于M-估计的稳健标靶球定位方法。通过M-估计剔除或减弱球体数据中异常点对参数估计的影响,获取稳健的标靶球球心坐标。利用仿真数据与真实数据进行验证,结果表明该法能克服异常值的影响,提高球体定位的准确度,具有稳健性。
Registraton sphere is one of the most important tools in terrestrial laser scanning. Different point clouds form different views, which can be transformed into the same coordinate system frame. But, the key point is the determination of sphere center. Because the traditional poor estimated parameter. Therefore, a new way to determine sphere center is given in this paper. This method is based on M-estimation. And the estimated parameters won＇t be affected by outliers of point clouds. Related experiments have been conducted, in which simulation data and real data were used. The results show that this new method can overcome the effect of outliers, and improve the accuracy of sphere center coordinates. Compared with the least square estimation, it should have the advantage of robust.
出处
《工程勘察》
CSCD
北大核心
2008年第10期42-45,49,共5页
Geotechnical Investigation & Surveying
基金
国家自然科学基金(40574008)
江西省教育厅科技项目(赣教技字[2007]237号)
江西省自然科学基金资助项目(0411005)
国防科工委资助项目子项目"激光扫描仪检校和作业方法研究"
关键词
地面激光扫描
点云
标靶球
M-估计
稳健性
terrestrial laser scanning
point clouds
registration sphere
M-estimation
robust
