基于凸包算法和抗差最小二乘法的激光扫描仪圆形标靶中心定位

Center location of circular target for TLS based on convex hull algorithm and robust least squares algorithm

针对激光扫描仪圆形平面标靶点云数据缺失或冗余时不易准确获取标靶中心的问题,提出一种稳健的中心定位算法。首先根据回光强度提取标靶点云数据;然后剔除标靶点云数据粗差点,拟合标靶的最佳平面;接着利用凸包算法提取标靶边缘点;最后利用抗差最小二乘求解标靶的中心坐标。实验表明,该算法定位精度能达到亚毫米级,可以有效解决标靶数据缺失或者冗余时标靶定位精度低的问题,提高标靶的中心定位的稳健性。

To solve the problem which is hard to acquire the accurate center of target when the point cloud is missing or redundant on the circular plane target of laser scanner,a robust center location algorithm is presented.Firstly,the point cloud of planar target is extracted based on return light intensity. Secondly,the noise points are eliminated and the best fitting planar is calculated. Thirdly,edge points are extracted based on convex hull algorithm. Finally, the coordinates of target center are calculated based on robust least squares algorithm.Experiments show that the locating precision of this algorithm is under millimeter and its efficiency is higher than Reference[8 ],which can also ensure the locating precision when the point cloud of planar target is missing or redundant,and improve the robustness of center location.