M估计的稳健平面标靶中心定位算法

标靶在地面激光扫描仪多测站点云配准，以及在全球卫星导航信号不良状况下改善移动激光雷达系统精度具有重要作用，标靶的精确提取与定位是关键问题之一。本文提出基于M估计的标靶中心定位方法：先根据标靶与周围环境反射强度差异，构建强度变化率特征，提取标靶边缘点云；再对标靶候选区域点云进行主成分分析，计算标靶平面法向量，对标靶边缘点云进行投影变换；最后通过M估计选权迭代法，剔除标靶噪声与粗差点云，拟合标靶形状，精确定位标靶中心。试验表明，本文算法能够在标靶存在冗余、缺损状态时排除非标靶圆形边缘的粗差点，获得高精度标靶中心。

Robust Planar Target Location Algorithm Based on M Estimation

Target plays an important role in the multi-site cloud registration of terrestrial laser scanner and the improvement of the precision of mobile LiDAR (Light detection and ranging) system under the bad condition of GNSS (Global Navigation Satellite System) signal. One of the key problem is to extract and locate the target accurately. In this paper, a target center location method based on M estimation is proposed. Firstly, the characteristic of the rate of intensity change is constructed to extract the edge point cloud of the target according to the difference between the target and the surrounding environment. Then, the principal component analysis is carried out to the point cloud of the target region, the target plane normal vector is calculated and the projection transformation is used to the edge point cloud of the target. Finally, M estimation, the method of weighted iteration, is used to eliminate the target noise and outlier cloud points, fit the target shape and locate the target center accurately. The experiments show that the algorithm can eliminate the gross errors of the non-standard circular target edges and obtain a high precision target center when the target is redundant and defective.