舱容计量中高精度点云配准研究

Research on High Precision Registration of Point Clouds

为了提高点云配准精度,研究了基于特征的配准方法。介绍了点云配准的六参数法。基于特征的配准方法利用特征点与周围特征的方位、距离等几何条件进行附加约束,实现点云配准。方法适用性强,可通过利用同名反射标志或直接利用点云数据实现,最后采用稳健模型参数估计算法——RANSAC算法提高配准参数估计的稳健性和可靠性。

In order to improve the accuracy of point cloud registration, a new registration method based on feature was applied in this paper. The basic method of point cloud registration six parameter method was also mentioned. By the method of shape matching method ,feature points and the surrounding features range were used and distance geometry constraints were added, so the registration of point clouds was achieved by constraining conditions. The applicability of the method was well through the use of the reflection sign or directly using the point cloud data. At the end of this paper,a robust model parameter estimation algorithm--RANSAC algorithm had been proposed to improve the robustness and reliability of parameter estimation.