针对移动三维激光扫描中分段点云数据闭合时首尾测段重叠区域存在闭合差的问题，分析了点云数据配准方法与原理，以点云拼接的转换参数为切入点，在粗略配准结果的基础上，顾及点云数据的整体性和关联性，提出了以测段间转换关系为参数的整体平差方法，实现了闭合差在各测段中的分配和改正。将改正数应用于转换参数中，对点云进行纠正，纠正结果与实测校核点的点位较差均在10 cm以内，数据质量得到了极大改善，证明了方法的可行性，具有较高的实用价值。

Addressing the issue of closure errors in the overlapping areas of initial and final survey segments during the closure of segmented point cloud data in mobile 3D laser scanning, we analyzed the methods and principle of point cloud data registration. Focusing on the transformation parameters for point cloud stitching, based on the preliminary registration results, considering the integrity and correlation of point cloud data, we put forward a global registration method taking the transformation relationship between survey segments as parameters, enabling the distribution and correction of closure errors across various survey segments. We applied the correction values into the transformation parameters to adjust the point cloud. The point position discrepancies of corrected results and actual verification points were all within 10 cm, significantly improving data quality, which could demonstrate the feasibility of method and its high practical value.