Two-stage point cloud registration for 3D measurement of large workpieces

3D measurement is an important method to ensure the precision of large-size workpiece.Due to the large spatial range of the large-size workpiece, the three-dimensional reconstruction of a single perspective cannot obtain its comprehensive three-dimensional measurement information. Therefore, it is necessary to join multiple perspective point clouds through point cloud registration to expand the measurement range of the workpiece.This paper proposes a two-stage point cloud registration algorithm, which can achieve efficient and high-precision point cloud stitching.At first, laser points are used as labels to achieve the first stage of point cloud registration. Then, AA-ICP is used to fine-tune the point cloud after registration.After two stages of point cloud registration, the proposed algorithm achieves the goal of measuring large size workpiece with high precision in a wide range.