Rotor surface defect and symmetry detection based on point cloud registration

Abstract

During the production process of the blower rotor, detecting its surface defects and symmetry can avoid the safety hazards of the rotor during operation. Therefore, this paper has improved the traditional ICP algorithm. For the rotor point cloud collected by the 3D measuring instrument, the standard point cloud and the test point cloud are unified in the same coordinate system, and the incomplete rotor surface is collected. The point cloud is locally registered, and stitching is completed, and then it is registered with the standard point cloud. The defect detection of the point cloud is completed by comparing the unevenness of the two-point clouds, and the symmetry detection is completed by comparing the shape features of the two-point clouds.