In recent years, it has become easier to save real objects as digital data using 3D scanning technology. Point cloud data obtained by scanning may contain missing parts due to occlusion or reflection of the object being measured. This paper proposes a method of detecting missing parts. The method generates new points using morphological operations on the point cloud and detects missing parts by analyzing the surface of the generated points. The structuring element of morphological operations is a sphere shape point cloud. By specifying the radius of the sphere, users can detect missing parts of any size. In addition, the processing time is reduced by introducing a voxel model. We verified the effectiveness of the proposed method through several experiments.

In order to prevent global warming, there is a need to reduce emissions of carbon dioxide, a greenhouse gas, and there is a demand for compressors for refrigeration and air conditioning to be more efficient and reliable. This paper focuses on scroll compressors and discusses weld deformation, which affects both efficiency and reliability. A scroll compressor consists of a cylindrical casing with a lid cap and bottom cap welded circumferentially to both ends, in which the compression element and electric motor are hermetically sealed and mounted. The frame and bottom frame, which comprise the bearings, are plug-welded to the casing, and a reduction in the coaxiality of these two parts is a problem that directly affects the performance and reliability of the compressor. Analytical investigation of the weld deformation of a scroll compressor revealed the following. To ensure the coaxiality of the frame and lower frame, it is important that the clearance between the casing and the frame/lower frame during plug welding be uniform. And if the clearance is not uniform, the frame and lower frame will tilt due to weld deformation, resulting in deterioration of the coaxiality.