TY - JOUR N2 - 3D scanning measurements are gaining popularity every year. Quick inspections on already captured point clouds are easy to prepare with the use of modern software and machine learning. To achieve repeatability and accuracy, some surface and measurement issues should be considered and resolved before the inspection. Large numbers of manufacturing scans are not intended for manual correction. This article is a case study of a small surface inspection of a turbine guide vane based on 3D scans. Small surface errors cannot be neglected as their incorrect inspection can result in serious faults in the final product. Contour recognition and deletion seem to be a rational method for making a scan inspection with the same level of accuracy as we have now for CMM machines. The main reason why a scan inspection can be difficult is that the CAD source model can be slightly different from the inspected part. Not all details are always included, and small chamfers and blends can be added during the production process, based on manufacturing standards and best practices. This problem does not occur during a CMM (coordinate measuring machine) inspection, but it may occur in a general 3D scanning inspection. L1 - http://journals.pan.pl/Content/120735/PDF-MASTER/23_02298_Bpast.No.69(5)_drukM.pdf L2 - http://journals.pan.pl/Content/120735 PY - 2021 IS - 5 EP - e138815 DO - 10.24425/bpasts.2021.138815 KW - 3d scan KW - flatness KW - turbine guide vane KW - small surfaces KW - point clouds KW - contour recognition KW - contour de-featuring A1 - Jamontt, Marcin A1 - Pyrzanowski, Paweł VL - 69 DA - 15.09.2021 T1 - 3D scan contour de-featuring for improved measurement accuracy – a case study for a small turbine guide vane component SP - e138815 UR - http://journals.pan.pl/dlibra/publication/edition/120735 T2 - Bulletin of the Polish Academy of Sciences Technical Sciences ER -