Defect Characterization on Complex Shape Aeronautical Parts via 3D Point Cloud Processing

This paper presents an approach to characterizing defects - a process that consists of accurate measurement of their geometric properties such as depth and surface area, assuming that the defect detection process has been successfully performed previously. Our methodology for addressing this problem involves three key steps. The first is to reconstruct an ideal or defect-free surface using scattered points obtained from a point cloud scan of the inspected part. Subsequently, the distance from each cloud point to this surface is computed, and points at a distance higher than a specified threshold are identified as defect points. The maximum of these distances corresponds to the depth of the defect. Finally, the minimal 3D bounding box encapsulating the defect points is determined, where the two largest dimensions of this box represent the length and width of the identified defect.