Fillets, rounds, grooves and sharp edges segmentation from 3D scanned surfaces. (May 2019)

• Record Type: Journal Article
• Authors: Di Angelo, L.; Di Stefano, P.; Morabito, A.E.
• Is Part Of: Computer aided design. Volume 110(2019)
• Page Start: 78
• Page End: 91
• Publication Date: 2019-05
• Subjects: Blending features -- Features recognition -- 3D mesh segmentation -- STL -- 3D scanning -- Fuzzy logic; Computer-aided design -- Periodicals; Engineering design -- Data processing -- Periodicals; Computer graphics -- Periodicals; Conception technique -- Informatique -- Périodiques; Infographie -- Périodiques; Computer graphics; Engineering design -- Data processing; Periodicals; Electronic journals; 620.00420285
• DOI: 10.1016/j.cad.2019.01.003
• Languages: English
• ISSNs: 0010-4485

Abstract

Abstract: Fillets, rounds, chamfers and grooves are secondary features which are typically present in real manufactured mechanical components to satisfy some manufacturing and functional requirements. Despite the broad array of research conducted on feature recognition, the investigation of secondary features is a relatively new topic. All of the pertinent studies have been focusedonly on the recognition of secondary features from B-Rep models. The recognition and segmentation of secondary features from a discrete model is a non-trivial problem due to the same geometric descriptors that may be applied to both primary and secondary features. Moreover, although in real-world mechanical parts primary features are planes, cylinders or cones, the secondary features may be non-analytical and complex-shaped geometries. Further sources of uncertainty are the measurement errors and non-ideal geometries of the real objects to which the method is applied. To overcome these problems, a new and original method to segment secondary features of tessellated geometric models is proposed. The method is based on the analysis of geometric-differential properties and provides specific strategies that reduce its sensitivity to all of the above-mentioned uncertainties without affecting its selectivity. The proposed method, described in detail in this paper, is tested in some very critical cases, and the results are presented and discussed. Highlights: A new method for secondary feature recognition from 3D scanned object is proposed. The method works from not well sampled and noisy 3D mesh. The method requires just a few setting parameters that are not critical. High-level semantic description of a scanned object is automatically generated.