Int J Interact Des Manuf DOI 10.1007/s12008-017-0426-8 ORIGINAL PAPER Secondary features segmentation from high-density tessellated surfaces L. Di Angelo 1 · P. Di Stefano 1 · A. E. Morabito 2 Received: 29 June 2017 / Accepted: 10 July 2017 © Springer-Verlag France SAS 2017 Abstract A new method for secondary features segmen- tation, performed on high-density tessellated geometric models, is proposed. Four types of secondary features are considered: fillets, rounds and grooves. Sharp edges are also recognised. The method is based on an algorithm that analyses the principal curvatures. The nodes, potentially attributable to a fillet of given geometry, are those with a certain value for the maximum principal curvature. Since the deterministic application of this simple working prin- ciple shows several problems, due to the uncertainties in the curvature estimation, a fuzzy approach is proposed. In order to segment the nodes of a tessellated model belonging to secondary features of a given radius, an appropriate set of membership functions is defined and evaluated based on some parameters, which affect the quality of the curvature estimation. A region-growing algorithm connects the nodes pertaining to a same secondary feature so that, for a given radius, one or more secondary features may be recognized. The method is applied and verified in some test cases. Keywords Region growing algorithm · Computational geometry · Features extractions · Mechanical engineering computing · Fuzzy logic B A. E. Morabito annaeva.morabito@unisalento.it 1 Department of Industrial Engineering, University of L’Aquila, Giovanni Gronchi 18, 67100 L’Aquila, Italy 2 Department of Engineering for Innovation, University of Salento, Via per Arnesano, 73100 Lecce, Italy 1 Introduction Mechanical components require that some features, usu- ally called secondary features (fillets, rounds, chamfers and grooves), must be used to satisfy proper technological and functional reasons. Secondary features generally occur in the transition between two intersecting surfaces (primary geo- metric features) of the object. Although they are necessary from a functional point of view, in the semantic evaluation of a mechanical component, the secondary features are not significant since their presence, usually, do not affect the engineering intent of the object. From a geometrical point of view, secondary features are cylinders, planar surfaces or tori, but they have to be distinguished from primary features of the object that could have the same geometric proper- ties. During CAD modelling, these secondary features are added, at a later stage, to dress-up the object by using ded- icated modelling features, relying on designers’ experience or on Knowledge Based Engineering applications [1]. The importance of distinguishing correctly secondary features from primary ones is particularly important in many appli- cations. It is the case of the features recognition finalized to tolerance inspection, which, usually, does not have the scope to investigate secondary features [2]. Essentially, secondary features segmentation greatly increases the possibility to han- dle interactively the geometric model. The interaction of the user with the tessellated model is facilitated when secondary features are automatically recognized and separated from the rest of the model so that they are not confused with the pri- mary ones. The secondary feature segmentation can be also functional to edit them. Despite a wide literature is available on feature recog- nition, investigation on secondary features is a relatively new topic. All the pertinent literature concerns only the secondary features recognition from B-Rep models. The 123