High-level feature recognition using feature relationship graphs

In this paper, we present a method to extract user-specific features from common features, This contrasts with other approaches which work directly off B-rep geometric models. Here, user-specific features are called high-level features which are a set of common features combined in a user-specific manner. A feature relationship graph is used to organize common features in a part and to define high-level feature patterns. The research presented in this paper focuses mainly on feature relationship graph construction and high-level feature recognition using subgraph isomorphic techniques. (C) 2002 Elsevier Science Ltd. All rights reserved.