A hybrid approach for identification of 3D features from CAD database for manufacturing support

A number of approaches for identification of 3D features are available toda y including syntactic recognition methods, graph based methods, expert syst ems and knowledge based approach. Work done on all such approaches shows la ck of generalisation, difficulty in implementation or little demonstrated c omputational viability. In this paper, a hybrid approach for non-interactin g 3D-feature identification from CAD database has been proposed, which is p artly mathematically iterative and partly based on heuristic rules. Initial ly a void non-polyhedral pocket feature has been undertaken to establish th e procedures and later this is extended for all features in general. A coup le of test results are presented with CPU times and inherent heuristics. Th e process is applicable for process planners and manufacturers in industrie s. (C) 2001 Elsevier Science Ltd. All rights reserved.