Title: Feature recognition from mesh models

Authors: Deepali Tatkar; Govind Kelkar; Venkatesh Kamat

Addresses: Department of Computer Science and Technology, Goa University, India ' Department of Computer Science and Technology, Goa University, India ' Department of Computer Science and Technology, Goa University, India

Abstract: In this paper, we propose a methodology for manufacturing feature recognition from a segmented triangulated mesh model. Proposed methodology has two phases, segment preprocessing and feature recognition. Input to the algorithm is triangulated mesh model which is simplified by segmenting the mesh model into high level regions approximated by a simple primitive such as planes, sphere and cylinders. In segment preprocessing phase, we gather adjacency information related to each identified primitive. For each feature to be identified, feature rules are defined using geometric properties of approximated primitives. Finally, feature recognition phase checks if any of the connected set of primitives satisfy the feature rules and highlights the recognised feature. At present, feature recognition is restricted only to simple features composed of plane primitives such as pocket, step, slot and cylindrical primitive such as holes. Given a segmented mesh model as an input, the algorithm automatically recognises different manufacturing features present in model. This extracted features information then can be used for downstream CAD/CAM application such as process planning, cost estimation and generating feature tree.

Keywords: CAD mesh models; reverse engineering; approximated primitives; segmented mesh; manufacturing features; feature recognition; segment preprocessing; feature extraction; CADCAM; process planning; cost estimation; feature trees.

DOI: 10.1504/IJCAET.2017.10003933

International Journal of Computer Aided Engineering and Technology, 2017 Vol.9 No.3, pp.289 - 306

Received: 07 Apr 2014
Accepted: 14 Aug 2014

Published online: 27 Mar 2017 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article