Automatic form-feature recognition using neural-network-based techniques on boundary representations of solid models

S. Prabhakar, Mark Henderson

Research output: Contribution to journalArticle

103 Citations (Scopus)

Abstract

A new technique for performing form-feature recognition using the principles of neural networks is discussed. Neural nets require parallel input of data, which, in this case, are B-rep solid models of parts. An input format has been developed which includes face descriptions and face-face relationships. An algorithm for recognition using neural-net-based techniques has been developed, and a suitable net architecture, which is similar to the multilayer perceptron in function and which implements the algorithm, has been designed. The net architecture is described, and a few examples are presented which highlight the strengths and weaknesses of the recognition algorithm.

Original languageEnglish (US)
Pages (from-to)381-393
Number of pages13
JournalComputer-Aided Design
Volume24
Issue number7
DOIs
StatePublished - 1992

Fingerprint

Feature Recognition
Boundary Representation
Solid Model
Neural Nets
Face
Neural Networks
Neural networks
Recognition Algorithm
Multilayer neural networks
Perceptron
Multilayer
Form
Architecture

Keywords

  • boundary representation
  • feature recognition
  • geometric reasoning
  • neural nets
  • pattern recognition
  • solid modelling

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Industrial and Manufacturing Engineering
  • Geometry and Topology

Cite this

Automatic form-feature recognition using neural-network-based techniques on boundary representations of solid models. / Prabhakar, S.; Henderson, Mark.

In: Computer-Aided Design, Vol. 24, No. 7, 1992, p. 381-393.

Research output: Contribution to journalArticle

@article{6fea740092d7450a9da5926d6431c515,
title = "Automatic form-feature recognition using neural-network-based techniques on boundary representations of solid models",
abstract = "A new technique for performing form-feature recognition using the principles of neural networks is discussed. Neural nets require parallel input of data, which, in this case, are B-rep solid models of parts. An input format has been developed which includes face descriptions and face-face relationships. An algorithm for recognition using neural-net-based techniques has been developed, and a suitable net architecture, which is similar to the multilayer perceptron in function and which implements the algorithm, has been designed. The net architecture is described, and a few examples are presented which highlight the strengths and weaknesses of the recognition algorithm.",
keywords = "boundary representation, feature recognition, geometric reasoning, neural nets, pattern recognition, solid modelling",
author = "S. Prabhakar and Mark Henderson",
year = "1992",
doi = "10.1016/0010-4485(92)90064-H",
language = "English (US)",
volume = "24",
pages = "381--393",
journal = "CAD Computer Aided Design",
issn = "0010-4485",
publisher = "Elsevier Limited",
number = "7",

}

TY - JOUR

T1 - Automatic form-feature recognition using neural-network-based techniques on boundary representations of solid models

AU - Prabhakar, S.

AU - Henderson, Mark

PY - 1992

Y1 - 1992

N2 - A new technique for performing form-feature recognition using the principles of neural networks is discussed. Neural nets require parallel input of data, which, in this case, are B-rep solid models of parts. An input format has been developed which includes face descriptions and face-face relationships. An algorithm for recognition using neural-net-based techniques has been developed, and a suitable net architecture, which is similar to the multilayer perceptron in function and which implements the algorithm, has been designed. The net architecture is described, and a few examples are presented which highlight the strengths and weaknesses of the recognition algorithm.

AB - A new technique for performing form-feature recognition using the principles of neural networks is discussed. Neural nets require parallel input of data, which, in this case, are B-rep solid models of parts. An input format has been developed which includes face descriptions and face-face relationships. An algorithm for recognition using neural-net-based techniques has been developed, and a suitable net architecture, which is similar to the multilayer perceptron in function and which implements the algorithm, has been designed. The net architecture is described, and a few examples are presented which highlight the strengths and weaknesses of the recognition algorithm.

KW - boundary representation

KW - feature recognition

KW - geometric reasoning

KW - neural nets

KW - pattern recognition

KW - solid modelling

UR - http://www.scopus.com/inward/record.url?scp=0001066659&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0001066659&partnerID=8YFLogxK

U2 - 10.1016/0010-4485(92)90064-H

DO - 10.1016/0010-4485(92)90064-H

M3 - Article

VL - 24

SP - 381

EP - 393

JO - CAD Computer Aided Design

JF - CAD Computer Aided Design

SN - 0010-4485

IS - 7

ER -