Automated feature recognition from 2D drawings

S. Meeran, M. J. Pratt

Research output: Contribution to journalArticlepeer-review


Automated feature recognition is likely to be an essential requirement for future integrated design and manufacturing systems. Most methods which have so far been proposed use as their input a solid model which represents only the purely geometric aspects of the design information. Much research is currently in progress on the augmentation of such models with technological and other lifecycle data to give what is referred to as a product model. The presence in such a model of design feature data, in particular, may make the recognition of manufacturing features easier in the future. On the other hand, little previous work has been reported on feature recognition from 2D drawings, which is the form in which the majority of design data currently exists. The paper describes an experimental system for this purpose. It has been designed to meet the automated process-planning requirements of today rather than tomorrow, and it has been demonstrated on a range of simple prismatic machined parts.

Original languageEnglish
Pages (from-to)7-17
Number of pages11
JournalComputer-Aided Design
Issue number1
Publication statusPublished - 1 Jan 1993


  • 2D drawings
  • automated feature recognition
  • NC machining
  • neutral file format

ASJC Scopus subject areas

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


Dive into the research topics of 'Automated feature recognition from 2D drawings'. Together they form a unique fingerprint.

Cite this