Free-form feature representation and comparison

David Ferrari*, Geoff West, Bernard Rolfe, Michael Cardew-Hall

*Corresponding author for this work

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    Abstract

    Knowledge Management systems have been applied to several areas in the manufacturing industry. These applications have been predominately text based, however, much of manufacturing's information is related to geometric features. In this paper we describe a representation and content-based search method such that given a geometric specification, the knowledge of previous designs having similar feature requirements is returned. The technique uses curvature based shape index segmentation to construct an attributed graph-like descriptor for 2.5D free-form surface features, such as those found on stamped components. Efficient edit distance methods are applied for search. Experiments on a database of freeform feature s indicate feasible runtimes with favourable results.

    Original languageEnglish
    Title of host publicationProceedings of 2006 ASME International Design Engineering Technical Conferences and Computers and Information In Engineering Conference, DETC2006
    PublisherAmerican Society of Mechanical Engineers
    ISBN (Print)079183784X, 9780791837849
    DOIs
    Publication statusPublished - 2006
    Event2006 ASME International Design Engineering Technical Conferences and Computers and Information In Engineering Conference, DETC2006 - Philadelphia, PA, United States
    Duration: 10 Sept 200613 Sept 2006

    Publication series

    NameProceedings of the ASME Design Engineering Technical Conference
    Volume2006

    Conference

    Conference2006 ASME International Design Engineering Technical Conferences and Computers and Information In Engineering Conference, DETC2006
    Country/TerritoryUnited States
    CityPhiladelphia, PA
    Period10/09/0613/09/06

    Fingerprint

    Dive into the research topics of 'Free-form feature representation and comparison'. Together they form a unique fingerprint.

    Cite this