P2P-based image recognition for component tracking in a large engineering domain

A. Amir, A. H.M. Amin, B. Srinivasan

    Research output: Chapter in Book/Report/Conference proceedingConference PaperResearch

    Abstract

    Large engineering design projects comprise very large and complex inventories of critical building components. Managing and searching components within multiple huge warehouses are complicated and time consuming tasks. Prompt availability of the components when there are needed is crucial to avoid the delay of the project, to guarantee the fast maintenance, and to ensure the smooth of the operation. Multi-Wheel Graph Neuron (mWGN) is a single-cycle, light-weight, and scalable associative-memory-based pattern recognition algorithm, and has been implemented over a structured P2P network. We propose the use of mWGN to connect multiple warehouse resources in order to facilitate a speedy and convenient searching of parts. For example an engineer can photograph a component and search to locate the part or obtain related information from using simple devices such as a mobile phone. In this paper, we show the accuracy of mWGN for component recognition and the efficiency of Chord network in linking the warehouses.

    Original languageEnglish
    Title of host publicationProceedings of the 2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering (PARENG 2011)
    EditorsP. Iványi, B.H.V. Topping
    Place of PublicationStirlingshire, UK
    PublisherCivil-Comp Press
    Number of pages14
    Volume95
    ISBN (Electronic)9781905088430, 9781905088447
    ISBN (Print)9781905088423
    DOIs
    Publication statusPublished - 2011
    EventInternational Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering 2011 - Ajaccio, Corsica, France
    Duration: 12 Apr 201115 Apr 2011
    Conference number: 2nd

    Conference

    ConferenceInternational Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering 2011
    Abbreviated titlePARENG 2011
    Country/TerritoryFrance
    CityAjaccio, Corsica
    Period12/04/1115/04/11

    Keywords

    • Associative memory
    • Component searching
    • Component tracking
    • Distributed system
    • Pattern recognition
    • Peer-to-peer

    Cite this