Robust Sonar ATR with pose corrected sparse reconstruction-based classification

John McKay, Vishal Monga, Raghu Raj

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

    2 Scopus citations

    Abstract

    Sonar imaging has seen vast improvements over the last few decades due in part to advances in synthetic aperture Sonar (SAS). Because of this, sophisticated classification techniques originally developed for other tasks can be used in Sonar automatic target recognition (ATR) to locate mines and other threatening objects. Among the most promising of these methods is sparse reconstruction-based classification (SRC) which has shown an impressive resiliency to noise, blur, and occlusion even in settings with little training. We present a coherent strategy for using SRC for Sonar ATR that retains SRC's robustness while also being able to handle targets with diverse geometric arrangements. Our method, pose corrected sparsity (PCS), incorporates state-of-the-art dictionary learning schemes on localized block extractions which we show produces compelling classification results on the RAWSAS dataset.

    Original languageEnglish (US)
    Title of host publicationOCEANS 2016 MTS/IEEE Monterey, OCE 2016
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    ISBN (Electronic)9781509015375
    DOIs
    StatePublished - Nov 28 2016
    Event2016 OCEANS MTS/IEEE Monterey, OCE 2016 - Monterey, United States
    Duration: Sep 19 2016Sep 23 2016

    Publication series

    NameOCEANS 2016 MTS/IEEE Monterey, OCE 2016

    Other

    Other2016 OCEANS MTS/IEEE Monterey, OCE 2016
    CountryUnited States
    CityMonterey
    Period9/19/169/23/16

    All Science Journal Classification (ASJC) codes

    • Instrumentation
    • Oceanography
    • Ocean Engineering

    Fingerprint Dive into the research topics of 'Robust Sonar ATR with pose corrected sparse reconstruction-based classification'. Together they form a unique fingerprint.

    Cite this