Signal preprocessing of deep-sea laser-induced plasma spectra for identification of pelletized hydrothermal deposits using Artificial Neural Networks

Soichi Yoshino, Blair Thornton, Tomoko Takahashi, Yutaro Takaya, Tatsuo Nozaki

    Research output: Contribution to journalArticle

    7 Citations (Scopus)

    Abstract

    This study investigates methods to analyze Laser-induced breakdown spectroscopy (LIBS) signals generated from water immersed deep-sea hydrothermal deposits irradiated by a long pulse (>100 ns) that are analyzed using Artificial Neural Networks (ANNs). ANNs require large amounts of training data to be effective. For this reason, we propose methods to preprocess full-field spectral signals into an appropriate form for ANNs artificially increase the amount of training data. The ANN was trained using a dataset of signals from immersed pelletized hydrothermal deposit samples that were preprocessed using the proposed method. The proposed method improved the accuracy of identification from 82.5% to 90.1% and significantly increased the speed of learning. The result shows that the ANN can be used to construct a generic method to identify hydrothermal deposits by long pulse underwater LIBS signals without the need for explicit peak detection.

    Original languageEnglish
    Pages (from-to)1-7
    Number of pages7
    JournalSpectrochimica Acta - Part B Atomic Spectroscopy
    Volume145
    DOIs
    Publication statusPublished - 2018 Jul 1

    Fingerprint

    plasma spectra
    preprocessing
    Deposits
    deposits
    Neural networks
    Plasmas
    Lasers
    Laser induced breakdown spectroscopy
    laser-induced breakdown spectroscopy
    lasers
    education
    deep water
    pulses
    learning
    Water

    Keywords

    • Artificial Neural Networks (ANNs)
    • Chemical analysis
    • Laser-induced breakdown spectroscopy (LIBS)
    • Signal processing

    ASJC Scopus subject areas

    • Analytical Chemistry
    • Atomic and Molecular Physics, and Optics
    • Instrumentation
    • Spectroscopy

    Cite this

    Signal preprocessing of deep-sea laser-induced plasma spectra for identification of pelletized hydrothermal deposits using Artificial Neural Networks. / Yoshino, Soichi; Thornton, Blair; Takahashi, Tomoko; Takaya, Yutaro; Nozaki, Tatsuo.

    In: Spectrochimica Acta - Part B Atomic Spectroscopy, Vol. 145, 01.07.2018, p. 1-7.

    Research output: Contribution to journalArticle

    @article{dad1280fe8bf41588ddc8eaef4bced40,
    title = "Signal preprocessing of deep-sea laser-induced plasma spectra for identification of pelletized hydrothermal deposits using Artificial Neural Networks",
    abstract = "This study investigates methods to analyze Laser-induced breakdown spectroscopy (LIBS) signals generated from water immersed deep-sea hydrothermal deposits irradiated by a long pulse (>100 ns) that are analyzed using Artificial Neural Networks (ANNs). ANNs require large amounts of training data to be effective. For this reason, we propose methods to preprocess full-field spectral signals into an appropriate form for ANNs artificially increase the amount of training data. The ANN was trained using a dataset of signals from immersed pelletized hydrothermal deposit samples that were preprocessed using the proposed method. The proposed method improved the accuracy of identification from 82.5{\%} to 90.1{\%} and significantly increased the speed of learning. The result shows that the ANN can be used to construct a generic method to identify hydrothermal deposits by long pulse underwater LIBS signals without the need for explicit peak detection.",
    keywords = "Artificial Neural Networks (ANNs), Chemical analysis, Laser-induced breakdown spectroscopy (LIBS), Signal processing",
    author = "Soichi Yoshino and Blair Thornton and Tomoko Takahashi and Yutaro Takaya and Tatsuo Nozaki",
    year = "2018",
    month = "7",
    day = "1",
    doi = "10.1016/j.sab.2018.03.015",
    language = "English",
    volume = "145",
    pages = "1--7",
    journal = "Spectrochimica Acta, Part B: Atomic Spectroscopy",
    issn = "0584-8547",
    publisher = "Elsevier",

    }

    TY - JOUR

    T1 - Signal preprocessing of deep-sea laser-induced plasma spectra for identification of pelletized hydrothermal deposits using Artificial Neural Networks

    AU - Yoshino, Soichi

    AU - Thornton, Blair

    AU - Takahashi, Tomoko

    AU - Takaya, Yutaro

    AU - Nozaki, Tatsuo

    PY - 2018/7/1

    Y1 - 2018/7/1

    N2 - This study investigates methods to analyze Laser-induced breakdown spectroscopy (LIBS) signals generated from water immersed deep-sea hydrothermal deposits irradiated by a long pulse (>100 ns) that are analyzed using Artificial Neural Networks (ANNs). ANNs require large amounts of training data to be effective. For this reason, we propose methods to preprocess full-field spectral signals into an appropriate form for ANNs artificially increase the amount of training data. The ANN was trained using a dataset of signals from immersed pelletized hydrothermal deposit samples that were preprocessed using the proposed method. The proposed method improved the accuracy of identification from 82.5% to 90.1% and significantly increased the speed of learning. The result shows that the ANN can be used to construct a generic method to identify hydrothermal deposits by long pulse underwater LIBS signals without the need for explicit peak detection.

    AB - This study investigates methods to analyze Laser-induced breakdown spectroscopy (LIBS) signals generated from water immersed deep-sea hydrothermal deposits irradiated by a long pulse (>100 ns) that are analyzed using Artificial Neural Networks (ANNs). ANNs require large amounts of training data to be effective. For this reason, we propose methods to preprocess full-field spectral signals into an appropriate form for ANNs artificially increase the amount of training data. The ANN was trained using a dataset of signals from immersed pelletized hydrothermal deposit samples that were preprocessed using the proposed method. The proposed method improved the accuracy of identification from 82.5% to 90.1% and significantly increased the speed of learning. The result shows that the ANN can be used to construct a generic method to identify hydrothermal deposits by long pulse underwater LIBS signals without the need for explicit peak detection.

    KW - Artificial Neural Networks (ANNs)

    KW - Chemical analysis

    KW - Laser-induced breakdown spectroscopy (LIBS)

    KW - Signal processing

    UR - http://www.scopus.com/inward/record.url?scp=85045047586&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85045047586&partnerID=8YFLogxK

    U2 - 10.1016/j.sab.2018.03.015

    DO - 10.1016/j.sab.2018.03.015

    M3 - Article

    AN - SCOPUS:85045047586

    VL - 145

    SP - 1

    EP - 7

    JO - Spectrochimica Acta, Part B: Atomic Spectroscopy

    JF - Spectrochimica Acta, Part B: Atomic Spectroscopy

    SN - 0584-8547

    ER -