• DocumentCode
    2010617
  • Title

    A new ground penetrating radar signal analysis method based on S transform

  • Author

    Weifeng, Lang ; Mingxing, Lin

  • Author_Institution
    Mech. Eng. Sch., Shandong Univ., Jinan, China
  • Volume
    3
  • fYear
    2010
  • fDate
    17-18 July 2010
  • Firstpage
    683
  • Lastpage
    686
  • Abstract
    The paper discussed the characteristics of GPR signals of one object and two objects respectively, and studied the rule of extracting twin-peak attribute when the size of object or the distance of two objects was changing through GST. For one object, the two reflections can´t be resolved in time when the diameter of object is less than or equal to 1/3 dominant wavelength within object medium, but the twin-peak attribute can resolve the reflections when the diameter is as small as 1/7 dominant wavelength, so the size can be estimated combining the attribute with time-domain waveform and the definition of a small object can be given. For two objects that each is a small object in time-frequency domain, we considered the situations where the permittivity were same or different, and it is pointed out that the attribute can improve the resolution of two objects and can estimate the distance combined with time-domain waveform. The selection rule of parameter and frequency point of twin-peak attribute was given as well so the unsystematic problem was solved.
  • Keywords
    Fourier transforms; ground penetrating radar; radar signal processing; time-domain analysis; S transform; ground penetrating radar signal analysis method; time-domain waveform; Ground penetrating radar; Permittivity; Reflection; Signal resolution; Time domain analysis; Time frequency analysis; Transforms; S transform; ground penetrating radar; twin-peak attribute;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Environmental Science and Information Application Technology (ESIAT), 2010 International Conference on
  • Conference_Location
    Wuhan
  • Print_ISBN
    978-1-4244-7387-8
  • Type

    conf

  • DOI
    10.1109/ESIAT.2010.5568442
  • Filename
    5568442