• DocumentCode
    2409394
  • Title

    An Improved EMD Analysis Method for GPR Images

  • Author

    Gan, Lu ; Zhou, Long ; You, Xinge

  • fYear
    2011
  • fDate
    21-23 Oct. 2011
  • Firstpage
    104
  • Lastpage
    107
  • Abstract
    The EMD method can be adopted to analyze the GPR echo image because it is adaptive for non-stationary and non-linear signals. However, the direct wave in GPR signal contains considerable energy which will generate false object, even aggravate the end effect. And the noise in the signal can influence EMD analysis for further characteristics extraction. Aiming at the problems above, an improved EMD analysis method for GPR images is put forward. Above all, the direct wave is suppressed in mean method, and then reducing the noise based on LWT. Finally it should carry on the EMD analysis on the new image. Through contractive experiment, the result shows the improved method can avoid the false information, furthermore, the IMF components contain more precise object information.
  • Keywords
    Ground penetrating radar; Mathematical model; Noise; Noise reduction; Time frequency analysis; Wavelet transforms; EMD; HHT; denoise; direct wave;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computational and Information Sciences (ICCIS), 2011 International Conference on
  • Conference_Location
    Chengdu, China
  • Print_ISBN
    978-1-4577-1540-2
  • Type

    conf

  • DOI
    10.1109/ICCIS.2011.77
  • Filename
    6086144