• DocumentCode
    1499612
  • Title

    Ultrawide-bandwidth fully-polarimetric ground penetrating radar classification of subsurface unexploded ordnance

  • Author

    Chi-Chih Chen ; Higgins, M.B. ; O´Neill, K. ; Detsch, R.

  • Author_Institution
    ElectroSci. Lab., Ohio State Univ., Columbus, OH, USA
  • Volume
    39
  • Issue
    6
  • fYear
    2001
  • fDate
    6/1/2001 12:00:00 AM
  • Firstpage
    1221
  • Lastpage
    1230
  • Abstract
    An ultrawide-bandwidth (UWB) ground penetrating radar (GPR) was used to collect fully-polarimetric backscattered data from 10 to 800 MHz using a dual-polarization version of the previously developed dielectric-loaded horn-fed bow-tie (HFB) antenna. Special processing algorithms were developed to extract the polarization, orientation, depth, and length features of the target under investigation. The polarization and resonance features were utilized to discriminate subsurface ordnance from false-alarm objects that do not have elongated bodies. The classification of a specific type of unexploded ordnance (UXO) was also performed using the known length information. The processed results obtained from an initial blind field test show a very encouraging discrimination performance.
  • Keywords
    buried object detection; geophysical techniques; military systems; radar polarimetry; radar theory; remote sensing by radar; terrain mapping; terrestrial electricity; 10 to 800 MHz; algorithm; backscatter; buried object detection; classification; dielectric-loaded horn-fed bow-tie antenna; dual-polarization; geoelectric method; geophysical measurement technique; ground penetrating radar; land surface; landmine; military system; mine detection; radar polarimetry; radar remote sensing; subsurface unexploded ordnance; terrain mapping; terrestrial electricity; ultrawide-bandwidth; unexploded ordnance; Data mining; Electromagnetic scattering; Feature extraction; Ground penetrating radar; Magnetic sensors; Polarization; Radar antennas; Radar detection; Radar scattering; Testing;
  • fLanguage
    English
  • Journal_Title
    Geoscience and Remote Sensing, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0196-2892
  • Type

    jour

  • DOI
    10.1109/36.927444
  • Filename
    927444