• DocumentCode
    3327257
  • Title

    Registering aerial photographs of farmland with satellite imagery

  • Author

    Saleem, Sajid ; Bais, Abdul ; Khawaja, Yahya M.

  • Author_Institution
    Dept. of Electr. Eng., N.W.F.P. Univ. of Eng. & Technol. Peshawar, Peshawar, Pakistan
  • fYear
    2010
  • fDate
    26-29 Sept. 2010
  • Firstpage
    945
  • Lastpage
    948
  • Abstract
    This paper presents a novel approach for registration of aerial and satellite farmland imagery. We have used field junctions as features for registration. Geometric properties of a group of neighboring fields connected to the junction are used as junction descriptor. These properties are calculated by fitting a rectangle to the field. Junction descriptor contains a group of field patterns capable of finding its match and indexing aerial image spatial location in satellite images in spite of repetitive structure of fields. It is invariant to in-plane rotation, scale and translation. Experimental results demonstrate the robustness of our approach against the junction localization errors and the errors introduced by splitting and/or merging of fields.
  • Keywords
    feature extraction; geophysical image processing; image matching; image registration; vegetation mapping; vocabulary; aerial photograph registeration; farmland photograph; feature extraction; image matching; image spatial location indexing; junction descriptor; rectangle fitting; satellite farmland imagery; Junctions; Merging; Pixel; Roads; Robustness; Satellites; Silicon; Registration; farmland image; field junction; junction descriptor; rectangle fitting;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Image Processing (ICIP), 2010 17th IEEE International Conference on
  • Conference_Location
    Hong Kong
  • ISSN
    1522-4880
  • Print_ISBN
    978-1-4244-7992-4
  • Electronic_ISBN
    1522-4880
  • Type

    conf

  • DOI
    10.1109/ICIP.2010.5651111
  • Filename
    5651111