• DocumentCode
    2383297
  • Title

    Coherent change detection in SAR images of harbors with emphasis on findings from container backscattering

  • Author

    Johnsen, Terje

  • Author_Institution
    Norwegian Defence Res. Establ. (FFI), Kjeller, Norway
  • fYear
    2011
  • fDate
    23-27 May 2011
  • Firstpage
    118
  • Lastpage
    123
  • Abstract
    The use of Coherent Change Detection (CCD) in analysis of TerraSAR-X images with high fidelity co-registration reveals a rich set of information. In this study, the harbor of Oslo is depicted and three container terminals with different orientation of containers towards the radar have been of special interest. In this paper, results are visualized by use of color composite images where the RGB channels contain day 1 and 2 intensity images and sample coherence, respectively. Extraction of details of container stacking, cluster configuration and rearrangements has been explored. The visualization shows clearly the appearance of various objects in the scene that has been removed, moved in or been replaced from day 1 to day 2 with distinct coloring differences. Examples also present other objects typically found in harbors and their appearance for a more complete understanding of the scene activity and as an important input for monitoring the area and scheduling new recording geometries.
  • Keywords
    containers; electromagnetic wave scattering; feature extraction; image colour analysis; image registration; image sampling; radar imaging; sea ports; stacking; synthetic aperture radar; RGB channels; SAR images; coherent change detection; color composite images; container backscattering; container stacking; container terminals; feature extraction; fidelity co-registration; harbors; image sampling; monitoring; recording geometries; scheduling; visualization; Coherence; Containers; Geometry; Image color analysis; Pixel; Radar;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Radar Conference (RADAR), 2011 IEEE
  • Conference_Location
    Kansas City, MO
  • ISSN
    1097-5659
  • Print_ISBN
    978-1-4244-8901-5
  • Type

    conf

  • DOI
    10.1109/RADAR.2011.5960512
  • Filename
    5960512