Analysis of single, dual and polarimetric synthetic aperture radar data for iceberg/ship detection and discrimination – ASAR and CONVAIR-580 SAR datasets from 2003 to 2007


  1. Power D.
  2. Howell C.
  3. Bobby P.
  4. Youden J.
Corporate Authors
Defence Research and Development Canada, Ottawa Research Centre, Ottawa ON (CAN);C-Core - St. John's, NL (CAN)
The RADARSAT-2 satellite is the most advanced spaceborne C-band synthetic aperture radar (SAR) available due to the options of resolution, swath width, beam position and polarization. These advances have the potential to enhance both the probability of detection and discrimination of point targets such as iceberg and ships over other spaceborne SAR sensors. This report includes an evaluation of the potential gains when using multi-polarization data for both detection and discrimination of iceberg and ship targets. Data used in this study consist of well-validated airborne Convair-580 quad polarization SAR and spaceborne ENVISAT dual polarization SAR. In total, the data set used here consists of 2207 validated iceberg and ship targets. Following a methodology for improved ship detection presented by Liu et al. (2005), receiver operator characteristic curves were derived that allow for the evaluation of detection versus false alarm trade-offs. This method also allows for direct comparison of single and multi-polarization target regionalized detection. In addition, a constant false alarm rate target detection was employed to evaluate the point target detection capabilities for single and dual polarization SAR. Discrimination between ship and iceberg targets was accomplished using a methodology successfully demonstrated by C-CORE (Howell et al., 2004, Howell et al., 2006, Howell, 2008). Here a quadratic discriminant with target feature optimization is accomplished via the sequent

Il y a un résumé en français ici.

synthetic aperture radar;SAR;iceberg;ship;detection;discrimination
Report Number
DRDC-OTTAWA-CR-2012-024 — Contract Report
Date of publication
01 Jan 2012
Number of Pages
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: