|Title||Detection and Mapping of Wetlands from Multi-Temporal Polarimetric Data|
|Author||Baghadi, N; Bernier, M; Gauthier, R P; Neeson, I|
|Source||International Symposium, Geomatics in the Era of RADARSAT (GER'97), Ottawa, Canada, May 25-30; 1997 p. 14|
|Alt Series||Earth Sciences Sector, Contribution Series 20042210|
|Abstract||This publication reports preliminary results of an experiment carried out to examine the potential of C-band Synthetic Aperture Radar (SAR) for mapping various wetland classes found in the Mer Bleue
region (near Ottawa, Canada). The Mer Bleue region was surveyed by the C-band (5.3 GHz) multipolarization (HH and HV or VV and VH) polarimetric SAR of the Canada Centre for Remote Sensing (CCRS) at various times in the hydrological cycle of wetlands.
Images collected on the following dates were selected; June 16, July 6 and October 3, 1995. Signatures of different cover types: peat bog (forested and nonforested), marsh, open water, clearing and forests have been derived as a function of incidence
angle. Separability between various classes were conducted to determine the relationships between season (s) and polarization (s) needed to differentiate various wetland classes. An algorithm has been developed for wetlands mapping by means of
multipolarization data. These investigations demonstrate some of the capabilities of SAR at C-band on mapping wetlands. The cross polarization data provided the best separation between the observed classes. The October dataset was more suited for
discriminating between the classes present than the others periods observed. The overall accuracies of the classification are of 72.8% from June, 73% from July, and 85% from October. Classification using a single-polarization has been investigated
and the results show that HH and the cross polarizations are better than VV polarization. For October, the percent of all pixels correctly classified is of 74.2% for HH polarization, 76% for cross polarization, and 59.1% for VV polarization.