Integration of optical and polarimetric SAR imagery for locally accurate crop classification

This research aims to find the best selection of imagery (optical and polarimetric radar) and methodology for very accurate and operational crop classification that could be used as a replacement for direct field observations and annual monitoring. We use RapidEye imagery as the optical source and RADARSAT-2 imagery as the Synthetic Aperture Radar (SAR) source. Both Dual polarization (Dual-pol) and Quad polarization (Quad-pol) imagery are used in this research. Optical, polarimetric SAR and the integration of both optical and polarimetric SAR images are tested for crop classification. Results show that the integration works better than single source of imagery, and Quad-pol outweigh Dual-pol in the integration with RE, especially for cereals, pasture and soybeans.