Pre-processing of Sentinel-1 C-band SAR images based on incidence angle correction for dark target detection

摘要

Wide swath synthetic aperture radar (SAR) has been commonly used for oceanographic applications, such as maritime safety, fisheries management and coastal monitoring, including dark target detection. However, the incidence angle (θ) effect on these data makes a direct visualization and classification of oceanographic targets troublesome. The recent availability of free-of-charge Sentinel-1 C-band data have the potential to assess methods for correcting the θ in wide swath SAR images, an issue that still remains to be undertaken. The goal of this work was to test two methods for correcting the θ effect of two Sentinel-1 SAR imaging modes and assess their implications for dark target detection in a sector of the South Atlantic Ocean. We assessed the empirical modelling and square cosine-based methods for correcting the 8 effect on co and cross-polarization from the Interferometric Wide Swath (IW) and Extra Wide Swath (EW) imaging modes. The results showed that the empirical modelling method improved the visualization and classification of dark targets for the co-polarized bands from both the imaging modes. However, further studies are necessary to test the potential of empirical modelling for correcting the θ for other different oceanic environments and SAR frequencies.