A unified comparative framework for multiscale geometric transforms in SAR and multispectral satellite image analysis
Rajakumar Krishnan
Satellite image analysis is essential for remote sensing analysis. Two types of data are captured via satellite: Synthetic Aperture Radar (SAR) imagery (which has structure) and multispectral imagery (which contains spectral information), so complementary data may present unique challenges because noise, image resolution, and image angles differ across modalities. This paper describes the creation of a unified cross-modality framework to evaluate alternative transforms (Fourier, Wavelet, Curvele