Abstract
A neural network-based method for speckle removal in synthetic aperture radar (SAR) images is introduced. The method rests on the idea that a neural network learning machine, trained on artificially generated input–target couples, can be used to efficiently process real SAR data. The explicit plus-point of the method is that it is trained with artificially generated data, reducing the demands put on real input data such as data quality, availability and cost price. The artificial data can be generated in such a way that they fit the particular characteristics of the images to be denoised, yielding case-specific, high-performing despeckling filters. A comparative study with three classical denoising techniques (Enhanced Frost (EF), Enhanced Lee (EL) and Gamma MAP (GM)) and a wavelet filter demonstrated a superior speckle removal performance of the proposed method in terms of quantitative performance measures. Moreover, qualitative evaluation of the despeckled results was in favour of the proposed method, confirming its speckle removal efficiency.
Acknowledgements
The authors are grateful to the Belgian Federal Science Policy Office for their funding of the SYNOPRA (Synergy of very high resolution optical and radar data in forest mapping and inventory) and HYDRASENS (Integrating radar remote sensing, hydrologic and hydraulic modelling for surface water management) projects respectively through the ORFEO (Optical and Radar Federated Earth Observation) and STEREO (Support to the Exploitation and Research of Earth Observation) data 2 Program. The Centre National d'Etudes Spatiales (CNES) (France) is thanked for providing the ADS40 and RAMSES imagery. The European Space Agency (ESA) is acknowledged for providing the Envisat ASAR data in the framework of AOE467.