A statistical approach for determining radiometric precisions and accuracies in the calibration of synthetic aperture radar imagery

A model that estimates a relative error bound for the radiometric calibration of synthetic aperture radar (SAR) imagery is presented. This model is based on a statistical 'Coefficient of Variation of Error Model', which produces a relative error bound by propagating the measured or estimated uncertainties in the radar system parameters utilized to correct digitally processed SAR image intensity values. Using this model, algorithms are generated for absolute and relative radiometric calibration of SAR imagery. These algorithms are parametrically exercised using radar system parameters from an existing airborne SAR system to determine their impact on the relative error bound.<
>