Schaefer, Christoph; Henrichs, Alois (EADS/Astrium, D-88039 Friedrichshafen, Germany)
Zahn, Rudolf (EADS, Germany)
The paper deals with on-board compression of airborne Synthetic Aperture Radar (SAR) intensity imagery. The new image compression standard R122 of the Consultative Committee for Space Data Systems CCSDS has been evaluated in this context. The corresponding compression performance is analyzed both visually and quantitatively. An argument is made that the conventional Root-Mean-Square-Error (RMSE) metric is not appropriate to measure compression distortion of the noisy SAR imagery. It is proposed to replace RMSE by distortion of Impulse Response Function (IRF) features. The status of implementation of R122 at EADS/Astrium is reported. It is concluded that SAR intensity images can be compressed by factors exceeding 10 in quasi-lossless modes.