This study proposes a method that improves the robustness of independent component analysis (ICA) by adding outlier rejection rule for solving synthetic aperture radar (SAR) image analysis problems. Since the noise in SAR images is multiplicative, the applicability of ICA is seriously reduced. The p