In recent years, synthetic aperture radar (SAR) technology is investigated for automotive applications. SAR images via matched filter suffer from noisy background, sidelobe artifacts and low resolution. Sparse imaging methods have the potential to restore the images. However, the conventional unweighted ℓ₁ regularization method restores the SAR images with systematically underestimated amplitude and inconsistent supper-resolution factor. In this paper, we propose a reweighted ℓ₁ regularization method for SAR image restoration. Our method is supposed to simultaneously enhance weak scatterers and suppress sidelobes even in the case of SAR image with wide dynamic range of RCS. Real data experiments are carried out to verify the feasibility and effectiveness of the proposed method.