The proposed weighted structure tensor total variation (WSTV) model effectively captures local image features and maintains details during the denoising process, outperforming other TV-based methods.
