Robust feature extraction based watermarking method using spread transform dither modulation

This paper proposes a local digital image watermarking method based on robust feature extraction. A robust feature extraction method is proposed based on DAISY to extract feature regions. Each feature region is decomposed into approximation and details sub-bands by wavelet transform and the watermark is then embedded into and/or extracted from the approximation coefficients. Spread Transform Dither Modulation is applied to embed and extract the watermark. Multiple copies of watermark images are embedded into the feature regions to increase the success rate of watermark extraction and meanwhile improve the robustness. Experimental results show the very good image quality of the watermarked image and the low Bit Error Rate of watermark extraction. Furthermore, the comparison results indicate the better performance and higher robustness of the proposed scheme when under various attacks comparing with the existing methods.