An improved image watermarking scheme using selective curvelet scales

The paper proposes an image watermarking algorithm on curvelet transform domain. Different from other conventional methods which embed the watermark in spatial domain, discrete cosine transform (DCT), or discrete wavelet transform (DWT), the present paper uses curvelet transform domain for image watermarking since curvelet transform is effective when representing curves with fewer coefficients than other transforms such as wavelets for the same accuracy. The paper will investigate the different scales of curvelet transform domain for embedding the watermark to achieve both invisibility and robustness. The performance of image watermarking in curvelet transform domain is compared with the same algorithm but applied on DCT-DWT combined domain. Peak signal-to-noise ratio (PSNR) and normalized correlation (NC) are used to evaluate the invisibility and the robustness of the algorithms, respectively. Experimental results show that invisibility and robustness are ensured. Performance in terms of invisibility and robustness of the proposed algorithm is better than that applying on DCT-DWT combined domain for the lossy JPEG compression attacks, speckle and Gaussian noise.