Improving the Performance of Color Image Watermarking Using Contourlet Transform

In this paper, a color image watermarking scheme based on combination of discrete cosine transform, contourlet transform and singular value decomposition has been proposed. The color image is transformed into three color components luminance (Y) and chrominance (Cb and Cr). After that, discrete cosine transform is applied on the components to obtain DCT coefficients. The DCT coefficients thus obtained are decomposed into directional sub bands using 2- level contourlet transform. A grey scale watermark is decomposed using contourlet transform. Laplacian Pyramid decomposition is applied on both, fourth directional subband from watermark and cover image to obtain Low pass subband. The singular values of lowpass subband of watermark are embedded into singular values of lowpass subband of cover image. The experimental results demonstrate that the proposed watermarking scheme is robust against common image processing attacks. The comparison analysis reveals that the proposed watermarking scheme outperforms the color image watermarking scheme reported recently.