A new watermarking method based on DWT green component of color image

A scheme of embedding a scrambled watermark into the green component of the color image is proposed. The embedding process is implemented in the discrete wavelet transformation (DWT) domain. The binary watermark image is first encrypted by scrambling technique, then is spread by two orthogonal pseudo-random sequences with zero mean, and finally, is embedded into the DWT low frequency subband of the green component. The largest coefficients are selected to hide watermark, and the embedding strength is determined by the energy ratio between the selected coefficients' energy and the subband's mean energy. A watermark extraction algorithm and a blind detection algorithm are designed according to the embedding scheme. Experimental results demonstrate that the proposed watermarking scheme is robust to the attacks such as additive noise, low-pass filtering, scaling, cropping image, rows (or columns) deleting and JPEG compression.