An Improved Blind Watermarking Scheme in Wavelet Domain

The trade-off between the imperceptibility and robustness is one of the most challenges in digital watermarking system. To solve the problem, this paper proposes an improved blind digital watermarking algorithm for images based on the combination of Discrete Wavelet Transform (DWT) decomposition and Particle Swarm Optimization (PSO). For protecting the copyright information of a digital image, the watermark preprocessed with an affine scrambling is embedded in the vertical subband ( ) coefficients in wavelet domain. Furthermore, the proposed algorithm based on PSO is performed to train scaling factors to accomplish maximum the watermark strength while decrease the visual distortion. Simulation results demonstrate the proposed watermarking procedure has the remarkable performance in robustness to various attacks and the superiority of the proposed scheme.