A wavelet-based particle swarm optimization algorithm for digital image watermarking

摘要

Digital watermarking techniques have been proposed for copyright protection and multimedia data authentication. The achieved tradeoffs from these techniques between imperceptibility and robustness are always controversial. This paper proposes the application of Discrete Wavelet Transform (DWT) into image watermarking by using Particle Swarm Optimization (PSO) which is an evolutionary technique with the stochastic, population-based algorithm for solving this problem. To protect copyright information of digital images, the original image is decomposed according to two-dimensional discrete wavelet transform. Subsequently the preprocessed watermark with an affined scrambling transform is optimal embedded into the vertical subband (HL_m) coefficients in wavelet domain without compromising the quality of the image. In the optimal process, the scaling factors are trained with the assistance of PSO. Furthermore, a novel algorithmic framework is proposed via a forecasted feasibility of the approach to parameters evaluation of hypothesized watermarked images in DWT domain. Simulation results show that the proposed watermarking procedure has better performances in imperceptibility and robustness under various distortions in the comparison of the previous scheme.