Research Into Robust Digital Image Watermarking Techniques

With the rapid development of information technology and computer networks, copyright protection of digital works has become more important. Digital watermark is an effective solution to the copyright protection and content authentication. As the most popular form of multimedia, digital images have become key research object. Robust image watermark technology offer an effective way. Main research efforts of this thesis are as follows.1. Overseas and domestic researches and advantages on image based digital watermark are sum upped firstly, and characteristics and basic model of digital watermarking are introduced. Then, a robust image watermark algorithm based on discrete wavelet transform is proposed. Considering the different texture features produced by wavelet transforms, a watermark is embedded adaptively. Experimental results demonstrate that the algorithm is robust to attack on data image.2. A robust image watermark algorithm based on Genetic Algorithm is proposed. Watermark embedding is modeled as a constrained optimization problem, and Genetic Algorithm is introduced to seek best solutions. By choosing proper fitness functions, the contradiction between robustness and fidelity is conciliated. StirMark experiment results illustrate the good effect.3. Based on the properties of chaos such as the random-like behavior and the sensitive dependence on initial conditions and parameters, a new robust image watermark algorithm using chaos is presented. Chaos sequences produced by a logistic map are used to scramble a watermark. The experimental results show that the imperceptibility and robustness of these methods in practice.At last, this thesis is sum upped with future researches pointed out.