Robust Reversible Watermarking Based On Biological Sensitivity

Due to rapid development of Internet as well as low cost and high convenience of digital information, more and more publishers and individuals publish their works in digital form and spread them on the network. Although digital works can be easily delivered and stored, their copyright protection and information security is greatly challenged. As an effective solution to copyright protection, digital watermarking has been extensively studied in recent years. In particular, by combining the merits of robust watermarking and reversible watermarking, robust reversible watermarking can not only recover host images and watermarks completely under the lossless environment, but also resist against some kinds of attacks under the loss environment. Base on this, the research on robust reversible watermarking is of great importance.In this paper, we make an extensive study of the challenging problems of robust reversible watermarking and propose some effective solutions. The main contents of this paper are as follows.(1) By taking the identification of biological information into consideration, a watermark generation method based on iris features is proposed to improve the security of watermarks. Firstly, this method selects appropriate local regions of iris automatically. By extracting and encoding the features of local iris regions, a watermark sequence is generated, which can solve the problem that traditional watermark sequences are not related to copyright owners.(2) For robust watermarking methods based on local feature regions, the selection of local feature regions is a key problem, which greatly affects the performance of the watermarking methods. Therefore, this paper proposes to select local feature regions based on the maximum weight clique and solve the corresponding problem with the novel optimization algorithm.(3) To optimize the robustness and invisibility, a novel watermarking embedding and extraction framework is designed. Based on this, two robust reversible watermarking methods are proposed, which utilize the stable components of the pixels and discrete cosine transform to improve the robustness and invisibility, respectively.In summary, this paper combines iris features with the watermarking generation and proposes two robust reversible watermarking methods based on optimized local feature regions, which not only enhances the security of watermarks but also improves the robustness and invisibility of robust reversible watermarking.