Digital Watermarking Method Based On Wavelet Transform

Digital watermarking is a new technique for information security. It embeds a special invisible mark into digital multimedia data, then carries out the functions such as information hiding, storing, copyright protection by tracking the use of the image by detecting the watermark. At present, watermarking technique has been a kind of effective means to multimedia copyright authentication and its integrity protection.Watermark technology and its application of digital image is comparatively analyzed and studied in this thesis.It mainly includes the following several aspects:Firstly, it introduces the development present status and future of Digital Watermarking Technology, and summary and discussion of the general methods of watermark embedding and watermark detecting is carried out ,and several existing digital image watermark algorithms are generalized.Secondly, the author summarizes the main knowledge for Digital Watermarking Technology and the relative theories, such as Discrete Cosine Transform, Wavelet analysis and Chaos Theory, and the scramble method and the bit-planes method in the digital watermark's pretreatment are introduced. The author proposes the aim of this dissertation.Thirdly, The DCT domain algorithms are studied and improved,the author puts forward a Blind Watermarking Algorithm basing on DCT domain and Chaos.That algorithm, in which a binary image with practical meaning is selected as the watermark, is encrypted by the Chaos Sequence generated from Composite Nonlinare Discrete Chaotic Dynamical systems. Such, when examining the algorithm, no original image is needed. Lastly, makes an analysis of the results of the imitated experiment to prove that the plan is practical. And the experimental result shows the robustness of this algorithm.Fourth, the present popular wavelet domain algorithms are studied and improved, a self-adaptive algorithm of watermark based on wavelet transform is put forward, which decomposes a gray-level image watermark treated with Arnold scrambling into eight binary bit planes,and embeds eight binary bit planes into the wavelet-transformed original image.The method uses the characteristic of Human Visual System to adjust the intensity of the watermark adaptively.A sign matrix is set to record information about embedding watermark and be used for blind detecting the watermark..