Research On A Wavelet-Based W-SVD Watermarking Algorithm

With the rapid growth of the Internet, providing great convenience for multimedia data access, the security of digital information has become increasingly concerned about. Especially, the copyright protection of digital product is of vital importance. The watermarking technology is one of the most effective one among all the solutions to the copyright protection. Therefore, the research on watermarking becomes a focus of academic study.Digital watermarking which embeds copyright information into the data, has been proposed to solve the problem. Digital watermark has incomparable advantages in digital products copyright protection, data authentication, etc. and is becoming new developing information security teleology. Compared with the spatial domain digital watermarking technology, the security and behavior of preventing attack of the transformed domain ones are better, which became the focus of digital watermarking technology. There are two important issues that the watermarking algorithms need to address. First, watermarking schemes are required to keep the quality of the original data and imperceptibility. Second, good watermarking schemes should satisfy the requirement of the robustness and resist distortions due to common image manipulations, such as filtering, cutting, etc.Based on abundant literature review, we first analyze the background, classification, modeling, application and recent development of the digital watermarking technique. After then the design requirements of watermarking systems are analyzed, evaluation methods for watermarking system are also provided. Adopting still images as research objects, the thesis focuses on wavelet transform domain singular value decomposition (SVD)-based image watermarking techniques. Wavelet transform theory is a new branch of mathematics emerged in recent years. Due to its excellent local properties; wavelet transform based digital watermarking methods have become hotspot of research. The basic wavelet theory and the application of singular value decomposition to image matrix processing are also included in the thesis.Experimental results indicate that SVD-based watermark algorithm may enhance the anti-geometry distortion of the embedded image, but it suffers from weak resistance against some copy attacks, such as Salt, Gauss, and Filtering, etc. This thesis adopts discrete wavelet transform (DWT) as the transform method, thus the resulting watermarking algorithm has the advantages of imperceptibility and stability. A novel watermark algorithm based on wavelet and singular value decomposition (W-SVD) is proposed, where SVD is carried out in the DWT domain. First, the singular values of the watermark are embedded to the corresponding low frequency bands of the singular values of the original image. The proposed algorithm improves the watermark transparency, robustness is also improved as watermark embedding is carried out based on frequency classification.According to the algorithm mentioned above, we use the Matlab software on the computer to prove it for the static image. In order to test the imperceptibility and robustness of the W-SVD algorithm, a lot of experiments are done. Experimental results demonstrate that the W-SVD algorithm is effective and useful. Because of the simpleness and convenience, they may be expected to have a better application prospect.Finally, the thesis concludes with summary and outlook of further work.