Digital Watermarking Algorithm Based On Fractional Fourier Transform And Capacity Analysis

Recently, with the rapid development of the multimedia technology and internet technique, it is very convenient for us to transmit, copy, store and process multimedia information, such as images, audio video and texts. At the same time, the problem of the digital media's information security and intellectual property protection is increasingly outstanding, and the problem has became an important research topic in the digital domain.For the strong points of invisibility and robustness of digital watermark, it is one of techniques to solve the above issue. With the aid of National Nature Science Foundation of China(NSFC)(No. 60472044), this dissertation focuses on research on digital watermark algorithm and capacity based on fractional Fourier transform. The main works includes:(1) After many documentation realated with digital watermark are read carefully, the development of the digital watermark is generalized systematically. Then the target and direction of the reaearch are grasped. Finally, we analyse the generic system model of digital watermark.(2) A digital waterark algorithm based on fractional Fourier transform is designed. In order to make the research convenient, we use pseudo-random sequencewhich submits Gauss distribution as watermark siginal, the watermark is embeded to the fractional Fourier domain of the digital image, and the embedding rule is additive. We computes the inserted watermark's maximum strength coefficient by using the Perk Signal-to-Noise Ratio (PSNR) which is considered as the image quality evaluation criterion. We use correlation detection as the watermark detection method,and further analyse the relationship between the embedding quantity and robustness. The result of the simulation indicates that the algorithm introduced in this paper can satisfy the invisibility, security and robustness of the watermark well.(3) A digital watermark algorithm using the combinational spatial and fractional Fourier domains is presented in this paper. The watermark image is splitted into two parts which are for spatial and fractional Fourier insertion respectively. This algorithm increases the watermark capacity, and the security of the watermark is better, but the robustness is not very good, so some improvements should be made in the future.(4) The watermark capacity under some watermark channel model is analysed, a method to compute the watermark capacity in the fractional Fourier domain is presented by using the noise visibility function and the theory of information.