| 英文摘要 | Digital media, manifested in text, image, audio, video, and combinations thereof, is a kind of digital works or material. Owing to the rapid development of the Computer Science and Internet Technology, web publishing, online purchasing and online playing have become the major forms of selling and transmitting digital media. These new forms of sale and communication are easy to use and cost-effective, and will play an important role in the prosperity of the digital media industry. However, digital media resources can be easily intercepted in the transmission, and illegally copied and spread by the users, so the digital media copyright should be protected, and user’s ability of accessing the digital media should be controlled. Based on the above application requirements, this paper studies and explores the digital watermarking technology and the digital media encryption technology. The innovations of this paper are as follows: (1) Proposed an optimal detection weighted coefficient method for spread spectrum watermarking schemes. An optimal detection weighting coefficients method is proposed to improve the imperceptibility and robustness of a high-capacity Spread Spectrum watermarking algorithm, and the corresponding calculating functions are also provided. Experimental results show that this method can effectively improve the imperceptibility and robustness of the Spread Spectrum watermarking algorithm. In addition, the method can also be applied to other Spread Spectrum watermarking algorithms, and improve their overall performances. (2) Proposed two digital watermarking algorithms, which are resistant to affine transformations. For affine transformation attacks, this paper proposed two affine transformation robust algorithms. The first one uses the original host image, estimates the affine transformation parameters by normalized correlation coefficients, and then uses the parameters to inverse transform the image. Experimental results show that the affine transformation parameter estimation algorithm is highly robust, and can be used in non-blind digital watermarking schemes to recover affine transformations. Instead of using the original image, the second one uses a Trace transform algorithm to normalize the image and embeds the watermark in the normalized image. Because the normalized image is invariant to affine transformations, the embedded watermark is invariant to affine transformations too. Experimental results show that the Trace t... |
修改评论