Research Of Digital Watermarking Algorithm In Wavelet Domain Based On Fractal And HVS

With the advancement of network communication technology, the exchange and transmission of multimedia message become very convenience and fast. We can promulgate products, transfer important message, learning communion and E-business. These technologies bring people many conveniences, while the side effects that they cause are thoroughly obvious. For example, anybody can get the original products of others easily. Someone can copy and tamper the original works arbitrarily without author's agreement, which infringes gravely the author's copyright. How to protect the copyright effectively and solve the problem of information safeties on network is an urgent subject.Presently, as a kind of technology to solve multimedia copyright authentication and its integrity-safeguarding problem, watermarking technology catches many people's attention in the academe. It embeds watermarks into the digital media, then carries out such functions as copyright safeguarding by detecting the watermarks that can track the use of the digital media. This paper takes still images as example. It brings image processing technologies such as DWT, fractal technology and human visual system (HVS) into digital watermark technology, and presents a watermark scheme in wavelet transformation domain based on fractal coding and HVS.A digital image watermark scheme based on fractal coding and wavelet transformation is presented in this paper. At first, the original image is decomposed through DWT. The domain subtree and range subtree are redefined. The optimum fractal-matched domain block is searched in the same orientation but lower resolution than that of the range block. According to tree structure relation of the wavelet coefficients, the domain blocks can extend through different resolutions along the same orientation to form a set of domain subtrees. A number of matched domain subtrees are selected as embedded trees in every resolution band of orientation. Employing the similarity of the domain subtrees and range subtrees of redefinition, the watermark can be embedded into the nodes of the domain subtrees of host image and retrieved from retrieved image with embedding data easily. In the scheme, in order to embed into the image the maximum, but still imperceptible, level of watermark, human visual system is referenced. At first, it is used as a parameter of watermark weighing. And considering how the eye perceives disturbs, the consideration of the eye is less sensitive to noise in high resolution bands having orientation of 45°, has been taken into account. More watermarks are embedded in less sensitive orientation resolution bands.At the end of the paper, the results of experiments show that the watermark is invisible and has good robustness against the attacks of JPEG image compression, Gaussian white noise, median filtering, cropped and rotated et al. And after each experiment there is an analysis with the result.