Chaos And Research Watermarking Algorithm Based On Local Fractal

With the development and popularization of multimedia technology, the Internet isplaying an irreplaceable role in people’s life. People can utilize the Internet to publish,spread and acquire information. People are enjoying the convenience the Internet hasbrought, while they are suffering from copyright infringement. It is an urgent issue thathow to maintain a safe internet environment and how to protect copyright in an effectiveway. Therefore, the focused digital watermarking technology has been one importantmethod to protect copyright.The chaotic system possesses the characteristics of sensitivity to initial value,equation parameters uncertainty and nonperiodicity. In this thesis, the author applies thechaotic theory in the digital watermarking for watermark encryption so that the securityof the digital watermarking can be improved. For images, fractal means that there is acertain degree of self-similarity between the part and the whole of images. It isconfirmed combining with fractal theory that increasing strength of embeddingwatermark will enhance anti-attack capability of watermark. The specific studies are asfollows:Firstly, this thesis presents the research background and significance of digitalwatermarking technology and its development status at home and abroad, introducesbasic framework and performance evaluation of digital watermarking and describeschaotic system and fractal theory in details. It provides basic theoretical foundation forthe follow-up research of watermarking algorithm.Secondly, this thesis proposes a kind of image watermarking algorithm based onlocal fractal and Lorenz chaotic encryption. This watermarking algorithm encrypts thewatermarking image by converting the Lorenz chaotic sequence into chaotic matrix,conducts contourlet transform on the host image to extract low-frequency coefficientsand make local fractal analysis on the low-frequency coefficients to acquire fractaldimension so as to confirm the strength of embedding watermark. The simulations ofthe experiment reveal that the invisibility and robustness of watermark with this algorithm are both good.Thirdly, this thesis puts forward a blind watermarking algorithm based on localfractal and Lorenz chaotic encryption as required by copyright authentication. Thisalgorithm can realize blind extraction without original image during the watermarkingextraction.Finally, this thesis makes respective contrast experiments between the abovetwo algorithms and literature algorithm and tests robustness of the three algorithms. Theresults show that the proposed algorithms in this thesis can better resist some commongeometric attack upon satisfying the invisibility condition.