Technology Research, Based On Fractal Theory Of Information Hiding

People have researched the information hiding technology which is considered source-channeling safety technology for a long time; it is concerned with digital rights, information security and covert commutation. As one branch of information hiding, digital watermarking, which is used for digital rights management (DRM), is becoming fast developing technology in the multimedia area and has already aroused widely attention from international academe and research organization. Fractal theory, which is a fresh and novel nonlinear science, has been widely applied to many cases or fields of natural and social science and has a remarkable guiding significance for computer graphics, image processing, curve/surface fractal interpolation and so on. In this paper, watermarking algorithm based on fractal theory is researched and analysed in conventional transform domain, the main working is as follows:1.In this paper, two digital image watermarking algorithms are proposed separately and simulated, which are based on fractal code and DCT/DWT under the precondition of watermarking invisibility. Simulated results show that the DCT-based method bears a good function against block effect of DCT with local self-affine transform, and the coded image with fractal quad tree is corresponding to the features of Mufti-Resolution Analysis (MRA)of DWT and Human Vision System (HVS). This method has a good robustness performance. Otherwise, on the basis of Iterated Function System (IFS), a novel method is proposed to enhances the resist of rotation attack in some degree, which uses four affine transform of IFS to create an image as a watermark2.The features of digital audio and compress coding schemes are studied in detail and a watermarking algorithm is presented. According to the size of fractal dimension, this algorithm embeds different capacity watermark with fractal interpolation function (FIF) and MRA of DWT, and vertical proportion scale and DWT decomposing level L are adopted as keys that make it safe. In order to assure watermarking invisibility and robustness, this method embeds watermark in coefficients of mediate and low frequency. Future more, a scheme is proposed that embeds two watermarks in both detail and approximate coefficients based on wavelet packet decompose, aiming at the attack of high pass or low pass filter and compress coding. Simulated results show it has a good performance.3.The features of video and some video compress coding scenarios are studied. Then, a algorithm is presented based on compressing video and fractal trees with parameters.After computing the difference of fore-and-aft I frames'DCT coefficients and comparing with two thresholds, this method embeds different fractal trees in video I frames. This scheme uses IFSP to create cartoon-like pictures, which is consistent with the motion characteristics of video frame, and embeds these pictures into video frames. Presetting thresholds guarantee not to increase the bit rate while reducing the cumulative distortion. Simulated results show that watermark invisibility is good.