The Research About Blind Watermarking Algorithm Based On Modulo

With the booming development of multimedia technology and the exceptional growing of internet ,we can get digital information conveniently and quickly .But how can we protect the author's copyright and economic profits ?It has become much urgent to provide efficient copyright and security protection means for the digital products. So the digital watermark technology has been put forward in recent years. Digital watermark is a new direction in the information security field .The author's personal copyright information can be embedded in the products. But we can't percept it, only special software can pick it up. In this dissertation, the application of digital watermark technology has been studied in the round. Firstly, this thesis analyzes the research background and the current development status of the digital watermark, describes the notion, principles of the digital watermark ,and discuss the basic frame and the general model of digital watermark processing system. Secondly, the paper analyzes four critical technology of digital watermark: the scrambling, embedding, detection and the attack analyses of digital watermark. In the mean time, the paper analyzes the technology of the digital watermark based on the wavelet domain., In chaper four, a blind watermark algorithm based on modulo in DCT domain is present. Firstly, the original image is manipulated by DCT and is split into 8*8 DCT Blocks . Secondly, The algorithm scrambles the watermark and produces the watermark embedding position in the DCT domain based on pseudorandom sequence. The algorithm inserts watermark into the important low frequency coefficients of DCT Blocks. The experiment results show the algorithm has imperceptibility and robustness. Finally, a blind watermark algorithm based on neighbor relationship in DCT domain is present.Firstly, the binary watermark image is encrypted by chaotic system and is transformed into one dimension vector; Secondly , the original image is split into 8*8 blocks and every block is manipulated by DCT; the blocks which are inserted into watermark information are controlled by pseudorandom sequence, so that we can know the first position of every block. As far as every block is concerned ,we choose one designed dot as the information hiding position to embed only one bit watermark information based on the relationship of the DCT coefficient and the neighbor mean of this dot. The experiment results show the algorithm has imperceptibility and robust to common image processing operations and some geometric attacks,such as JPEG lossy compression,additive noise ,cropping,and so on.