Research Of Digital Watermark Algorithm Based On DWT

The multimedia information safeties problems become an urgent subject along with the development of digital technology and the popularization of the internet. As a kind of main means to multimedia copyright protection and a new information hiding technology, digital watermark has been developed rapidly in recent years.The blind watermark algorithm has more superiority than the nonblind in some net condition where original images haven't been gained conveniently. At present the research of watermark mainly focus on invisible watermark. The research of visible watermark still is less. But visible watermark can indicate copyright more straightly, and can conceal part important information of original image. It cause that visible watermark have important effect in some special situation. Two watermark algorithm based on DWT and a new scrambling encrypting pretreatment method are proposed in the paper. The main contents are as follows:1. We discuss essential theory frames and representative algorithm of digital watermark systematically. Then we introduce theory base of digital watermark algorithm-discrete wavelet transform (DWT) : multi-resolution analysis, Mallat algorithm and application of DWT in image processing. Finally digital image scrambling encrypting algorithm are studied and a rapid convenient image scrambling encrypting algorithm based on chaos theory and random matrix scrambling is proposed. All those provide necessary guarantee to the efficient implement of the back watermark algorithm.2. Two watermark algorithms based on DWT (blind watermark algorithm and visible watermark algorithm) are presented. The front uses the same watermark repetitions embedding technology as the watermark embedding means. It not only realizes blind extraction, but also provides rectify information to enhance the extraction rate when we do watermark detection. Moreover we propose multiple extraction methods to cope with various attacks. The back confirm embedded information strength factor adaptively by HVS. The decomposed hiding information of different resolution is embedded into chosen bigger coefficients of corresponding resolution of fused image.A lot of experiments in the paper validate the two algorithms can achieve...