Research Of Digital Watermark Technology On Image Authentication

Along with the computer network and the communication quickly developing, the digital media (including digital image, digital video, digital audio) obtained the widespread application, following digital question and so on media information security, intellectual property rights protection and authentication problems also changes day by day prominently, and becomes in the digital world an extremely urgent important subject. The digital watermark (Digital Watermarking) as the effective supplement method of the tradition encryption, is a new technology that may protect the copyright and the authentication origin under the open network environment, in recent years it had drew more and more attention of people, and had became a research hot spot on the multimedia signal processing field .This article has conducted the comprehensive research to the digital watermark technology.First, introduced the digital watermark basic characteristic, the principle, classified as well as the present domestic and foreign research situation and so on the basic question, induces the digital watermark system frame, summarily introduced the present common digital watermark algorithm.Next, in detail introduced the wavelet encoding technology, discussed the zero tree encoding good and bad points, and realizes a specific encoding process, then absorbed embedded zero tree wavelet encoding thought, utilized it in the digital watermark technology.After that, has analyzed in the former digital watermark technology the threshold value selection limitation, combines human visual system (HVS), proposes one kind based on the image characteristic auto-adapted algorithm.Finally, proposed one kind based on the wavelet zero tree encoding auto-adapted digital watermark algorithm. Before based on in the wavelet transformation watermark algorithm, the low frequency coefficient always by obvious removed in outside, the algorithm which in the article proposed first searched the low and middle frequency important coefficient, then embedded the watermark into the low and middle frequency important coefficient.The experimental result proved that, the algorithm kept the watermark robustness at the same time guarantee image quality, not only adapted to human visual system, also had a good robustness to general image processing like cutting, filtering, smoothing, peaking and JPEG compression.