Analysis And Application Of Digital Watermarking Technology Based On Transform Domain

The rapid development of digital technology and the popularity of the Internet not only have brought people much convenience in the daily life and work, but also have provided a more convenient way for some malicious behaviors, such as copyright infringement, information falsification and so on, which have seriously damaged the interests of the creators of digital works and the users. Digital watermarking technology is to hide special and invisible marks into a digital image, audio or video file by the digital embedding method, so as to identify the copyright owners, the authenticity of digital content sources and the purchasers, to provide the other additional information on the digital content for the purchasers, to confirm the ownership authentication and to track the pirate. Transform domain digital watermarking became the focus of attention for its significant advantages. Especially, the DCT domain digital watermarking and the DWT domain digital watermarking will be the main trend in the future development of digital watermarking. Based on the above reasons, this paper will focus on the watermark system design by taking a static gray image as original image, taking a binary image as watermark signal and embedding the watermarks in DCT and DWT domain.This study mainly focused on the following five aspects:1. DCT and DWT transforms were performed to the images, their coefficients were extracted and the characteristics of coefficient distribution were analyzed for the images after the two transforms;2. The comparative advantages and disadvantages of DCT and DWT transforms were compared and analyzed from the aspects of energy distribution, direction of selectivity, human visual characteristics (HVS) and so on;3. According to the characteristics of coefficient distribution of the images after the two transforms, watermark embedding strategies were analyzed in the different frequency domain bands, and in terms of the invisibility and robustness of watermarks, the relative optimum embedding positions were given;4. The watermark was embedded in the given relative optimum embedding positions by taking a binary image as watermark and combining image scrambling technology, and taking advantage of neighboring symbol's mean value and odd-even adjudgement rule algorithms referred to the image adaptive blind watermarking algorithm based on the wavelet transform proposed by Wang Pei, et, al. The original image was not required when the watermark was extracted and detected, and blind extraction and test was carried out in the algorithm. Antiattacking test was performed for the algorithm by certain methods such as the various noise, filtering, compression and gray transforming, and so on. The secret key was both used in the generation and extraction of watermark, and combining the image scrambling technology the pretreatment of watermark was conducted, which improved the security and safety of watermark;5. The digital watermarking was applied into digital museums. The necessity and inevitability of occurrence for Digital Museum was analyzed, as well as the feasibility of combination of digital watermarking and Digital Museum was discussed. The watermarks were found in good invisibility and having certain anti-interference ability when the digital watermarks were embedded into illustrations from the ancient books and the Calligraphy and Painting by using the new algorithm. Owing to the advantages of the new algorithm such as simplicity, easy to realize, extraction and test without the original image and the safety guarantee of secret keys, the possibility to apply successfully the watermarking into Digital Museum maybe come true.