The Study Of Digital Image Watermarking Technology

With the development of computer networks and communication technology, it makes people's life convenient, but it has also created information security and copyright protection problems. Information hiding technology is called as the last line of defense of copyright protection, a complementary technology of encryption.In this paper, we have studied digital watermarking technology of information hiding, In order to explore how digital watermark applied research into new ideas. This paper focused on the close relationship between intelligence calculation and digital watermarking, and tried to use the methods and the theory of computational intelligence for digital watermarking research. The works include the following:In existing watermarking algorithm, the related parameters is to be defined relying on the experience or several experiments. This paper presents a genetic algorithm-based algorithm to search the optimal parameters of watermark image, minimizing the damage of digital watermark which is caused by embedding in the original image and it will achieve the mutual unification between fidelity and robustness.Wavelets have good local characteristics of time-frequency, space-frequency, we can fully utilize the human visual characteristics to embed watermark. An improved blind watermarking scheme in multi-wavelet domain was proposed, making full use of the image characteristics after CARDBAL multi-wavelet transform to embed watermark, establishing the relationship model between the watermarked image and the watermark, BP neural network model has a strong nonlinear approximation character to realize the blind watermark detection, The method can effectively improve the correct rate of watermark detection.