Research On Zero Watermarking Algorithm For Geometric Attacks

With the rapid development of digital information, a large number of digital images are transmitted on the Internet, to facilitate people’s lives at the same time, it brings about a series of problems, such as copyright security, and so on. Digital watermarking technology is an important way to solve the problem of digital image copyright. Therefore, digital watermarking technology has become a hot research topic.This paper describes the security requirements for digital image copyright, and based on this gives following two zero watermarking algorithm against geometric attacks.Zero watermarking algorithm based on improved Sobel8 direction, this paper analysis the Sobel edge detection principle and the former used Sobel edge detection methods, and uses the Sobel8 direction of edge extraction, Sobel4 direction edge detection, log polar coordinates and other core technology, and presents the zero watermarking algorithm based on improved Sobel8 direction. According to the given algorithm conducts program design and simulation, and we can see the simulation results improve the robustness of noise,JPEG compression and other common signal processing attacks, and make better the robustness of the translation, rotation and other geometric attacks. This paper also analysis some defects of the algorithm, and base on it gives out the future of the research direction.Zero watermarking algorithm based on the SUSAN edge detection and Hu correction invariant moments, this paper studies the principle of SUSAN edge detection and Hu invariant moments, and uses the SUSAN edge detection, Hu correction invariant moments,log polar coordinates and other core technology, and gives the zero watermarking algorithm based on the Hu correction invariant moments and SUSAN edge detection.Completed the algorithm construct and program design of the zero watermark registration and zero watermark detection, to realize the function of digital image copyright protection,improve the robustness of the circulation translation, small angle rotation, scaling and other geometric attacks, and make better the robustness of the noise, mean filter, JPEG compression, and other common signal processing attacks, and through the inadequacies of the digital image anti aggression to exploration and research, has been clear about the next research direction.