Zero-watermark Algorithm Based On Joint Feature Of Digital Image And Laplacian Of Gaussian Edge Detection

Digtal watermark technology is important to protect the copyright of digital information, it’s a new direction of researches of information security, Research on digital watermarking technology will be better to improve the efficiency of information copyright protection of digital media.Zero-watermark is a new watermark technology to protect the copyright of digital information which need not modify the original picture,how to improve the robustness of the zero-watermark, it is one of the important contents of research,to improve the robustness of the zero watermarking algorithm can start from the extraction method of characteristic information of image, however, there are a lot of image feature extraction method, but for the zero watermarking technology is not enough,so the propose of this paper is to find the right method for extraction.In order to meet the needs of copyright protection of digital media information and improve the robustness when digital media information is being attacked by digital signal attack, proposed paper had researched following the two aspects:Firstly,Proposed a new zero-watermark algorithm based on joint feature of digital image on the basis of Chen and Chang’s zero-watermark algorithms and combining the human visual system and the digital image joint feature.In the first place,proposed algorithm gain the scaleimage by using mean value method,in the second,obtain the feature image through feature of human visual system included brightness value、texture value、contrast value and entropy value,then in sum of the value,to obtain the feature matrix of the scaleimage,through exclusive or and scrambling operation to protect the copyright of the digital picture.Secondly,on the basis of Chang’s copyright protection scheme,replace Sobel edge detection by Laplacian Of Gaussian edge detection.Because of the Laplacian Of Gaussian edge detection gain less points than Sobel,so Laplacian is better than Sobel when the digital picture being attacked.Last, By comparing with Chen and Chang’s algorithms’ in the signal processing attack,the proposed algorithm has the better performance than the other two algorithms.Last prove conclusion through simulation experiment.