Digital Watermarking Algorithm Based On Image Edge Detection And DCT Domain

With the rapid development and wider application of computer, multimedia technology, the exchange of digital information becomes more fast and convenient. And in the case of fast exchange of information, information acquisition is also becoming more easily, thus it caused a lot of problems. These digital products such as images, video or audio copyright are destroyed, unlimited being copied to resell and malicious tampering. These phenomena to the development of information industry have been severely hampered. In this case the digital watermarking technology gradually is presented. Digital watermarking technology as an important branch of information hiding technology is an effective means to realize the digital image product copyright protection. In recent years has become a research hotspot in the field of multimedia information security.This paper focuses on the digital image watermarking technology based on DCT domain, and the main contents include:First, the research background and significance of digital watermarking technology are introduced, and the research status of digital watermarking, the basic concept, basic framework, characteristics, classification and its some typical algorithms are summarized and explained.Second, a watermarking algorithm based on DCT domain edge detection is given. Algorithm uses a key sequence to scramble the watermark image for encrypting. Get the edge of the original image according to edge detection operator. Then using two different methods, the encrypted watermarking information is embedded into the dc component of the original image after block with different embedding strength. Experimental results show that the algorithm for scaling, compression, cropping and other attacks have good robustness and invisibility. Third, an improved blind watermarking algorithm based on edge detection and the positive and negative of quantitative is given. Algorithm first uses Arnold scrambling watermark image to encrypt, then carries on the positive and negative change. For vector image first uses edge detection operator for edge extraction. According to the setting threshold to select different quantitative values, then the watermark information is embedded. And extract the watermark does not require the participation of the original carrier image, thus to realize a blind extraction. Experimental results show that the algorithm for cutting, compression, noise, filtering and scaling attack has better robustness and invisibility.Finally, Digital watermarking technology involves several disciplines and technologies, although we have studied for many years, in this area there are still many problems to be solved and some places have not been touched. Papers on the work of this paper are summarized and the development of digital watermarking technology is prospected.