Matrix Singular Value Decomposition Method In Steganography And Digital Watermarking Technique

With the development of computer network and multimedia technology, the application of digital media is becoming wider and wider, various requirements such as copyright protection, completeness authentication and detection of digital media and hidden transmission of digital works are urgently needed. Information hiding technique which is a newly-developed technique in information security area is gradually becoming hot research area, and it has a huge research and application perspective.Matrix singular value decomposition (SVD) is an important technique in matrix decomposition area. It has been used in digital image processing. Because SVD's ability to resist a small perturbation is good and it represents image intrinsic properties instead of visual characteristics, it has been widely used in steganography and digital watermarking techniques in recent years. A thorough reseach is given in this paper.With the analysis of present correlative algorithms in steganography, the fault of statistical character in these algorithms is pointed out, and the improved algorithm is given. With the analysis of a present algorithm in digital watermarking, its fundamental fault in using SVD is indicated, and a new algorithm is proposed. With the combination of Watson Visual Model and SVD, a new algorithm used in completeness authentication and detection of digital works is put forward in this paper. This is used to explore the possibilities of the combination of SVD and other methods. Experiments show these proposed improved and new algorithms all have good performance.According to this paper's research, SVD is a valid method in information hiding techniques such as steganography and digital watermarking. It has a huge application value and wide prospect.