Research On The Digital Watermarking Algorithms For Color Images In Contourlet Domain Based On Schur Decomposition

With the rapid development of multimedia technology,video,image,audio and other digital media technologies are applied more and more widely,but they are vulnerable to damage in the process of transmission.As an effective means of copyright protection,digital watermarking technology has attracted extensive attention and in-depth study by scholars.Digital watermarking technology has achieved remarkable results since it was put forward.However,digital watermarking technology involves a wide range of content.In this paper,based on the existing research foundation of digital watermarking technology,by analyzing the principles of Contourlet transform,DCT transform and Schur decomposition technology,combined with the research of copyright watermarking and zero-watermarking technology,two robust digital watermarking algorithms are proposed.The main work of this paper is as follows:1?A zero-watermarking algorithm for color image in Contourlet domain based on Schur decomposition is proposed.The algorithm uses Co component of YCoCg color space to do Contourlet transform,and combines DCT transform and Schur decomposition technology to construct the feature matrix of the image.The zero-watermarking can be obtained by exclusive or operation between the feature matrix and the watermarked image,and saved to the third party copyright authentication center.The zero-watermarking is extracted from the copyright extraction center,and the feature matrix of the image is constructed to do XOR operation with the zero-watermarking to obtain the watermarking image.2?Based on NSCT transform,a dual watermarking algorithm for color image is proposed,which combines copyright watermarking and zero watermarking.In order to solve the problem of strong correlation among three components of RGB color space,the algorithm chooses YCoCg color space with strong decorrelation to embedding the copyright watermarking in the luminance component.In the embedding process,the luminance component is transformed into NSCT.Combining with Schur decomposition technology,the improved adaptive quantization algorithm is applied to embedding the copyright watermarking in the low frequency sub-band.Using the CO component of the watermarked carrier image to do NSCT transformation,combining with Schur decomposition technology,the image feature matrix is constructed and the zero-watermarking is obtained by XOR operation with the watermarking.The zero-watermarking is saved to the third party authentication center.The extraction of copyright watermarking is based on the inverse process of the previous embedding process.3?The proposed two color image watermarking algorithms are tested for robustness and invisibility.Compared with similar algorithms,the two algorithms in this thesis can better resist the impact of noise,cropping,JPEG compression,scaling and rotation attacks to a certain extent.Especially when resisting the impact of JPEG compression,shearing and noise attacks,it has outstanding advantages.