Digital Watermarking Improvement Algorithm Based On Discrete Wavelet Transform And Singular Value Decomposition

Digital watermarking technology is an effective means to solve copyright authentication disputes,digital products anti-counterfeiting and tampering,and to protect the security and integrity of digital products.This master dissertation focuses on two improving image digital watermarking algorithms and validation of their effectiveness through the experiment on the transform domain.Aiming at the problems of image distortion and poor robustness caused by the same embedded parameter among all sub-blocks in the existing block digital watermarking algorithm,a"dynamic digital watermarking algorithm on the block wavelet domain"isproposed.According to the human vision model on the wavelet domain and the image information entropy theory,the embedding amount and embedding intensity of sub-block are adjusted repeatedly to make the embedding amount and embedding intensity factor conform to different sub-block characteristics,and the watermark information is embedded dynamically.The experimental results show that the PSNR,i.e.peak signal-to-noise ratio of the embedded image is 43.25d B,and the correlation coefficient between the extracted watermark and the original watermark is close to 1 in none attacks on the above mentioned"dynamic digital watermarkingalgorithm on the block wavelet domain",the invisibility and robustness of which reach a good level under several different types of attacks.Aiming at the problem of image distortion caused by unreasonable modification of singular vector coefficient for the existing digital watermarking algorithm on the SVD domain,and the application of blind watermarking,a"blind watermarking algorithm based on DWT-SVD"is proposed.The stability of the first column vector of the image SVD singular matrix is verified by experiments in this dissertation.The carrier image is transformed by discrete wavelet transform,and the low frequency sub-bands are divided into 8×8 blocks on the above mentioned"blind watermarking algorithm based on DWT-SVD".On the basis of the singular value decomposition of sub-blocks,the coefficient pairs U₂₁with U₃₁ similarity is selected to design an embedding strategy,and the coefficient is modified in a small extent to realize the above mentioned image digital blind watermarking scheme.The experimental results show that the proposed"blind watermarking algorithm based on DWT-SVD"in this dissertation has little effect on the quality of carrier images,and the PSNR of carrier images embedded with watermarks reaches a high standard of about 50d B.Moreover,there is a good robustness against common attacks such as noise,JPEG compression and little area cutting on above mentioned algorithm.As the noise intensity reaches a high level,such as 2%Gaussian noise and 2%salt and pepper noise,the correlation coefficient NC value of watermark before and after extraction can still reach above 0.7 in terms of anti-noise attack.The proposed"dynamic digital watermarking algorithm based on block wavelet domain"and"blind watermarking algorithm based on DWT-SVD"in this dissertation can be used for design of digital image watermarking and its blind watermarking.Figure[29]Table[13]Reference[59]...