Research On Digital Watermarldng Algorithm Based On Whale Optimization And Singular Value Decomposition

With the development of the Internet,it is more and more convenient for people to obtain information resources,which also brings about the copyright issue of digital multimedia works.Digital watermarking technology is produced under this background,and it has become one of the rapidly developing fields.Digital works usually have redundant information.According to this feature,digital watermarking technology is not easy to detect,but can determine the different watermark information.According to the extracted watermark information,the copyright of the digital work is determined,effectively protecting the rights of the original author.Robustness and invisibility in digital watermarking restrict each other.When invisibility is guaranteed,it is particularly critical to improve its robustness as much as possible.This dissertation optimizes the embedding strength based on the whale optimization algorithm,and balances invisibility and robustness.The main research contents are as follows:(1)A novel digital watermarking scheme is proposed for gray-scale images,which is based on discrete wavelet transform,singular value decomposition and whale optimization algorithm.Firstly,the carrier image is subjected to three-level wavelet transform,and the obtained three-level low-frequency subbands are extracted,and then the low-frequency subbands and the watermark are subjected to singular value decomposition.Finally,the intensity factor is introduced,and the watermark is embedded using the additive rule.The strength factor controls the trade-off between invisibility and robustness,and its value is determined by the whale optimization algorithm.The normalized correlation coefficient and peak signal-to-noise ratio are used in the optimized objective function design to satisfy invisibility and robustness.The simulation results show that the proposed digital watermarking scheme not only meets the demand for invisibility,but also has strong robustness.(2)For color images,a digital watermarking algorithm based on the idea of block division,non-sampled contourlet transform,singular value decomposition and whale optimization is proposed.This method performs block processing on the H,S,and I channels of the carrier image,respectively performs non-sampling contourlet transformation on these non-overlapping blocks to obtain low-frequency subbands,and performs singular value decomposition;the watermark image is divided into H,S and I channels,which are respectively scrambled by Arnold and then decomposed by singular value.Finally,the intensity factor is introduced and the additive rule is used to embed the singular values of the watermark information H,S,I channels into the low-frequency subband singular values of the corresponding channels of the carrier image.Simulation results show that the proposed watermarking algorithm satisfies the need for invisibility and can effectively resist attacks such as clipping,compression,noise,filtering,and sharpening.This dissertation examines the performance of the whale optimization algorithm through gray-scale image and color image analysis.The experimental results show that the watermarking scheme in this dissertation uses the whale optimization algorithm to optimize the embedding strength to ensure invisibility and at the same time have strong robustness.