Research On Compressive Sampling Theory And Its Application In Digital Watermarking Technology

With computer,Internet technology and information technology developing rapidly,almost all kinds of multimedia products exist and share in the form of digitalization.However,how to avoid multimedia products being malicious stolen and falsified in complex network environment,how to protect the rights and interests of users and producers are important issue that needs to be solved urgently in the digital industry.This technique of digital watermarking has an irreplaceable application value in copyright protection.The compressive sampling theory is different from the Shannon Nyquist theorem,which can be sampled and compressed at the same time no longer limited by “two times”,Achieved the goal of reducing the cost of information acquisition and storage.Some of its characteristics are urgently needed in many aspects of digital watermarking system.For example,there exists a corresponding sparse technology in each transform domain of watermarking,which adds an operational environment for its embedding,and the observation matrix is the reason why digital watermarking can improve the product security index.As a key,it protects information.In this paper,the above two theories are studied and discussed in detail.By synthesizing the advantages of the two theories,a more comprehensive and effective new watermarking algorithm is designed.The main contents of this paper are summarized as follows:Firstly,combing the three core links necessary for the compressive sampling framework: signal sparseness,observation matrix and reconstruction algorithm,and based on this,the five common greedy reconstruction algorithms are simulated and implemented to complete comparative analysis.Secondly,according to the characteristics of waveforms in contour domain,which have many propagation directions and resolution,multi-scale geometric analysis is used to solve the limitations of horizontal,diagonal and vertical decomposition directions in the process of decomposition of discrete wavelets.The core idea of block theory is analyzed and analyzed.In view of the shortcomings of block technology used in the past,the block effect is reduced by means of sub-sampling.Finally,the Contourlet domain compressive sampling algorithm based on the idea of sub-sampling and block is designed.Choosing MATLAB software to simulate and verify the algorithm,and based on this,the algorithm is judged to be feasible.Thirdly,this paper combines the two techniques of compression sampling and digital watermarking.Watermarking is embedded in the DWT compressed sampling transform domain of image.The experimental results show that it is almost impossible for this kind of algorithm to extract or destroy the watermarking information when the observation matrix and other information are unknown,which proves that the algorithm is robust to most attacks.After complete data analysis and research,this paper considers that the Contourlet domain algorithm based on sub-sampling and sub-block is a practical and effective improvement direction of digital watermarking technology.The simulation results show that the digital watermarking of this scheme has the characteristics of information not easy to be found,strong robustness and good protection effect.At the same time,because the technology compresses the products,it shortens the running time and improves the running efficiency.