Research On Digital Image Watermarking Algorithm Based On Image Scrambling And DWT_SVD

Visual sensation is the most important feeling of human,so digital media such as images and video are becoming one of the most common parts of human life.As the development of digital camera and other equipment functions,people are more likely to make images to record and disseminate important information.It also makes people pay more and more attention to the security issues of digital media information.In order to further solve the protection problem of copyright,the digital watermarking technique has turned out that embedding watermark data with specific meaning into the digital media in an invisible manner.In this paper,we firstly introduce the background and significance of digital watermarking technology,and the present research situation of digital watermarking.Then,the basic theory of digital watermarking is expounded.For the application of full-blind watermarking,a watermarking algorithm based on DWT_SVD can be used for full blind detection.Based on the characteristics of the watermark image and the carrier image,the theoretical basis of the wavelet transform and the image scrambling technique and the singular value decomposition theory,we further propose a method with Double Color Image Watermarking Algorithm Based on Arnold Transform and DWT_SVD.The innovation of DWT_SVD-based domain full-blind watermarking algorithm is that the copyright can be protected by only the attack image.The first process is DWT transform the vector image,and then divide the wavelet low-frequency sub-band into different sub-blocks and no overlap each other,and carry out SVD singular value for each sub-block.Then the key process is that based on the singular value relationship of nearest two sub-blocks produces a watermark sequence and embeds it into the original image,where uses the parity quantization rule in the embedding process.Finally,the SVD synthesizes the digital image which containing the watermark.In the subsequent anti-attack test through the comparison of original image and the final inclusion of watermark image,it shows that its invisibility and security is good,and achieve full blind safety certification testing.In the attack robustness test,the algorithm performs good performance on median filtering and Gaussian low-pass filtering.The innovation of double color image watermarking algorithm is that the watermark image and the carrier image are color.After studying the existing color digital image algorithm,we pretreat the RGB component of the color watermark image by Arnold transform.Based on the human vision system characteristics,the wavelet transform and singular value decomposition are used to find the embedding position and intensity in combination with the color component of color image.We propose an optimized double color image digital watermarking algorithm.We carry out the simulation experiments and robustness analysis of embedding and extracting watermark.The algorithm has good invisibility and security,and is good for watermark and vector image restoration.It not only works well in lena image and emblem image test,still valid in other application scenarios.Experimental results and attack test show that the algorithm is effective in anti-robust attack test and is an effective double color digital image watermarking algorithm.