Research On Texture Complexity Watermarking Algorithm For Tampering Location And Recovery

Times are changing,technology is developing,and the network era with a hundred flowers is undoubtedly convenient for people's life and work,but there are a series of information security problems hidden in the convenient life,such as digital information forgery,copyright infringement and so on.In order to solve these problems,some scholars put forward cryptography in the early stage,which guarantees the security of information by establishing public key cryptosystem,and uses digital signature to protect the integrity of messages and identify the source of messages.However,this method has drawbacks.If the information is received successfully,the protection effect will be lost.Digital watermarking solves the shortcomings of traditional cryptography.Its basic idea is to use certain algorithms to embed the marked information into video,audio and image,but it does not affect the use value of the original content.People's perception system can not be found,only rely on certain equipment to extract,so as to achieve copyright protection of digital products.But the traditional digital image watermarking can only play the role of surveillance,covert communication,content authentication,etc.It can not locate the modified area and restore the tampered area.In order to solve this problem,scholars have developed a digital watermarking technology for content location and recovery of tampered images.This method can accurately locate the tampered areas and restore tampered images with high quality.It plays an important role in military,medical,legal and other fields.By classifying and summarizing the existing algorithms,focusing on the study of image texture features and edge information,an image tampering location and watermarking restoration algorithm based on texture complexity is proposed.The main work of this paper is as follows:(1)Aiming at the problem of inadequate utilization of texture correlation in existing methods,which leads to the inadequate selection of image block eigenvalues and insufficient concealment,a novel tamper location and recovery watermarking algorithm based on texture complexity and Laplace operator is proposed.Firstly,the algorithm divides the image into 3×3 blocks,calculates the texture complexity of each block separately,merges homogeneous blocks with texture complexity to reduce the influence of watermarking embedding on image quality;then calculates the convolution values of parent blocks and individual blocks as their eigenvalues,and takes the location information of parent blocks as the eigenvalues of subblocks.The new eigenvalues enhance the sensitivity to tampering and reduce the impact of watermarking embedding on image quality.Finally,the eigenvalues are embedded into the lowest significant bit.(2)Aiming at the problem that the classification of image sub-blocks is not meticulous enough to cause high quality recovery after image tampering,an image tampering location and recovery watermarking algorithm based on texture features and edge information is proposed.Firstly,the image is divided into 3×3 sub-blocks,and the texture complexity of the image is calculated by gray level co-occurrence matrix to get the most complex blocks.Then,the edge blocks,sub-complex blocks and simple blocks are divided by Canny edge detection and variance.Finally,the watermarking information is generated adaptively according to four block types.Experiments show that when the image is tampered with,the algorithm can accurately locate and restore the image with high quality,which is 4.12% higher than the peak signal-tonoise ratio(PSNR)of the existing algorithms.