Research And Design Of Steganography Algorithm Based On Low-density Multi-carrier

Since the 1990 s,information hiding technology with "digital watermarking" and "steganography" as the main technical branches has received extensive attention in the field of information security.Among them,"steganography" may be used by criminals and intelligence personnel to covertly communicate and engage in activities that endanger public safety such as national politics,military,economy,etc.,so steganography and steganography analysis have been increasingly widespread in academia Attention.The current research on steganography is mostly based on a single carrier,but the single carrier is limited by the number of carriers,and the algorithm cannot further improve the embedding capacity and security.If we use multiple carriers to hide secret information can not only improve the embedding capacity,but also improve the performance of steganography because of the distribution of secret information.In addition,the rapid development of information technology has driven the development of image acquisition and other technologies.Different digital carriers explore on the Internet.Digital carriers available to steganists are massive.Multi-carrier steganography is more in line with the actual application of digital carrier big data scenarios.This paper focuses on the design of steganography algorithms based on low-density multi-carriers,focusing on multi-carrier processing and effective carrier labeling,carrier embedding area selection,secret information preprocessing,and the formulation of embedding rules.On the basis of completing the above research,two types of steganography based on low-density multi-carriers are proposed:(1)Multi-carrier steganography based on angular structural features.Firstly,extract the Angle Structure Descriptor(ASD)feature vector of the image to classify the carrier set,convert the image to the ASD direction field and encode it;second,segment the secret information according to the number of carrier classifications and scramble and optimize;Finally,based on the encoded data of the ASD direction field,the information is hidden on a large number of different types of carriers.In the extraction of secret information,the judgment of the color quantization difference when extracting the ASD feature vector is added to further improve the performance of the algorithm.(2)Three-dimensional model multi-carrier steganography based on distance characteristics and volume integral invariant.First,the distance feature,a global variable of the 3D model,is used to classify multiple 3D model carriers using the similarity formula;secondly,the volume integral invariant is used to calculate the volume inside and outside the sphere of the 3D model vertex,and the 3D model vertex is classified;The unevenness of the vertices of the model hides the secret information in the 3D model and sends it to the recipient for extraction.In this paper,the related simulation experiments on the multi-carrier steganography based on ASD are carried out.The comparison experiment results show that the advantage of the algorithm over the comparison algorithm is that the invisible PSNR value is enhanced by an average of 17.92% and 32.83%,respectively.The performance Q value is enhanced by 26.48% and 37.61% respectively.The compound attack experiment is performed on the algorithm.Under the general attack of 13% shear,13% compression and 20% noise,the PSNR value is as high as 39.0018.In the face of high-intensity compound attacks of 37% shear,40% compression and 22% noise,the PSNR value can still reach 35.0026.Another three-dimensional model steganography based on distance features and volume integral invariants improves the invisibility skeleton similarity by an average of 7.78% and 16.63% compared to the comparative algorithm.Facing non-uniform compression attacks,random noise,and re-networking with gridding,mesh simplification,and patch cut,the robust BCR values are improved by 4.97%,33.92%,7.27%,21.61%,4.35%,24.88%,16.19%,27.31%,respectively.19.72%,29.97%.The steganography proposed in this paper using multiple three-dimensional models or digital images as a carrier provides a reference for the hidden communication of large-volume secret information with high security requirements.