Research On Reversible Information Hiding Method Based On Pixel Correlation

With the rapid development of information technology,a large amount of information is transmitted in the network world.In the process of information transmission,how to ensure the specificity of information and the concealment of communication has become an urgent problem in the field of information security.As a branch of information hiding,reversible information hiding has the characteristic that the original carrier can be completely restored after extracting secret information,and has attracted the attention of scholars in the field of security at home and abroad.Compared with many carriers such as video,audio,and text.Images have the advantages of simple acquisition and large redundant space,so images have become the main research carrier of reversible information hiding algorithms,and images are more widely used in military,medical,legal and other fields.This makes the research on image-based reversible information hiding of great significance.Based on the pixel correlation of images,this paper researches the reversible information hiding algorithm of plaintext images to improve the quality of encrypted images,thus ensuring the concealment of communication and the security of information.And study the reversible information hiding algorithm of encrypted image to improve the hiding capacity of the carrier image and the security of the encrypted image.The main work of this paper is as follows:(1)A histogram translation reversible information hiding algorithm based on gradient block is proposed.The reversible information hiding algorithm based on histogram translation only increases the embedding capacity by increasing the number of peak points,but the texture complexity of different regions of the image is different.The region with complex texture not only has a low embedding capacity but also destroys the original semantic information of the image.To solve the above problems,this algorithm divides the image into smooth texture area and complex texture area according to the pixel correlation of the image,and only performs information embedding on the smooth area,which reduces the loss of image semantic information.Then block the smooth area,increase the number of peak points,and improve the embedding capacity of the carrier image.Experimental results show that,compared with similar algorithms,this algorithm has higher embedding capacity,lower distortion of secret-carrying images,higher information security,and more concealed communication.(2)Propose a Reversible Information Hiding Algorithm Based on Global Prediction and DNA Encoding in Binary Tree Marked Ciphertext Domain.Most reversible information hiding algorithms in the ciphertext domain do not make full use of the pixel correlation of the original image or do not use the pixel correlation of the image at all,which eventually leads to the embedding capacity of the carrier image not reaching the ideal state.In view of the above problems,this algorithm preprocesses the image before encryption,and proposes a higher-precision prediction method to predict image pixels according to the pixel correlation of the image.The improvement of prediction accuracy means that the space that can be embedded is larger.According to the prediction error value,the embedding space is marked by the parameter binary tree marking method,and then the marked image is encrypted by a more complex chaotic encryption system.Then the information embedder embeds the information by bit replacement according to the hidden key and the embedding space of the mark.Finally,the receiver can either restore the original carrier image according to the encryption key or extract the information according to the hidden key.Experimental results show that the hidden capacity of this algorithm is higher,and the secret-carrying image has high sensitivity to the secret key,which ensures the security of the semantic information of the carrier image.