Research On Reverisible Data Hiding Based On Encrypted Domain Of Image

With the advent of the era of big data and Internet,multimedia data plays an important role in people's life and greatly facilitates people's life.However,in the process of data transmission and processing,information security is particularly important.Especially with the popularization and application of cloud computing technology,the security protection of mass data stored in the cloud is more serious.How to ensure the security of information transmission and the integrity of information content has become an extensive research direction of scholars.The reversible data hiding technology in the image encryption domain combines the image encryption technology and the reversible data hiding technology to encrypt the image content during the image transmission process to ensure the concealment of the image content.Meanwhile,the secret information hiding is carried out in the encrypted image to complete the image authentication,management and distribution.Its main feature is to ensure the reversible extraction of image and secret information from the encrypted image.In this paper,the following works are carried out for the reversible data hiding algorithm in encrypted images:1.In order to improve the performance of the reversible data hiding algorithm in encrypted images which combines the “reverse room before encryption algorithm” with the “separable algorithm”,this paper studies the reversible data hiding algorithm in encrypted images based on error expansion and image partition.The algorithm first uses the designed image segmentation algorithm to divide the carrier image into three parts,which are respectively used for pixel value prediction,self-information embedding and secret information embedding.Then the algorithm of prediction error expansion and multi-histogram modification is used to create redundant space for embedding secret information.The stream cipher algorithm is used to encrypt the image and generate the encrypted image.Secret information is embedded with least-significant bit substitution.In the process of information extraction and image restoration,the reversibility of secret information and original image is guaranteed.In the image segmentation algorithm,the utilization of the correlation between adjacent pixels and the embedding sequence based on the regional smoothness make the performance of the algorithm greatly improved compared with that of the comparison algorithm.2.In order to improve the performance of the reversible data hiding algorithm in encrypted images which is combined the “reverse room after encryption algorithm” with the “separable algorithm”,this paper studies the reversible data hiding algorithm in encrypted images based on the cross template and homomorphic encryption.The algorithm first uses the cross template to divide the original image into several groups.The additive homomorphism system is used to encrypt the image and the difference histogram is generated inside the cross template to embed the secret information.At the image receiving end,due to the homomorphism of the encryption mode,the identical difference histogram can be generated in the plaintext field and the ciphertext field in the same way for information extraction and image recovery.Compared with the classical homomorphic encryption algorithm,it has lower computational complexity and can guarantee the security of the algorithm.Simulation results show that the performance of the algorithm is superior.3.In order to further improve the performance of the reversible information hiding algorithm in the encryption domain which is combined the “reverse room after encryption algorithm” with the“separable algorithm”,this paper studies the reversible data hiding algorithm in encrypted images based on the adaptive multi-mode region division.According to the embedding capacity requirement,the algorithm uses the adaptive region division strategy for the original image,which corresponds to the different proportion of the number of center pixels and reference pixels in the unit template.Different types of difference histograms are generated by using homomorphic encryption in the template.Through theoretical analysis and experimental verification,the value of image distortion degree under the multi-mode region division strategy is given,and the superiority of the algorithm is proved by comparing the algorithm with various homomorphic encryption algorithms with a large number of experimental images.