Research On Reversible Data Hiding In Encrypted Image With Public Key Cryptography

With the continuous development of cloud computing,users can quickly and easily access large amounts of image data stored on cloud servers.However,while bringing convenience,some problems have been exposed.Once the cloud is attacked,attackers will directly steal digital images and cause privacy leaks.Therefore,reversible data hiding in encrypted images(RDHEI)was proposed by researchers to solve these problems.In the schemes of RDHEI,the image provider encrypts the image and uploads it to the cloud server.The nature of the encryption algorithm allows additional data to be embedded directly into the encrypted images.According to the secret key,the additional data can be extracted in the ciphertext,and the original image can be obtained by the corresponding decryption key.This ensures that the cloud server analyzes or processes the hidden data without knowing the original content,thereby ensuring the privacy of the cover images.In the entire RDHEI process,on the one hand,the security and applicability of the encryption algorithm must be considered,and on the other hand,the two important standards of embedding rate and image quality must be guaranteed.Based on these two aspects,this thesis proposes Reversible Data hiding in Encrypted Images with Public Key Cryptography.A research scheme of High-Capacity Reversible Data Hiding for Encrypted Images with Somewhat Homomorphic Encryption is proposed.This scheme uses SHE(Somewhat Homomorphic Encryption)to encrypt images.It can not only ensure the high security of the image,but also reduce the computational complexity.In order to improve the embedding rate and image quality,we select three adjacent pixels in the image as a group,and each group can use the intermediate pixel as a reference to calculate two differences.Then embed the secret data by way of the difference histogram expansion.At the same time,the recovery of the image and the extraction of secret data in this solution can be operated separately.Experiment shows that the decrypted image quality and embedding capacity of this scheme are better than other methods proposed in recent years.A reversible data hiding research scheme based on Dual binary tree and prediction error expansion in the public key system is proposed.SHE is used to not only perform arithmetic operations directly in the ciphertext image,but also has a shorter ciphertext length and lower computational complexity.In this scheme,the dual binary tree is combined with prediction-error expansion to embed data,and the dual binary tree address the problem of transmitting multiple peak points.In addition,the operation of sorting the prediction errors can preferentially embed the secret data into the pixels with small change in pixel value,which improves the visual quality in some degree.Finally,the receiver has two options.One option is extracting additional data in the ciphertext domain by using the data hiding key.The other of the two is decrypting the ciphertext by using the decryption key and extracting the data in the plaintext domain,and restoring the original image with the data hiding key.Experimental results demonstrate that our scheme has a better image quality than the previous algorithms in the same amount of embedding.