Block cipher based separable reversible data hiding in encrypted images

Abstract

While most reversible data hiding in encrypted images (RDH-EI) are based on stream cipher, this paper aims to present an alternative method feasible for block-enciphered images. Before uploading data to a remote server, the content owner encrypts the original image with a block cipher algorithm using an encryption key. Then, the server embeds additional bits into the encrypted image with an embedding key to generate the marked encrypted image. On the recipient side, the additional bits can be extracted if the receiver has the embedding key. In case the receiver has only the encryption key, the marked encrypted image can be directly deciphered to a plaintext image with good quality. When both the embedding and encryption keys are available for the receiver, he can recover the original image without any errors. Compared with the existing block cipher based RDH-EI method, drawbacks of the encryption and the recovery are avoided, and good embedding payloads are achieved.