Authenticated reversible image hiding algorithm based on blockchain technology

An authenticated reversible image hiding algorithm based on blockchain framework is presented in this paper. First, global eigenvalues of the plain image are extracted as plaintext keys. Meanwhile, a novel key transformation model is built, combining with public key cryptography RSA and random numbers to compute the initial values for Lorenz system. Thereby, chaotic sequences are produced. Secondly, a preprocessed model and an integer getting model are constructed to process these chaotic sequences and generate preprocessed sequences. The digital signature image and the plain image are then encrypted to obtain a cipher signature image and a cipher image. Thirdly, integer wavelet transform is executed on the carrier image to embed the cipher signature image and the cipher image in the frequency domain matrices by the least significant bit. Next, \(2^{k}\) $2^k$ correction method is employed to correct these coefficient matrices. Finally, inverse integer wavelet transform is performed on the coefficient matrices to generate the carrier image containing secrets. Especially, cipher signature image is stored in an interplanetary file system, producing content identifier hash value. Then content identifier hash value is written into smart contract and saved on Ethereum blockchain. In the stages of decryption and authentication, the receiver inquires the corresponding transaction file from blockchain by the transaction address, obtaining content identifier hash value and downloading cipher signature image from interplanetary file system. Consequently, the receiver extracts cipher signature image from carrier image containing secrets to do decryption and authentication. Experimental results show that the proposed image hiding algorithm has strong sensitivity to both key and plaintext, can resist clipping attack, and etc.