Adaptive reversible image-data hiding algorithm based on blockchain and zero watermarking

An adaptive reversible image-data hiding (ARIdH) algorithm is proposed. First, the variance of the secret plain image is calculated and taken as feature values, by a new integer extraction model (NewIEM). Then, a new key transform model (NewKTM) is constructed and all feature values are converted into initial values. Second, discrete wavelet transform (DWT) is applied to the secret plain image, and a new adaptive threshold sparse model (ATSM) is designed to sparsify the DWT coefficients. Third, the sparse matrix is compressed and measured by a measurement matrix produced by NewNM to obtain a middle cipher image (MCI). Then, MCI is divided into blocks, and executed by both XOR diffusion between blocks and spiral scanning diffusion within blocks, to generate a new cipher image (NCI). After that, a carrier image is confused by random sequence and NCI is embedded into it by matrix encoding technique. To realize copyright authentication, an authentication framework based on blockchain and zero watermarking is designed. Specifically, perform an XOR operation between the feature matrix of NCI and the QR code containing copyright information to generate a zero watermarking. Then, smart contracts are designed to store the data and used for authentication.