Research And Implementation Of High Capacity Reversible Steganography Scheme

Steganography is a technology which can protect information security by hiding its existence.By using the redundant information of multimedia itself and the limitations of human perception system,the secret data such as tag data and additional information can be embedded to the original cover data without affecting the value of the original data.At present,most steganography algorithms only focus on the embedding and extraction of secret information,which usually brings permanent distortion to the original cover data.There are many challenges in scenarios that require high data authentication,such as remote sensing applications,medical image processing and military applications.In this paper,digital images are taken as the research object.Compare with existing data hiding schemes,we would research on how to embed more data under high integrity requirements,and to satisfy the situation of image tampering in unsecure channels.The main work and research results are as follows:(1)A reversible steganography algorithm based on homomorphic encryption and pixel prediction is proposed.In this algorithm,the redundant space from the plaintext domain of the image is directly transferred to encrypted domain.According to the characteristics of the Paillier homomorphic encryption algorithm,a reversible data hiding method is used to embed the secret data directly into the encrypted image.By using the decryption key,secret data can be directly extracted in the encrypted image,and the cover image can be obtained by the reconstruct algorithm.Experimental results show that our algorithm is completely reversible.When experiment on a dataset of 1000 images and 90 complex texture images,the results also have a satisfied capacity performance over 0.9bpp.(2)A cover image robust reversible steganography algorithm based on block chain authentication is proposed.The algorithm uses a fragile watermark to mark the tampered location of the cover image,and can recover the tampered information through authentication of the related blocks.The proposed scheme can improve the peak signal-to-noise ratio of the reconstructed image when tampering occurs.At the same time,it can be combined with the reversible data hiding scheme in the above research and carried out independently.In both secure channel and unsecure channel,the security and integrity of the cover image and secret data can be guaranteed.Experimental results show that when marked image is tampered with 10%,our method can restore the peak signal-to-noise ratio of the restore cover image over 50dB.Proved our scheme has effective recovery ability.