Research On Distributed Reversible Steganography Method Based On Secret Sharin

The current steganography methods can be divided into one-to-one steganography and distributed steganography with multiple participants.Complex scenarios of multi-party communication such as multi-departmental joint computing and multi-source data query are often difficult to ensure the privacy of personal information when implemented by one-to-one steganography,and it is also difficult to recover and less robust when unexpected situations occur during the communication process.In contrast,distributed steganography can be used to communicate using secret sharing,quantum channels,vector products,and neural networks,which can significantly improve privacy and robustness.However,the sub-secrets generated by the current secret sharing methods used in secret sharing-based distributed steganography can appear to have no readability,which is not in line with the idea that steganography requires that the secret-containing carriers are transmitted in a common channel and differ slightly in readability and statistics from the non-secretcontaining carriers,and it is extremely dependent on the complete operation of the secret sharing protocol itself.In this thesis,we propose two secret-sharing-based steganography methods to address the current problems of secret-sharing-based distributed steganography.The specific work is as follows:(1)A communication method based on secret sharing and distributed steganography is proposed,followed by a many-to-one reversible steganographic algorithm based on this secret sharing scheme.The steganographic communication method is divided into three steps: 1)the senders processes the secret message using transformation;2)the distributed storage backup and sharded communication of the cover containing secrets after the secret message is embedded in the cover using the secret sharing scheme,which has a computational complexity and strong robustness;3)the receiver receives the sharded containing secrets and uses zero-knowledge proof to distinguish the sender’s identity locally,which realizes the senders to send the secrets to the receiver without revealing the personal identity,which can avoid the attention of external attackers to a certain extent.Then this chapter extends the second step of the secret sharing scheme to support more senders and also gives a detailed correctness proof.The many-to-one steganography algorithm based on this secret sharing scheme uses the sub-secrets in the secret sharing scheme as covers,and the conditional random numbers in the secret sharing scheme are associated with the covers,whereby the covers are modified and the modified covers can proceed to the next secret sharing step.The experiments show that the steganography algorithm can realize the secerts transmission without affecting the original secret sharing sheme’s function.(2)A many to one reversible steganography scheme based on primitive polynomial and code is proposed.First,the first sender combines the primitive polynomial and secret into bit strings,which is the message to be embedded;Afterwards,the message is matched with the cover,and the cover is modified based on the matching results to obtain a cover containing secrets.Secondly,the secret of another sender is also embedded into the copy of the encrypted image using primitive polynomial to obtain another encrypted image;After receiving the above two encrypted images,the receiver can obtain both the secrets and the original image by using primitive polynomial to take modulus in the binary domain.This scheme has minimal changes to the cover and can hide multiple bits with each modification.The shared image obtained with the same embedding amount has better quality,making it more suitable for steganography compared to other secret sharing schemes.