Research On Image Data Hiding Based On Fusion Of Compressive Sensing

The 21 st Century is the Internet Information Age.The rapid development of Internet has connected the world,solved the time and space limitation problems,and played a significant role in promoting people's production and lives.However,its development has also triggered many security issues.Multimedia digital images are vivid information carriers and have a wide range of applications in people's production and daily life.Unlike text,since digital images have their own unique characteristics,they are vulnerable to malicious attacks during transmission.In addition,considering the unreliability of the transmission network,cloud service providers,et al.,it is necessary to perform in-depth research on how to ensure the security of digital images.Image encryption uses keys and encryption algorithms to ensure the security of images,which protects the image content from unauthorized users.Data hiding uses the self-redundancy of the image to conceal additional information in the image to achieve the functions of copyright protection,identity authentication,covert communication,and so on.Compressive sensing(CS)is an emerging sampling technique proposed in recent years,which can realize encryption while sampling.Moreover,compared to the traditional Nyquist sampling theorem,CS can recover the original signal with a lower sampling rate.These advantages bring CS the potential to be widely used in the research of data hiding.The main work of this thesis includes the following three aspects:1)In-depth understanding of the research status of image data hiding,in-depth study of image encryption,data hiding and CS theoretical knowledge,laid a solid foundation for the follow-up research.2)A robust dual watermarking algorithm based on CS is proposed.The algorithm solves the single function problem of single watermark using dual watermark.Experimental results show that the algorithm has good imperceptibility.The robust watermark in the algorithm is more robust than other algorithms,and it can resist common attacks better.The semi-fragile watermark in the method makes full use of the advantages of the emerging CS technology,which achieves image tamper positioning and recovery.3)A controllable encrypted domain data hiding scheme combining CS and data pretreatment is proposed.The scheme is a novel high-capacity hiding scheme in encrypted domain,which is suitable for privacy protection and ciphertext management in the cloud environment.There are three roles in the scheme.The image owner first carries out the secondary wavelet decomposition of the image,and then compresses the wavelet coefficients using CS and the traditional encryption method to obtain the encrypted image,and transmits the encrypted image to the data hider.Then,the data hider preprocesses the secret data and embeds it in the received encrypted images.Finally,the authorized users can recover images and/or extract the embedded secret data based on different keys.Compared with the traditional algorithms,the proposed algorithm achieves the advantages of high quality recovered image,large data embedding capacity,separablity between image recovery and data extraction,etc.