Research On Reversible Data Hiding In Encrypted Images Based On Pixel Prediction

In recent years,with the rapid development of cloud computing technology,the widespread adoption of mobile intelligent devices,and ubiquitous network connectivity,more and more users are choosing to upload their private information to cloud servers.However,while technological development provides new momentum for people’s lives and work,it may also bring serious information security issues.The confidentiality,integrity and user authentication of information are facing potential threats at any time.Reversible data hiding in encrypted images(RDHEI)algorithms combine image encryption technology with advantages of previous data hiding techniques,providing a perfect solution for privacy protection of users in cloud environments.Therefore,proposing RDHEI algorithms that meet current application requirements is a hot topic in relevant research fields today.After analyzing existing RDHEI algorithms,according to the spatial correlation of digital images,this paper proposes two pixel-prediction-based RDHEI algorithms.The main works are summarized as follows:(1)Reversible data hiding in encrypted images based on nearest neighbor pixel predictionA novel RDHEI algorithm based on nearest neighbor pixel prediction is proposed to address the problem of insufficient embedding capacity in some application scenarios.To fully utilize the spatial correlation of natural images,the algorithm first designs a nearest neighbor pixel prediction method.By further exploring the prediction errors,the classification rule of prediction error blocks is obtained.Secondly,according to this rule,block shift and block rearrangement methods are designed to retain the correlation between adjacent pixels in the block.Finally,by dynamically marking the error interval within the block to maximize the space for redundancy,the embedding capacity is improved.Experimental results show that under the same test conditions,the proposed algorithm has better performance compared with other similar algorithms.(2)Reversible data hiding in encrypted images based on pixel prediction and codingIn order to further improve the performance of RDHEI algorithms in terms of security and embedding capacity,a high-capacity RDHEI algorithm based on pixel prediction and coding is proposed.By using high-precision pixel predictor and adaptive Huffman coding,the proportion of additional information is sufficiently compressed on the basis of improving prediction accuracy,effectively improving the embedding capacity.Experimental results show that,under the premise of meeting the requirements of reversibility and security,the proposed algorithm can achieve higher embedding capacity compared to other similar algorithms under the same conditions.