Reversible Information Hiding Based On Dynamic Prediction And High-Dimensional Prediction Error Histogram Shifting

With the rapid development of Internet technology and the ever-changing multimedia information technology,the generation,transmission,storage and change of multimedia digital information such as texts,images and videos are becoming easier and easier,and people can easily download,upload,tamper and distribute the multimedia information through the Internet.The ensuing series of security issues are related to personal privacy,corporate secrets,social stability and national security.Therefore,we must take measures to protect the information security.Cryptography is one of the most effective ways to protect the security of digital information,but it also has some limitations.Cipher technology protects information by encrypting the plaintext into an incomprehensible ciphertext,and it is the incomprehensibility of the ciphertext that exposes the importance of the information,making the protected information more vulnerable to attack and sabotage by attackers.Information hiding,as another important technology in the field of information security,can solve this problem well and complement each other with cryptography.Information hiding technology is widely used in hidden communications,copyright protection and content integrity certification,and the effect is remarkable.In recent years,the proposal of reversible information hiding has solved the problem of non-recoverable cover medium,and thus has received close attention in some sensitive areas such as military,medical,biology and law.Reversible information hiding can hide the secret information into the cover medium.Legitimate users can use the decryption algorithm to extract the secret information completely,and at the same time,it can recover the original carrier without distortion.Most of the reversible information hiding has a serious distortion of the carrier after embedding the secret information,while some sensitive areas are unacceptable for serious carrier distortion.In order to effectively improve the quality of stego image and solve the problem of serious carrier distortion,this thesis presents a reversible information hiding algorithm based on dynamic pixel prediction and high-dimensional prediction error histogram shifting.In terms of pixel prediction,the pixels are divided into four independent categories to expand the neighborhood of the central pixel and more strongly related pixels are selected as the prediction pixels according to the pixel selection threshold.According to the pixel fluctuation in one direction,different weights are allocated to the filtered prediction pixels,so that the prediction weights are reasonably allocated.The pixel prediction algorithm after dynamic pixel screening and weight redistribution can make full use of the adjacent pixel correlation,and the prediction accuracy is better.The difference between the original pixel and its predicted value is calculated to obtain the prediction error sequence,and then the prediction error is expanded to embed the secret information.In terms of prediction error expansion,unlike the traditional singlepixel processing to embed secret information,the prediction error is synthesized into a prediction error group and used as an embedding unit.The local texture complexity is used to distinguish the image smooth region and the rough region.The local texture complexity of each embedding unit is calculated,the prediction error group of the relatively smooth region of the image is selected according to the threshold value,and finally the secret information is embedded by high-dimensional prediction error histogram shifting.The high-dimensional prediction error mapping scheme discards the extended direction of serious distortion to derive a new reversible mapping,which improves the embedding performance of the algorithm.The experimental results show that the proposed algorithm improves the accuracy of pixel prediction,ensures the embedding capacity and effectively controls the image distortion,which has certain research significance and application value.