Research On Reversible Data Hiding Algorithm In Encrypted Images Based On Pixel Bit Flipping

Reversible data hiding in encrypted image is a technique(RDH-EI)to hide additional information into encrypted images,which can extract information and recovery original image without error.This technology can be effectively applied to cloud storage that cloud manager can hide relevant information into encrypted images to improve the management efficiency of encrypted images when the users upload encrypted images to the cloud.With the development of cloud technology,cloud manager's demand for hidden ability of encrypted images and the users' requirements for image security and information integrity become higher and higher.To meet the demands,this thesis proposes two separable RDH-EI algorithms which take image security,embedding capacity,and the quality of decrypted image as main research objects.The main research work is as follows:To improve the security and reversibility,a reversible data hiding in encrypted image with capable of ciphertext-only attack(COA)is proposed.The image owner segments the original image into a number of non-overlapping blocks,the encrypted image is generated by combining block scrambling and the designed XOR encryption which three MSBs and five LSBs of pixels in the block are encrypted respectively with different key stream.Then the data hider uses the pixel fluctuation value and lossless compression to gain extra space to embed the additional information into pixels by bit flipping.To achieve full reversibility,the fluctuation value of the pixels is recorded in marked image.At last,the receiver extracts the embedded data correctly and recovery original image.The proposed algorithm is evaluated on uncompressed color image database.Experimental results show that the correct rate of stream cipher obtained by COA on encrypted image sets of 20,50,70,100 is about 50%,which can effectively resist COA.Compared with three similar algorithms,the embedding capacity is improved respectively by 240.85%,148.21% and 145.69%,and the quality of decrypted image is improved about 5d B.In order to further increase the embedding capacity,a reversible data hiding in encrypted image based on distortion value sorting is proposed under the above algorithm.The data hider can hide additional information by flipping the 4 LSBs of pixel to improve the accuracy of the wave function,thereby increasing the embedding capacity.And the distortion value sorting strategy that pixels which can be embedded information are sorted in descending order according to distortion value is used.Pixels with smaller distortion value are preferentially selected to hide information to improve the quality of decrypted image.At the image receiving end,the receiver can either extract the additional information before recovering original image,or recover original image before extracting the additional information.The experimental results show that the algorithm can extract information and recovery original image,meanwhile can get better quality decrypted images and higher embedding.At the same time,a GUI simulation system is designed based on MATLAB that can verify the effectiveness and feasibility of the algorithm intuitively.