Research On Reversible Data Hiding Algorithm Of JPEG Image Based On AC Coefficient Coding

With the rapid development of the Internet,the security issue of digital information transmission in an open network environment is becoming more and more serious.Reversible data hiding(Reversible Data Hiding,RDH)technology is an effective means of information protection,and has now become a research hotspot in the field of information security.JPEG images have good visual quality and high compression,and are often used as carrier images.First,the important secret information is embedded in the JPEG carrier image using the RDH algorithm,and then the JPEG image containing the secret information is transmitted on the network to achieve the hiding of secret information and protection.The JPEG image is a lossy compressed image.Compared with the embedding capacity of the RDH algorithm using the spatial image as the carrier,the embedding capacity of the JPEG image RDH algorithm is relatively low.In order to improve the embedding capacity of the RDH algorithm using JPEG images as the carrier,starting from the analysis of the encoding and decoding process of the JPEG images,this thesis carried out related research on the structural characteristics of the JPEG images and the main performance indicators of the RDH algorithm,proposed a JPEG image RDH algorithm based on adding AC coefficients(AC Code,ACC)encoding and improve it.The main work of this thesis is as follows:(1)In order to improve the embedding capacity of RDH algorithm with JPEG image as the carrier,a RDH algorithm of JPEG image based on adding ACC is proposed.The algorithm first parses the JPEG image bitstream into an independent minimum coding unit(Minimum Coding Unit,MCU);then classifies the MCU according to the rules and divides it into usable MCU and useless MCU;secondly,it establishes secret information and ACC Mapping association table;finally,the secret information is embedded by adding ACC before the end of block(End Of Block,EOB)in the available MCU.Design algorithm verification experiments and analyze performance evaluation indicators such as the embedded capacity of the algorithm.The experimental results prove that the embedding capacity of the algorithm has increased by about 980%,170%,and 305% respectively compared with the existing three typical algorithms,but in terms of file size increase,the file size increases by 1.75 for each bit of embedded data bit.(2)In order to further optimize the file size growth of JPEG image RDH algorithmbased on adding ACC,a JPEG image RDH algorithm based on iterative adding ACC is proposed.The algorithm optimizes file size growth by establishing a better secret information and ACC mapping association table;iteratively adding ACC to further increase the embedding capacity;by exclusive-OR encryption of location map,the mapping association table is scrambled periodically,by scrambling the ACC position in the steganographic MCU to improve the security of the RDH algorithm.Design algorithm verification experiments to analyze the optimized file size growth,embedded capacity,and RDH security.Experimental results show that,in terms of file size growth,compared with the pre-optimization algorithm,when the algorithm embeds 1 bit of information,the file size increases by 1.34 bits,and the file size growth is optimized by about 25%.In terms of embedding capacity,compared with the pre-optimization algorithm,when the number of iterations is one,the embedding capacity of the algorithm is increased by about 100%,and the security of the RDH algorithm is also improved.(3)In order to verify the effectiveness of the optimized RDH algorithm for JPEG images based on iterative addition of ACC,this thesis designs an algorithm function verification system.The system is mainly composed of two modules: the sender system and the receiver system.It simulates the process of reading the image,setting the key,saving the image,embedding the data,extracting the data and restoring the image for the data hider and the legal receiver.The system verification results show that the algorithm further improves the embedding capacity,and the secret information can be completely extracted,and the original image can be recovered without loss.