Research On Color Image Encryption Algorithm Based On Chaos

The rapid development of communication technology and real-time information interaction technology has made people transmit more and more data on the Internet,and information security is facing more and more severe challenges.Nowadays,more and more images are generated,transmitted and shared on the Internet.More and more people like to share their personal image with relatives and friends in cyberspace.In addition,many fields require a large number of images for transmission,such as remote diagnosis,traffic monitoring and satellite observation.As a typical multimedia resource,digital images are facing more and more prominent security issues.Compared with traditional text data,the characteristics of image information are quite different,such as strong correlation properties between adjacent pixels,high repetition of pixel information,and huge amount of data.With the development of chaotic systems,chaotic systems have become a widely used encryption technology due to their pseudo-randomness,ergodicity,and initial state sensitivity.With the research of scholars,a large number of image encryption systems based on chaotic technology have been proposed,but they all have some shortcomings to some extent.Therefore,it has become a top priority to propose algorithms with better performance and efficiency on the basis of existing algorithms.Based on the existing research on image encryption algorithms based on chaos technology,this thesis proposes two image encryption algorithms,and conducts experimental simulation and related security analysis on C++.One algorithm is a parallel remote sensing image encryption algorithm of chaotic system and DNA encoding,which solves the problems of poor speed and insufficient security when encrypting remote sensing images and large-capacity images.The algorithm uses GPU parallel computing key sequence and image grouping to use GPU parallel computing to improve security,and uses DNA-S box to non-linearly replace the scrambled DNA code to improve the security of the algorithm.First,calculate the digest of the second-generation secure hash algorithm SHA-256 of the plaintext image to update the parameters and initial key of the chaotic system to improve the plaintext sensitivity of the algorithm,and to scramble the image and pixels through the two-dimensional Hénon-Sine mapping.Then use GPU to calculate the key sequence in parallel to shorten the encryption time,select multiple high-dimensional chaotic systems and modify the initial value of the chaotic system to ensure the randomness of the key sequence;finally use the key sequence and GPU to pair the image Perform DNA parallel encryption to get the final ciphertext image.In the DNA parallel encryption process,a DNA-S box is generated to perform non-linear replacement of the DNA code.Simulation experiments and security analysis results on remote sensing images and classic Lena color show that the algorithm in this thesis has a speed of more than 80 Mbps on encrypted remote sensing images,the key space is increased to 10 to the 213 th power,and the information entropy approaches 8.The histogram and the distribution of adjacent pixels obey uniform distribution,and have passed the NIST random test and chi-square test;compared with other algorithms,the algorithm in this thesis is in key space,adjacent pixel correlation,pixel change rate,uniform average change intensity,Information entropy and other evaluation indicators are closer to the ideal value.The algorithm not only greatly increases the encryption speed,but also ensures that the algorithm is secure enough to resist various attacks,and is suitable for the confidential storage and network transmission of remote sensing images and large-capacity images.The other is aimed at the practical application requirements of high efficiency and security of image transmission in general equipment.In order to further improve the versatility and applicability of the first proposed algorithm,while ensuring the security of the algorithm,it combines chi-square check and pixel frequency.A dynamic S-box color image encryption algorithm based on chaos and plaintext correlation is designed.The algorithm first calculates the SHA3-512 message digest value of the plaintext image to update the initial value of the chaotic system,then performs Arnold scrambling encryption on the plaintext image,and then uses the chi-square check and the pixel frequency to generate the interference value of the dynamic S-box and the key sequence stream Finally,perform a diffusion operation on the image.The diffusion process uses a dynamic S-box to perform non-linear replacement in the process of encrypting pixels one by one,and uses interference values to interfere with the value of the key sequence.In the diffusion stage,the algorithm enhances the diffusion of the algorithm by generating a dynamic S-box related to the plaintext,updating the interference items through the parameters related to the cipher pixel stream,and enhancing the randomness of the key stream,thereby enhancing the security of the algorithm.The simulation experiment analysis on the classic Lena color image shows that the key space of the algorithm is large enough to effectively resist brute force attacks.The value of information entropy is close to the theoretical value of8.The key sensitivity and plaintext sensitivity analysis proves The diffusibility of the algorithm has been enhanced.