Research On Chaos-based Image Encryption Technology And Cryptanalysis

With the rapid development of network technology,the dissemination and sharing of multimedia data(such as images,videos,etc.)among users has become more convenient.However,due to the reliance of digital technology,the issue of data security has also attracted widespread attention.In some applications,image data(such as personal electronic photo albums,sensitive medical images,military images,etc.)sent through the Internet or other channels must be protected to prevent eavesdropping,interception,and tampering by unauthorized persons,which may cause inability to estimated loss.Therefore,the problem of secure transmission and access of images in the network has attracted more attention,and has become the focus of research by many scholars.The image has the characteristics of huge amount of data,high information redundancy,scattered important information,and strong correlation between adjacent pixels.Due to its inherent characteristics,such as sensitivity to initial conditions,ergodicity,unpredictability and good pseudo-randomness,chaotic systems provide an effective way for image encryption technology.As a branch of modern cryptography,chaotic cryptography has been extensively studied in recent years.However,many image encryption schemes based on chaos currently proposed cannot resist cryptanalysis attacks and cannot be used for secure communication.Therefore,in view of the low security and low encryption efficiency of existing image encryption algorithms,this article focuses on the cryptanalysis of chaotic image encryption algorithms,the design of chaotic image encryption algorithms,the design of chaotic systems,and the use of chaotic image encryption methods in electronic medical care.The following research is carried out:(1)Cryptanalysis and improvement of image encryption algorithm at bit-level based on multiple chaotic mapsAiming at an image encryption algorithm(IESBC)based on multiple chaotic maps and bit-plane extraction recently proposed by Shafique et al.,this paper finds its security loopholes through cryptanalysis,so a known plaintext attack and a chosen plaintext attack that successfully cracks the ciphertext image without the original key.Then,in view of the security flaws of the IESBC,an improved probabilistic image encryption algorithm is proposed,which add the statistical characteristics of the plaintext image in the diffusion stage,so that each pixel of the ciphertext image is closely related to the plaintext image.At the same time,a random number seed is introduced,making the encryption algorithm a probabilistic algorithm.Even if the same plaintext image is encrypted multiple times,different ciphertext images will be obtained.The security analysis and performance evaluation of the improved algorithm are carried out.The simulation experiment results show that the improved algorithm can effectively resist various common types of attacks while maintaining the advantages of the original algorithm.(2)Cryptanalysis and improvement of image encryption algorithm based on coupled sine chaotic map and blockAiming at a recently proposed image encryption algorithm(CSMIE)based on coupled sine chaotic mapping and by Behzad et al.,this paper found that when the size of the plaintext image and the block are the same,the key-stream is fixed,so CSMIE cannot resist plaintext attacks.Therefore,this article proposes a chosen plaintext attack to crack the CSMIE,analyzes three cases according to the number of blocks,and solves the equivalent key of CSMIE with the least overhead,and successfully recovers the plaintext image.Then,in view of the flaws of the CSMIE,this paper proposes an improved image encryption algorithm to conduct security analysis and performance evaluation of the improved algorithm.The experimental results show that the improved algorithm has better encryption effects than the original algorithm,and can resist various common types of attacks effectively.(3)Image encryption algorithm based on 2D-HCMM and genetic operationsAiming at the problems of one-dimensional chaotic mapping with few parameter variables,simple chaotic trajectory,and easy to be cracked by brute force;high-dimensional chaotic system calculation is complicated,the implementation cost is high,and it is not suitable for real-time applications.This paper takes into account the security of chaotic mapping and the complexity of implementation.By cascading Henon mapping and Chebyshev mapping,a new two-dimensional Henon-Chebyshev chaotic system(2D-HCMM)is proposed.The dynamic analysis of 2D-HCMM through phase diagrams,bifurcation diagrams,Lyapunov exponents and information entropy proves that it has better ergodicity,unpredictability,and larger chaotic range and more parameters than the original chaotic map,so it is more suitable for application in image encryption.Then,using the principles of gene recombination and gene mutation,an image encryption scheme based on 2D-HCMM and gene manipulation methods is proposed.Security analysis and performance evaluation experiments show that the proposed encryption algorithm can achieve a good balance between security and efficiency.(4)A privacy-preserving system for electronic medical image based on double chaotic mapsAiming at the security,integrity,and real-time requirements of electronic medical images in the transmission and storage process,a privacy-preserving system for electronic medical image based on double chaotic maps is proposed.The system is in ciphertext form during transmission and storage,and only authorized users are allowed to decrypt,preventing patients’ medical images from being attacked by man-in-the-middle,thereby protecting privacy of patients.This paper adopts a parallel mode cryptographic system to reduce the time consumed by encryption to meet the real-time processing requirements of medical data collected in real scenarios.The system has designed a key update mechanism,so that the key used for each medical image is different,which ensures the high security of one-time pad.The decrypted medical image is completely consistent with the collected original image data,and the integrity of the image can be verified to prevent important medical image information from being tampered.Finally,the security analysis and performance results verify that the system has obvious advantages in safety,efficiency and robustness.