Research On Fingerprint Image Encryption Technology Based On Chaos Theory

In the 21st century,biometric identification technology is gradually replacing the traditional security technology of password and certificate.Fingerprints are the most widely used feature in biometric identification due to its uniqueness and permanence.The fingerprint database is an important basis for identification and authentication,and then becomes an important target of hacker attacks.In 2018,the Indian citizen database Aadhar,which contains fingerprint information,was attacked and private information was leaked.It is imperative to protect this type of information.In order to better protect images,especially fingerprint images,this thesis studies and realizes fingerprint image encryption technology based on chaos theory based on the existing research results of chaotic cryptography,and has achieved the following results:(1)As Zigzag transform is not suitable for non-square matrix and low encryption degree,A fingerprint image encryption algorithm based on Lorenz chaotic map and chaotic shift transform is proposed.Firstly,the plaintext image is scrambled and the first round XOR diffusion is performed to change the statistical information of plaintext image.Secondly,the improved Zig Zag transform is used to scan and transform the middle ciphertext image,the pixel values are changed by using XOR operation.Finally,a lot of experiments are carried out,and the results show that the method can not only realize the encryption of fingerprint image of any shape,but also has a good image protection function.(2)In view of the fact that some chaotic encryption algorithms can not effectively resist the chosen plaintext attack,a fingerprint image encryption algorithm based on Lorenz hyper-chaotic system and sub-matrix rotation is proposed.Firstly,a chaotic sequence is generated by Lorenz hyperchaotic system for the rotation scrambling of sub-matrice.The size,position and rotation mode of the sub-matrices are determined by the values of the chaotic sequence,and all the sub-matrices are rotated in turn to scramble the whole plaintext image.Finally,the forward and reverse XOR operations are used.SHA-256 hash algorithm is introduced in the whole encryption process to associate with plaintext image.The difference in the plaintext image will make the encryption key different.The simulation results show that the encryption algorithm can resist common attacks and has a large key space.(3)In view of the single DNA coding rules and the inability to effectively resist the chosen plaintext attack,a fingerprint image block encryption algorithm is designed by using Chen hyperchaos and Logistic chaotic systems,combined with dynamic DNA coding operation.Firstly,the chaotic sequence generated by hyperchaotic system is processed to make it suitable for encryption operation.Then,one of the sequences is selected for plaintext association scrambling.In the process of diffusion,the scrambled image and the random matrix generated by logistic are evenly divided into blocks,and the dynamic DNA coding operation are adopted.This method can solve the problem of single DNA coding rule Simulation results show that the algorithm has good ciphertext statistics and high key sensitivity,and can effectively resist differential attack.Experimental simulation of the proposed algorithm shows that the three algorithms have good encryption effect,with high information entropy,large key space,high key sensitivity and easy implementation.They can be applied to network security communication can solve the fingerprint image for stolen in the process of transmission and suffered the problem of replay attacks,which has certain theory value and broad application prospects.