Research And Implement Of Self-adaptive Image Encryption Algorithm Based On Chaos

With the fast development of Internet, the information-rich and intuitive images are getting more and more widely used. Comparing to text, images are more easily accepted by people. There are numbers of sophisticated theoretical studies and practical applications for traditional text encryption algorithms. How to protect the security of images with large amount of data and high redundancy is becoming a hot research topic in the cryptography.In this paper, the existing image encryption technique, performance requirements, evaluation indicators, the characteristics of chaotic system and its relationship with image encryption are summarized. The self-adaptive image encryption algorithms based on spatial-domain and frequency-domain are studied. The conclusions are listed as follows.Firstly, the concepts of cryptograph, the design principle and evaluation indicators of image encryption, the definition and characteristics of chaos are summarized. Additionally, some of the common chaos systems are analyzed.Secondly, the relationship of chaos and cryptograph, the pros and cons of common image encryption algorithm based on chaos are analyzed.Thirdly, a self-adaptive image encryption algorithm based on spatial-domain and chaos system is presented. First, for any size of gray image or colorful image, it is permuted by expanded Arnold with 1 round and then divided into 2×2 sizes of blocks. Furthermore, a matrix, which size is equal to the right-top block, is constructed with the left-top block under Chebyshev mapping. Substitution operation is performed to the right-top block with the new matrix. Encrypt each block in clockwise until the left-top block is encrypted. Several evaluation indicators are analyzed through quantities of simulated experiments.Fourthly, a self-adaptive image encryption algorithm based on frequency-domain is proposed. First, for any size of gray image or colorful image, it is expanded to meet the requirement of self-adaptive and decomposed in L level by discretize wavelet transform. Then a matrix, which size is equal to the pre-expanded image, is constructed under Logistic mapping. The wavelet coefficient matrixes are permuted by sub-matrix of the matrix. After that the permuted matrixes are composed to the cipher image by inverse discretize wavelet transform. The relationship between decomposition level and permutation times, the advantage and disadvantage of self-adaptive image encryption algorithm based on spatial-domain and frequency-domain are analyzed.Finally, some open issues of image encryption as well as the future work are presented.