Bit-level encryption of images based on hyper-chaotic system

Cryptography is one of the effective methods to prevent unauthorizedaccess to sensitive and critical data such as medical images andmilitary information. In this paper a novel method is presented toencrypt color images based on self-adaptive, bit level decompositionand hyper-chaotic functions. At first, a color image with the size ofM×N is converted to a gray M×3N image. Six pseudo-random arraysare generated using Lorenz system which is applied for imagepermutation based on rows and columns of image, respectively.Accordingly, the pixels of color components are combined togetherwhich led to change the histogram of each color components. Then,the permutated image decomposed into eight bit levels. Four bit levelsclassified into one group and the other four bit levels classified intoanother group. Six pseudo-random arrays using the Chen systemgenerated and applied to encrypt bit levels based on self-adaptiveencryption. Experimental results and security analyses show that theproposed encryption algorithm has a good encryption effect and highsecurity. Moreover, it has a strong robustness for resists varioustypical attacks.