A self-cited pixel summation based image encryption algorithm* *Project supported by the National Natural Science Foundation of China (Grant Nos. 61602124, 61273021, 11526057, and 11301091), the Natural Science Foundation of Guangdong Province, China (Grant Nos. 2016A030310333, 2015A030313614, and 2015A030313620), the Science & Technology Planning Projects of Zhanjiang City, China (Grant Nos. 2015B01098 and 2015B01051), the Project Foundation of Chongqing Municipal

Abstract

In this paper, a novel image encryption algorithm is presented based on self-cited pixel summation. With the classical mechanism of permutation plus diffusion, a pixel summation of the plain image is employed to make a gravity influence on the pixel positions in the permutation stage. Then, for each pixel in every step of the diffusion stage, the pixel summation calculated from the permuted image is updated. The values from a chaotic sequence generated by an intertwining logistic map are selected by this summation. Consequently, the keystreams generated in both stages are dependent on both the plain image and the permuted image. Because of the sensitivity of the chaotic map to its initial conditions and the plain-image-dependent keystreams, any tiny change in the secret key or the plain image would lead to a significantly different cipher image. As a result, the proposed encryption algorithm is immune to the known plaintext attack (KPA) and the chosen plaintext attack (CPA). Moreover, experimental simulations and security analyses show that the proposed permutation-diffusion encryption scheme can achieve a satisfactory level of security.