Grayscale image confusion and diffusion based on multiple chaotic maps

Abstract

With fast evolution of wireless telecommunication networks, all types of multimedia data may be accessed easily like digital images, texts, audio and video data. Therefore, encryption approaches are required to ensure a high level of security, confidentiality and integrity. In this work, a new version for grayscale image encryption has been proposed. Firstly, the image is divided equally to four blocks, and then each block is rotated with 90 angles in clockwise direction. Secondly, generating permutation sequence to scramble the image pixels (confusion), and generating masking sequence to alter the image pixels value (diffusion). Four different chaotic maps are used one for each block which are Quadratic map, Cross map, Chebyshev map and Ikeda map. Finally, by combining the four encrypted blocks, the cipher image is obtained. The results showed a significant security and validity to resist differential and statistical attacks. Experiments conducted showed that the presented version has robust security characteristics such as large key space (2419) high sensitivity to changes in keys and plain image (Number of Pixel Change Rate (NPCR) 99.6506% and Unified Average Change in Intensity (UACI) 33.5743% for Baboon image) and the ability to highly de-correlate adjacent pixels in the cipher image (correlation coefficient is near to zero).