Confusion and diffusion of color images with multiple chaotic maps and chaos-based pseudorandom binary number generator

Abstract

In this communication, a chaos-based symmetric key encryption of RGB color images with multiple chaotic maps and pseudorandom binary number generator (PRBNG) has been proposed. The original image is first scrambled using generalized Arnold cat map to achieve confusion. The scrambled image is then encrypted using chaotic sequences generated by multiple one-dimensional chaotic maps. Selection of the map is made pseudorandom by letting the selection be determined by the output generated by a chaos-based pseudorandom binary number generator. Encryption based on multiple chaotic maps where the choice of map is governed by the output of PRBNG leads to the most important contribution of this work: It affords a high degree of security by providing a very large key space. Statistical tests like histogram analysis, correlation coefficient analysis, measures of central tendency and dispersion along with key sensitivity test, key space analysis, information entropy test have been used. To demonstrate the efficiency and superior performance of the proposed approach as compared to existing techniques, the metrics mean squared error, peak signal-to-noise ratio, number of pixels change rate and unified average changing rate have been used for a quantified assessment of the encryption quality.