Toward robust image encryption based on chaos theory and DNA computing

Abstract

Due to the significant importance of image security for various users, there is an ongoing need to develop innovative algorithms to enhance this security. Image security typically involves encryption techniques. This study has tackled the challenge of creating an efficient, secure, and resilient image cipher by using pixel-swapping techniques at both DNA and decimal levels. The swapping methods include four different approaches that involve randomly selecting pixel pairs to swap with adjacent pixels—either left, right, upper, or lower—based on random numbers generated by a chaotic map. Specifically, the 2D Tinkerbell chaotic map was used to generate the necessary random numbers for diffusion and confusion processes in the encryption. Additionally, through careful arithmetic operations, two more random number streams were derived from the main streams produced by the chaotic map. Thorough performance analyses and computer simulations have shown that this image cipher is robust, secure, and resilient against various threats, making it suitable for practical applications. Notably, the cipher achieved a very high information entropy value of 7.9975, indicating its effectiveness in encryption.