A novel permutation-diffusion technique for image encryption based on the Imperialist Competitive Algorithm

Recently, there has been explosive growth in the sharing of images on social media networks. This has occurred primarily because people are using smart phones with high quality imaging facilities and a mobile Internet connection. Accordingly, digital image security has formed a particular focus of recent research. Image encryption is the most common method used to keep images safe during their transmission over the Internet. There are two main phases in image encryption: permutation and diffusion. In this paper, a hybrid model is proposed, which is a combination of a novel permutation technique in conjunction with a new diffusion method. In the permutation phase of the suggested model, pixels are relocated based on their indices. The resulting permuted image has a high difference (more than 99.3%) in the gray level of pixels compared to the corresponding pixels in a plain image. The proposed diffusion method is based on a combination of an Imperialist Competitive Algorithm (ICA) and adjacent lattices as a chaotic function. To date, the ICA, as an evolutionary algorithm, in combination with adjacent lattices, has never been exploited for image encryption. The result analysis demonstrates that the proposed method has excellent resistance against brute force and statistical attacks, in addition to obtaining a 7.9993 entropy score for the encrypted image.