Acousto-optic cryptosystem based on dynamic DNA encoding and hyperchaotic system in gyrator domains

Abstract

Traditional encryption algorithms like DES suffer from the issue of key length (128 bits to 256 bits) and the inconvenience of memorization. Users typically store keys in a file or device, or even directly on a computer hard drive, releasing the key through a relatively easy-to-remember passphrase. Addressing these issues, this paper proposes a novel remote sensing image encryption system based on a hybrid model incorporating a 4D Lorenz chaotic system, Gyrator transform, Acousto-optic Conversion system, and dynamic DNA encoding. Firstly, the original image is scrambled using a 4D Lorenz mapping function. Subsequently, employing an Acousto-optic modulation algorithm, audio information is fitted into image information and serves as part of the encryption system key, significantly enhancing the correlation between the key and the user. Then, employing a dynamic DNA encoding algorithm, the image is encoded using eight different DNA encryption schemes and three distinct computational rules, achieving an exponential expansion of the key space with minimal computational overhead. Finally, through comparative analysis, the Gyrator transform, demonstrating superior performance, is selected to further encrypt remote sensing image data, expanding the key space. The proposed remote sensing image encryption system exhibits high security performance and robustness against various attacks.