A New Image Encryption Scheme Based on Fractional Order Neural Network

To effectively deal with the complex and changeable network environment and meet the high standard of information transmission security, this paper designs a convenient and efficient image encryption algorithm on the incommensurate fractional-order neural network model system. The algorithm uses the generated random number sequence to construct an index matrix and combines the Hilbert curve traversal principle to displace each position of image pixels. The image is converted into a DNA-encoded image by DNA coding technology. The DNA addition principle completes the diffusion operation, and the ciphertext image is obtained. Simulation results and security tests show that the proposed algorithm can encrypt image information effectively and has good anti-attack capability.