Spatiotemporal Chaotic System Based Image Encryption Algorithm Research

Chaos is an important subset of nonlinear science. It has several significant features, such as sensitive dependence on initial conditions, pseudo randomness, and ergodicity, etc. These features make chaos have broad applying prospect. Up to now, many chaotic cryptosystems have been proposed, but most of them are proved to be insecure. The most serious problem in individual chaotic systems is that the chaotic dynamics degrade rapidly when they are realized with finite precisions in digital computers, spatiotemporal chaotic systems are widely used in chaotic cryptography in recent years for their excellent chaotic dynamical properties. In this paper, I proposed several chaos based image encryption algorithms. The main works are as follows.(1) Propose a block image encryption algorithm based on spatiotemporal chaotic system and chaotic neural network. The employed chaotic neural network is4-input4-output, and spatiotemporal chaotic system participants in generating its weight matrix and other parameters. In encrypt process, the algorithm uses information of plaintext, so every encryption is different from each other, it is beneficial to resist known-plaintext attack and chosen-plaintext attack.(2) Propose a selective image encryption scheme using a coupled map lattice and a selector. The selector like a switch is controlled by two cyclic shift register, and the result of selection have a direct effect on the applied expression. Since adopt the selective encryption scheme, users can select appropriate proportion to encrypt the plain image for their different demands of security and efficiency.(3) An encryption algorithm for color images using a zigzag path scrambling is proposed. Initially, use the R, G, B components to form a matrix with a height of3. Then the matrix is permutated using the zigzag path scrambling. The resultant matrix is at last passed through substitution process using the typical exclusive OR operation.