Optical double-image cryptosystem based on generalized singular value decomposition and five-dimensional hyperchaotic maps.

In this paper, we propose an asymmetric optical double-image cryptosystem based on generalized singular value decomposition (GSVD) and five-dimensional (5D) hyperchaotic maps. In the proposed cryptosystem, the two plain images are first decomposed into five components by the GSVD operation. The two unitary matrices obtained by GSVD are encoded as a complex function, which is then modulated by the chaotic random phase masks (CRPMs). The private key and the final encryption result are generated by phase-truncation and amplitude-truncation operations. The GSVD operation can decompose two images at the same time and is used to generate the private key that enables the encryption process to be asymmetric. Compared with the existing phase-truncated-based cryptosystems, our cryptosystem can improve security against a special attack. In addition, the CRPMs are generated by 5D hyperchaotic maps, which have a larger parameter space and better randomness. Numerical simulation results are shown to verify the feasibility and robustness of our cryptosystem. Furthermore, the proposed cryptosystem can be extended to encrypt multiple images conveniently.