An efficient approach for encrypting double color images into a visually meaningful cipher image using 2D compressive sensing

Abstract

An efficient visually meaningful double color image encryption algorithm is proposed by combining 2D compressive sensing (CS) with an embedding technique. First, two color images are measured by measurement matrices in two directions to achieve simultaneous compression and encryption, in which low-dimensional matrices generated from Logistic-Sine system (LSS) are extended with Kronecker product (KP), and the resulting high-dimensional matrices optimized by singular value decomposition (SVD) are employed as measurement matrices. Second, the compressed cipher images are confused by index sequences produced by a 6D hyperchaotic system. Finally, a visually meaningful cipher image is obtained by embedding permutated cipher images into a color carrier image. The final cipher image and plain image are of the same size, which greatly reduces the storage space and transmission bandwidth. To enhance the relationship of our algorithm with plain images and prevent vulnerability to known-plaintext and chosen-plaintext attacks, SHA 256 hash values and feature parameters of plain images are combined to generate the initial values of the LSS and 6D hyperchaotic system, and these parameters are both embedded into the carrier image to avoid additional transmission and storage. Simulation results and performance analyses demonstrate the effectiveness and security of the proposed image encryption scheme.