Novel image compression-then-encryption scheme based on 2D cross coupled map lattice and compressive sensing

Mingxu Wang, Xingyuan Wang, Chunpeng Wang, Zhiqiu Xia, Shuang Zhou

Abstract: A 2D cross coupled map lattice (2D-CCML) model with large key spaces is constructed for image compression and encryption, and this model exhibits better cryptographic features in dynamics than the traditional cross coupled map lattice (CCML) system. Furthermore, by combining the 2D-CCML model and compressive sensing (CS), we present an image compression-then-encryption scheme, the core of which includes image compression by CS and image encryption based on the 2D-CCML. In addition, the proposed 2D-CCML model is utilized to generate a measurement matrix for use in CS. Furthermore, the key stream used in our algorithm and the 2D-CCML model is based on the hash values calculated with the SHA-512 hash function and plain images. We experimentally demonstrate reasonable chaotic behaviors of the 2D-CCML model in terms of its image encryption effect, compression capability, and high security.