A novel compressive sensing-based framework for image compression-encryption with S-box

Abstract

In this paper, we find that compressive sensing (CS) with the chaotic measurement matrix has a strong sensitivity to plaintext. Because of the quantification executed after CS, however, the plaintext sensitivity produced by CS may be weakened greatly. Thus, we propose a novel CS-based compression-encryption framework (CS-CEF) using the intrinsic property of CS to provide a strong plaintext sensitivity for the compression-encryption scheme, which takes a low additional computation cost. Meanwhile, a simple and efficient Substitution box (S-box) construction algorithm (SbCA) based on chaos is designed. Compared with the existing S-box construction methods, the simulation results prove that the proposed S-box has stronger cryptographic characteristics. Based on the above works, we develop an efficient and secure image compression-encryption scheme using S-box (CSb-CES) under the proposed CS-CEF. The simulations and security analysis illustrate that the proposed CSb-CES has the higher efficiency and security compared with the several state-of-the-art CS-based compression-encryption schemes.