DNA coding and chaos based image encryption using compressive sensing in MSVD domain

This paper introduces a novel image encryption technique using Compressive Sensing (CS) and DNA encoding. At first, the plain image is decomposed into the low and high-frequency sub-bands through Multi-Resolution Singular Value Decomposition (MSVD). The core idea behind the use of MSVD is to decompose the whole information into significant and less significant information. Significant information contains the information which can best represent the signal, so the low-frequency sub-band values are encrypted through the confusion and diffusion process. Here first diffusion is applied through DNA XOR operation; after that, Arnold scrambling is used to permute the encrypted information. Next, simultaneous compression encryption is employed on less significant information as high-frequency sub-bands through CS. At last, all sub-bands are concatenated to form a single encrypted image. Finally, the proposed algorithm is tested against statistical, differential, clipping, and noise attacks. Simulation results illustrate that the given algorithm is secure and robust against attacks.