A novel dual-layer security scheme based on 3-cell cellular neural network encryption and data hiding scheme for color images

The security of digital information is one of the significant problems of the existing world. In this work, we have proposed a dual-layer information hiding mechanism. We have utilized 3-Cell Cellular Neural Network (CNN) system to generate a pseudorandom number generator to fulfill diffusion requirements of a cryptosystem, and then utilizing highly non-linear substitution boxes (S-boxes) to complete the confusion process of cryptosystem design. The encrypted data are the hide in the cover image, the data hiding process comprises three steps, generation of transform matrix from the cover image, embedding matrix from transform matrix and secret encrypted image and the last step is to obtain a stego image. Less computational complexity without affecting the strength of the security scheme is the need of the day. Due to the simplicity and computational efficiency of the diffusion process of a 3-cell cellular neural network, the confusion created by predefined S-boxes, and the data hiding scheme effectively reduces the time for encryption and data hiding. The proposed scheme has been tested using state-of-the-art key performance indicators including information theory-based analysis, Histogram based analysis, analysis based on pixels difference, correlation-based measurements, and distance-based analysis. the results of this analysis show the efficiency and effectiveness of the proposed system.