Modified bernoulli map-based scramble and s-box supported colour image encryption

Abstract

The exponential growth of digital image sharing has amplified concerns regarding data privacy and security, especially for colour images of varying sizes and resolutions. Traditional encryption algorithms often fall short in balancing speed, scalability, and robust security for such diverse image datasets. Addressing this gap, we introduce a novel colour image encryption scheme that synergizes modified Bernoulli map-based random number generation for pixel scrambling with an S-Box-supported diffusion process. Our approach first employs a chaotic random number generator to effectively reorder pixel positions, enhancing confusion. This is followed by a diffusion phase utilizing a robust Khan S-Box to introduce nonlinearity and further obfuscate pixel values. To evaluate the security and efficiency of our method, we conducted extensive tests including differential cryptanalysis using NPCR (Number of Pixel Change Rate) and UACI (Unified Average Changing Intensity) metrics. The results demonstrate that our encryption system exhibits high resistance to differential attacks and achieves superior performance compared to existing methods. By combining fast random number generation with strong S-Box diffusion, our scheme offers a scalable and secure solution for real-time colour image encryption, contributing significant advancements to the field of cryptographic image processing.