An Advanced Image Encryption Algorithm Integrating Orthogonal Latin Square Transversal for Shuffling With Lorenz Chaotic System for Diffusion
Balmukund Jha, Piyush Kumar Singh, Rohitesh Pradhan, Pankaj Mishra, Vivek Kumar Jain, Fahmi Ansari · Security and Privacy · 2025
ABSTRACT In this work, we introduce a new image encryption method designed to overcome key weaknesses found in many existing techniques—like low randomness, poor sensitivity to key changes, and vulnerability to statistical attacks. Our algorithm combines two powerful ideas: first, we use orthogonal Latin square transversals to shuffle pixels in a way that avoids repetition and boosts spatial unpredictability, and second, we apply a chaos‐based diffusion process using the Lorenz chaotic system to change pixel values, which makes the encryption highly sensitive to initial conditions and resistant to intrusion attempts. The two primary steps of the encryption process are pixel permutation using organized shuffling and diffusion using chaotic sequences produced by the Lorenz chaotic system. Standard grayscale images (256 × 256) were used to evaluate the method. Results of keyspace analysis, histogram analysis, correlation analysis, information entropy analysis, NPCR and UACI analysis, and key sensitivity analysis indicate strong and robust security of the algorithm against various kinds of attacks. It only took 0.1689 s to encrypt each image, demonstrating the method's efficiency and security. Our algorithm offers a dependable solution for safe image encryption since it is complicated, resilient, and resistant to statistical and differential assaults when compared with other methods.