Hybrid Image Encryption Model Based on Zernike Moments Keys Generation and Rubik’s Cube Scrambling
Mohammed Salih Mahdi, Hayder Najm, Ali Hakem Alsaeedi, Riyadh Rahef Nuiaa, Zaid Abdi Alkareem Alyasseri, Zuraida Abal Abas · IEEE Access · 2026
This paper proposes a novel hybrid image encryption model that combines key generation based on Zernike moments and Rubik’s Cube scrambling to achieve a high level of security. The proposed approach uses Zernike moments based on a secret image to generate dynamic cryptography keys. These keys cause an initial diffusion layer to be processed with bitwise XOR, followed by a spatial permutation using a scrambling technique similar to that employed in a Rubik’s Cube, in which row and column transformations are repeatedly applied. Experimental analysis shows that the hybrid model has very good security performance, information entropy 7.9998 bits/pixel, a Number of Pixel Change Rates 99.5995%, and the Unified Average Changing Intensity 32.3631%. Spatial correlation analysis shows that the coefficients are close to zero in the horizontal (−0.0124), vertical (0.0025), and diagonal (−0.0069) directions, indicating a significant break in the relationship between pixels. The results show strong resistance to statistical, differential, and chosen-plaintext attacks, making the proposed model an effective tool for guaranteeing secure image communication in multimedia systems.