Purpose: This study aims to conduct a comparative analysis of image encryption techniques, focusing on the performance and security of systems based on Linear Feedback Shift Registers (LFSRs) against those based on chaotic maps. The research addresses a gap in the literature by systematically evaluating higher-order and combined LFSR designs, and demonstrating their viability as a lightweight and secure alternative for image encryption.

Methods: We implemented and tested several image encryption algorithms, including novel LFSR-based ciphers and established chaos-based systems using logistic, tent, and Henon maps. We evaluated their performance using a suite of security metrics, including statistical analysis (histograms and correlation), and differential attack analysis (NPCR and UACI). The efficiency of each cipher was also assessed through execution time measurements.

Results: The experimental results demonstrate that all tested ciphers, including the novel SCLFSR24, achieved robust security, with NPCR values exceeding 99% and UACI values greater than 40%. This is associated with high sensitivity to pixel changes and strong resistance to differential attacks. The LFSR-based ciphers generated similarly random keystreams and secure encrypted images, performing comparably to the well-known chaos-based methods.

Conclusion: The findings suggest that LFSR-based ciphers offer a compelling and practical alternative for image security. Their lightweight design and strong security performance, comparable to chaotic systems, are particularly well-suited for real-time applications in fields such as healthcare and multimedia communication.