Bird swarm optimized chaotic lattice framework for parallel multi image encryption

The exponential growth of multimedia communication demands secure, efficient, and scalable image encryption frameworks capable of handling high-dimensional correlated data. Existing chaotic and hybrid encryption schemes demonstrate strong nonlinearity but often suffer from redundant computations, weak inter-channel coupling, and limited adaptability in multi-image settings. In this work, we propose BLINK (Bird Swarm Optimization (BSO)-Hypertuned Lattice-Based Image eNcryption Kernel), a novel entropy-preserving multi-image encryption framework integrating chaotic lattice dynamics, eigen-coupled channel interaction, and optimization-driven parameter tuning. The proposed approach introduces a PSI-based amortized eigen-coupled chaotic lattice mechanism that enables efficient reuse of invariant chaotic structures while maintaining non-deterministic behavior. BSO is employed for adaptive hyperparameter tuning and secure seed generation, with cryptographic safeguards ensuring randomness quality. Furthermore, the framework aligns with post-quantum security principles through lattice-based constructs inspired by learning with errors. Experimental evaluations demonstrate superior entropy, reduced pixel correlation, and enhanced resistance to statistical and differential attacks compared to existing schemes, making BLINK suitable for secure multi-image encrypted transmission.