With the rapid escalation of cybersecurity threats, the demand for robust and efficient encryption techniques to secure sensitive visual data has become increasingly critical. Conventional algorithms such as AES and DES, although widely used, exhibit limitations when applied to digital images due to their high computational overhead and inadequate adaptability to the intrinsic characteristics of visual content. This study introduces a novel image encryption scheme grounded in chaotic systems, incorporating an optimized selection mechanism for system parameters and initial conditions to enhance both security and computational performance. Leveraging the inherent properties of chaotic dynamics - such as extreme sensitivity to initial states and pseudo-random behavior - the proposed approach achieves high levels of confusion and diffusion. Experimental evaluations confirm that the encrypted images exhibit uniform histogram distributions, effectively concealing visual structures and thwarting statistical analysis. Entropy measurements ranged from 7.9993 to 7.9998, indicating an exceptionally high degree of randomness. Furthermore, correlation analysis revealed a substantial reduction in adjacent pixel correlation, with values between -0.0003 and -0.0025, signifying strong decorrelation and noise-like behavior in the encrypted outputs. The method also demonstrated strong resilience against differential attacks, achieving a 100% NPCR (Number of Pixels Change Rate), underscoring its sensitivity to minor input alterations. UACI (Unified Average Changing Intensity) values ranged from 33.48% to 33.50%, highlighting the algorithm’s effectiveness in uniformly diffusing changes throughout the image. In terms of efficiency, the proposed system outperforms traditional methods by offering reduced encryption and decryption times, rendering it highly suitable for contemporary digital environments where both security and performance are paramount.