<p>Image encryption technology based on chaotic systems is an important guarantee for the secure transmission of information. In this paper, a novel cellular automata controlled coupled map lattices (CACCML) is proposed, in which the coupling structure is dynamically controlled by cellular automata to enhance structural complexity, and two enhanced one-dimensional maps are presented to improve the dynamics performance of the system. The dynamics analysis reveals that all lattices of the system exhibit chaotic behavior within the given parameter range, demonstrating its strong potential for cryptographic applications. Based on the proposed CACCML system and two-dimensional cellular automata, an novel image encryption scheme is presented. First, the plain image information and the initial security keys are jointly computed to generate chaotic initial values, ensuring strong sensitivity to plaintext image. Next, a dual-state transformation permutation method is introduced, which leverages the iterative evolution of the two-dimensional Von Neumann neighborhood structure cellular automata (2D-VNCA) to indirectly scramble pixel positions. Subsequently, an iterative HL-Bitplane diffusion method is proposed, cleverly integrating the separated high-four-bit and low-four-bit planes into the iterative process of the two-dimensional reversible-like cellular automata (2D-RLCA), thereby achieving an efficient diffusion operation. The results from comprehensive security analyses and comparative evaluations demonstrate that our proposed image encryption scheme possesses exceptional security strength and robustness.</p>