As digital products communicate over public networks, digital image security becomes essential. Despite advancements in encryption technology, protecting data from secure images remains a complex computational issue that necessitates using an encrypted environment to protect data transmitted from devices over networks. Encryption is critical for protecting sensitive information, especially images, from unauthorized access or failure. Image encryption is distinct from text encryption. Images contain extensive data with high redundancy and a high correlation between nearby pixels, making them difficult to process with traditional technology. In recent decades, chaos maps have become popular in the crypto community. This work proposes a new encryption technology based on chaotic map substitution boxes (S-box) and cellular automata (CA) to address the frequent challenges in chaotic encryption methods. To address the inadequate randomness provided by the 1D chaotic map, this work presented a 4D memristive hyperchaos with a more excellent chaotic range, increased uncertainty, and ergodicity as an alternative to the software-based approach, which is vulnerable and offers low throughput. The suggested encryption technique was implemented on an Intel Cyclone IV EP4CE115F29C7 FPGA. It is less prone to tampering and has a better throughput. This suggested encryption scheme employs 1384 LEs. In comparison to traditional algorithms, the suggested Number of Pixels Change Rate (NPCR) (99.9706%), unified averaged changed intensity (UACI) (33.3956%), and information entropy (IE) (7.984) accomplish more effective. Statistical studies such as IE, histogram, correlation, peak signal-to-noise ratio (PSNR), differential analysis such as NPCR and UACI, and noise attack analyses are employed to validate the suggested encryption design. Regarding security, the suggested hybrid method outperforms conventional algorithms while protecting image quality.