Rapid digital advances have made image security crucial, especially in fields like banking, biometrics, and social media. This study proposes a novel image cryptosystem combining DNA and chaos-based cryptography to ensure secure image transmission. A Cross Cosine Map (CCM) generates chaotic keys for grayscale image encryption, while a hyper-chaos-based system secures color images. A lightweight architecture offers dual-layered protection for biometric data. The system's performance is validated using statistical, differential, and robustness analyses, along with standard image quality metrics.