A CHAOS-BASED HYBRID IMAGE ENCRYPTION SCHEME USING DNA ENCODING AND DEEP NEURAL NETWORK KEY GENERATION
alsaadi, lamis alsaadi · Zenodo (CERN European Organization for Nuclear Research) · 2025
This repository contains the preprint of the research article entitled “A chaos-based hybrid image encryption scheme using DNA encoding and deep neural network key generation.” The study proposes a novel hybrid image encryption framework that integrates convolutional neural networks (CNNs) for plaintext-dependent dynamic key generation, a two-dimensional hyper-chaotic map for pixel permutation and diffusion, and dynamic DNA encoding for bit-level confusion. Unlike conventional chaos-based or DNA-based encryption schemes, the proposed method forms a fully coupled, adaptive architecture in which the plaintext image directly influences the cryptographic key stream and subsequent encryption stages. Extensive experimental evaluations on standard benchmark images demonstrate near-ideal cryptographic performance in terms of information entropy, differential attack resistance (NPCR and UACI), pixel correlation, key sensitivity, and robustness against noise and data loss.