ARCHIVES

In the past, cryptographic key derivation has been achieved by using a password, a random number generator, or a hardware-based entropy source to generate secure encryption keys. The authors introduce a novel framework, called Paradox: Recursive Visual Entropy Key Derivation Engine (RVE-KDE), which presents an experimental solution to the problem of exploring digital images as an alternative deterministic source of cryptographic entropy. The proposed method is based on recursive image traversal, pixel-level visual entropy extraction and state evolution using hash-chains to generate reproducible cryptographic keys from image data. Unlike conventional processing methods, which process linear inputs, Paradox recursively examines multiple areas of a picture to gather color, spatial and structural data and incorporates it into an entropy pool. This entropy is then passed through a key derivation step to create deterministic symmetric keys that can be used with existing and proven encryption methods like AES-256-GCM and ChaCha20-Poly1305. To test the proposed framework, a Python implementation was created and experimentally evaluated using avalanche effect analysis, Hamming distance distribution, collision testing, latency measurements, and a comparison against several other key derivation functions (KDFs) such as PBKDF2, HKDF, Argon2id, and BLAKE3. Experimental results show deterministic repeatability and statistical properties matching the assessed cryptographic parameters. Paradox is not designed to be the replacement for standard key derivation function, but rather a framework for research in visual entropy as a new input space and a reproducible base for further studies in image based cryptographic key derivation