Robust Steganography Technique that can withstand data tampering or degradation
Sellami Benaissi , أكرم عابد, مصطفى طهراوي
كلية العلوم الدقيقة-جامعة اكلى محند أولحاج - البويرة · الجزائر
In the digital age, the exponential growth of multimedia communication has made information security a critical concern. Steganography, the art of hiding secret messages within innocent-looking cover media, has emerged as a powerful tool for covert communication. However, existing steganographic methods remain highly vulnerable to unintentional distortions and deliberate attacks, raising a fundamental question: how can hidden data survive real-world degradations while remaining imperceptible? This thesis proposes DASRS (Damage Aware Self-Recovering Steganography), a robust steganographic framework that prioritizes payload survivability over mere concealment. The message is fragmented into six data chunks plus one XOR parity chunk, with each fragment protected by Reed– Solomon error correction and embedded simultaneously across four independent transform domains distributed across the YCbCr color space. A damage-aware cascade extraction with triple-validation and automatic XOR reconstruction enables partial message recovery even under severe compound distortions. Experimental evaluation against signal processing, geometric, and compound attacks demonstrates that DASRS achieves substantially higher robustness than classical single-domain baselines while maintaining acceptable imperceptibility. The framework offers a principled, modular alternative to existing approaches through structural redundancy rather than attack-specific optimization.