Image Encryption technology based on Advanced Signal and Image processing Approaches

Abstract: The rapid increase in the digital image transmission in medicine, surveillance, and the Internet of Things (IoT) networks has made image security a vital issue in cryptography. Digital images are highly redundant, highly correlated between adjacent pixels, and in medical images, highly bit-depth, making standard techniques such as Advanced Encryption Standard (AES) and Rivest–Shamir–Adleman (RSA) wholly unsuitable. Despite significant Advances in chaotic encryption, optical encryption, DNA-inspired encryption, and quantum cryptography, there are some significant limitations: static Substitution Boxes (S-Boxes), poor sensitivity to plaintext, limited key space, lack of static randomness validation, and ignoring high-bit-depth images.
The proposed thesis systematically deals with these issues using the following five complementary frameworks. First, Enhanced Random Phase Encryption with Feature Iterative XOR Image Encryption (EDRPE-FIXIE) algorithm eliminates the linearity of the classical DRPE by employing Substitution Boxes (S-Boxes) from chaos systems and generating keys based on image features, thus making exponential key expansion in each round, with bits in just three rounds. Secondly, a new Lorenz, Secure Hash Algorithm 256 (SHA-256), and Discrete Time Quantum Walks (DTQWs) combination produces S-Boxes that are dependent on plaintext data, providing an entropy value of 7.9999 and passing all 15 tests of NIST. Third, a DNA-based encryption employing Generalized Arnold Cat Map Transform (GACMT) with Logistic-Sine-Tent (LST) chaos maintains 16-bit diagnostic information of medical images while achieving entropy close to the ideal value of 15.95 bits. Fourth, a revolutionary framework ensures NIST compliance not as an afterthought but as an iterative design consideration through adaptive parameters turning of the LST hybrid map. Fifth, a Hash-based Message Authentication Code (HMAC) based scheme uses Piece-Wise Linear Chaotic Map (PWLCM) with weighted image fusion steganography.
Extensive validations on standard test images indicate Number of Pixels change Ratio (NPCR) above 99.6 %, Unified Average Change Intensity (UACI) between [33.3-33.7] %, correlation values at the level of, and perfect decryption. The computational complexity is of the order, allowing for real-time applications. Performance benchmarks compared to more than 25 existing works have proven the superiority of our schemes in entropy, differential attacks, randomness, and key space. Future research directions include hardware implementations, color image encryption, and quantum walks-based photonic encryption
Keywords: Image encryption, DRPE, chaotic cryptography, quantum walks, DNA computing, SHA-256, NIST SP 800-22, medical image security, post-quantum cryptography.