Domain-flexible selective image encryption based on genetic operations and chaotic maps

• Published in: The Visual computer Vol. 39; no. 3; pp. 1057 - 1079
• Main Authors: Murali, P., Niranjana, G., Paul, Aditya Jyoti, Muthu, Joan S.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.03.2023; Springer Nature B.V
• Subjects: Artificial Intelligence; Chaos theory; Computer Graphics; Computer Science; Data encryption; Design; Domains; Encryption; Fitness; Genetic algorithms; Hierarchies; Image coding; Image databases; Image Processing and Computer Vision; Image retrieval; Literature reviews; Mutation; Neural networks; Original Article; Polynomials; Transformations (mathematics); Chaos; Selective encryption; Region of Interest (ROI); Edge detection; Orthogonal polynomials; Genetic operators
• ISSN: 0178-2789; 1432-2315
• DOI: 10.1007/s00371-021-02384-z

Image encryption research has seen massive advances in recent times, but many avenues of improvement still remain nascent. This paper takes head on various research challenges, giving the user fine grained control over their encryption requirements, by proposing a domain-flexible and selective image encryption scheme based on genetic algorithm, chaotic map, square-wave diffusion and orthogonal polynomials transformation. Initially, the proposed cryptosystem separates the image into important and unimportant regions making use of edges in the image with the orthogonal polynomials transformation. Important blocks, termed as Regions of Interest (ROI), are encrypted based on genetic operators and fitness score with chaos and unimportant blocks are encrypted with shuffling operations in the orthogonal polynomial domain. Then, square-wave diffusion is carried on the entire image to obtain the final encrypted image. The novel feature of the proposed encryption scheme is the unique design of the fitness function, wherein the fitness value can be varied between 1 for maximum speed and 10 for maximum security, to suit the user’s requirements and can operate in frequency or spatial or hybrid domain suitable for a vast range of real-time applications. Extensive experiments and analyses have been conducted to demonstrate the efficiency of the proposed work.