Despite current developments in the discipline of selective image encryption, some of the most pressing challenges still need to be solved concerning the degree of user-controlled encryption level. These research gaps have motivated the proposal of an encryption algorithm that gives the user control over the encryption level. It is based on JSMP map, the creation of random artificial image, OPT, and dynamic permutation matrix using lower upper (LU) decomposition. The secret key generated with the novel JSMP map makes it highly secured as it exhibits chaos in the enormous range of. The security analysis is conducted for different user choices u∈10,100, where u refers to the incremental degree of encryption from selective to full image encryption. Results show that the proposed technique achieves high efficiency when and degrades towards u=10, whereas the speed increases for to as it ranges from 10.97 MB/s to 23.27 MB/s, respectively. Therefore, the proposed cryptosystem can be chosen for various applications requiring different security levels.
TopIntroduction
Rapid communication and Internet technology growth has facilitated the transmission of multimedia like text, image, video and audio over public networks, social media platforms and cloud storage. This transmission has several advantages, such as high speed, no temporal limit, cost-effectiveness and independence from geographical distance between the sender and receiver. At the same time, vital information must be protected against illegal use by using efficient encryption. Encryption of text data has taken a good shape with the introduction of traditional methods such as DES, AES, RSA and IDEA as discussed in (Schneier 1993; Dhanalaxmi et al. 2017). However, image and multimedia data encryption need special schemes and rules as they have intrinsic features, including bulk data, strong correlation between pixel values, and high pixel redundancy. Encryption of these intrinsic features lowers the performance of encryption schemes. Hence, traditional encryption methods are unsuitable for images Stallings (2019), and researchers started paying attention to encrypting images and multimedia content using various chaos-based techniques, one of the most pivotal of which is selective image encryption [4].
Chaos-based encryption (S. Muthu, 2022; Iqbal N, 2021; Roy M, 2021) became popular as it increased the uncertainty of an encryption process by creating random sequences, besides its known characteristics like sensitivity to initial conditions, transitivity and ergodicity. Images contain various regions, some of which are important and some not, and this information can be exploited to create selective or partial encryption schemes, wherein the critical regions of the image are more encrypted, considerably reducing the computational cost. In the work proposed by Murali et al. (2019), Saw-tooth SFC was used for the confusion process, and diffusion was performed with SVD and chaotic maps. Murali et al. (2022) presented an image encryption scheme based on genetic computing, chaos, OPT and square-wave diffusion, which included a novel fitness parameter that gives the user control over the security speed tradeoff.
Despite massive progress in selective encryption, some of the most pressing challenges remained unsolved, like creating an efficient windowing methodology for ROI-based pixel selection or giving the user or deployment endpoint dynamic control over the encryption level. For these, the algorithm must smartly vary the number of iterations per block, have an efficient essential structure and thus be suitable for various devices from high-powered workstations to handheld mobiles.
To meet all these requirements, a new selective image encryption technique based on chaos, the creation of random artificial image, orthogonal polynomials transformation and dynamic permutation matrix using Lower Upper (LU) Decomposition has been proposed in this research paper.
The primary innovation of the proposed research is the construction of a random artificial image using the JSMP chaotic map. Additionally, we aim to generate a permutation matrix through LU decomposition, which will be utilized for shuffling purposes.
This research work introduces a groundbreaking approach to creating random artificial images by leveraging the JSMP chaotic map. By harnessing the power of this map, we can generate visually captivating and unpredictable images that push the boundaries of traditional image construction techniques.