An Extensive Analysis of Deep Learning Techniques for Improving Picture Encryption's Security & Data Embedding Capability
Yandapalli Venkata Sree Vaishnava Reddy · Communications on Applied Nonlinear Analysis · 2024
Because data digitisation is so common, it is necessary to encrypt messages while they travel over unprotected channels. The communication is encrypted before it is sent. Encryption cannot function without both a technique and a key. It is crucial to take additional measures to safeguard the privacy and confidentiality of the receivers when transmitting sensitive images over an unsecured network such as the internet. Regardless of the fact that there is no standardisation for these algorithms Combining standard public key cryptosystems like RSA and El-gamal with chaos-based cryptosystems like Chebyshev polynomials can improve security in a manner similar to AES, DES, RSA, etc. Although AES and other classical algorithms have been standard for some time, many security experts now recommend chaos-based encryption approaches for media files due to their computational efficiency. Because existing algorithms are unsuited for real-time picture encryption due to their poor encryption speeds and the high processing needs of secure encryption methods, researchers have proposed bespoke algorithms for image encryption. One of the better techniques is chaotic encryption, which uses cypher text that is unpredictable enough to greatly decrease the likelihood of decipherment when applied to images. Consequently, research into digital picture encryption algorithms based on chaotic technology has become an important tool for modern digital image encryption. We go over all the new features and improvements in picture encryption. With an eye towards evaluating them and giving academics and practitioners the context they need to understand the methods' current position, this study aims to introduce and summarise the currently used photo encryption algorithms and metrics. This study rates the different photo encryption algorithms based on their strengths and weaknesses after examining them. Furthermore, the assessment matrices used to evaluate the performance and security of the encryption algorithms in recent studies have been validated by comparative analysis.The article also reviews these metrics in detail and provides upper and lower bounds for a set of efficiency, quality, and security evaluation criteria for image encryption algorithms.