Medical Images Encryption Using Two Improved Vigenère Laps Separated by Genetic Crossover

Abstract

Mariem Jarjar, Abid Abdellah, Hicham Rrghout, Mourad Kattass, Abdellatif JarJar* and Abdellhamid Benazzi

In this article, we present a novel method for encrypting medical images, based on a significant improvement of the classical Vigenère technique. This method utilizes dynamically generated large-sized substitution tables derived from multiple pseudorandom vectors. After vectorizing the original image and applying confusion with a pseudo-random vector, the first Vigenère round encryption is performed using two substitution tables, followed by a genetic crossover adapted for image encryption and controlled by a pseudo-random table. Next, a second Vigenère round encryption is applied, implementing new confusion and diffusion functions to maximize the avalanche effect and protect our system against all differential attacks. The output vector is then fully subjected to permutation to increase attack complexity. Simulations are conducted on a large number of images of various sizes and formats to ensure that our approach is not vulnerable to known attacks.

PDF