Implementing digital image security framework with hybrid approach of chaotic map and singular-value decomposition

Abstract

Various encryption techniques, mostly based on mathematical and logical principles, are used for protecting sensitive data from attacks meant to modify or unauthorizedly distribute them. The importance of these techniques has grown significantly as various real-life applications in fields like medicine, banking, or transport are accompanied by increasing security concerns. Various effective cryptography schemes were proposed so far in the literature; however, each of them exhibits certain flaws and limitations with respect to different vital aspects. To overcome these issues, we propose a hybrid scheme for securing digital images by encrypting them using a dual security approach. More precisely, we use first a chaotic map for scrambling the image pixels, and then, we apply the singular-value decomposition method for decomposing the permuted image to provide very strong security. Individually each of these steps has already been considered in the cryptographic literature; however, their combination has not been proposed before this contribution. Experimental results on benchmark data validate our proposed scheme and various performance evaluation metrics indicate that it shows promising qualities in terms of security (against various attacks) and sensitivity in comparison with baseline methods.