Efficient selective image encryption based on chaos and Tchebichef moments.

In this work, we introduce a novel selective encryption scheme based on Tchebichef moments and chaos theory. The goal is to propose a fast and secure encryption algorithm that can deployed in real world scenarios. The proposed algorithm operates in the domain of Tchebichef moments, we encrypt only the most significant coefficients of the moments transform. First, we down-sample the computed moments' matrix coefficients, then we use two logistic maps for confusion and diffusion of the down-sampled Tchebichef's coefficients. The resulting matrix is the encrypted image. This approach improves drastically the time performance of the encryption algorithm while keeping a "good" security level. In order to prove the performance of our algorithm, we run different experiments and test the algorithm on different criteria: MSE, correlation coefficient, differential analysis, entropy and time performance. The presented results prove that the encryption algorithm proposed is secure outperforms state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]