Research Of Image Encryption Algorithm Performance Evalution

Since the encryption technology based on d blind source separation has been proposed, the performance evaluation of these blind theories encryption have become the focus to scholars, the aim of this paper is evaluating the two typical encryption technology using signal to noise ratio, and then singular value decomposition method, aiming at the evaluation of only the image itself, would be proposed in order to expound performance evaluation and comparison of the different blind encryption technologies.This paper describes t he encryption method based on underdetermined blind source separation and blind encryption based on BSS/NMF, and gives detailed argument of ICA and NMF principle which are the theoretical basis of blind encryption. And the MATLAB simulation images of two blind encryption algorithms are given, including explicit images key images, encryption images and decryption images, meanwhile the effect of encryption and decryption from the sensory point of view is given. The focus of the paper is the image encryption evaluation using signal to noise ratio and singular value decomposition. A large number of experimental simulations using Lena, Cameraman and other pictures and image contrast are provided for the detail of two blind encryption algorithm evaluations. The results show that BSS/NMF based blind encryption has, under the same correlation, low signal to noise ratio after encryption, high decryption signal to noise ratio value, and low entropy value of singular, indicating that the BSS/NMF based Blind encryption technology is better than the encryption method which based on underdetermined blind source separation.