Image Hiding Technology Based On The CNN Hyperchaos And Shearlet

Nowadays, the network communication has become the main channel for people to exchange information. The Internet has brought us great convenience for information transmission, but the openness of the Internet has also posed new challenges to the information security. However, for the information as the personal medical record, the company’s key technology, the state secret and so on, it can easily be illegal destroying and stealing to transport in public directly, the consequences are inestimable. Therefore, as an important means of information security, the information hiding has become the focus of study.The image hiding has become an important part of the information hiding, as the image plays an increasingly important role in the network communication. The image hiding is usually first to encrypt the secret image, and then to hide. Image encryption algorithms are mainly based on matrix transformation, secret sharing and chaos, the safety of these algorithms is not high and some of them are cyclical. The algorithm of image hiding mostly focused on the transform domain because the robustness of transform domain is good.However, there are few algorithms to study the contradiction between invisibility, robustness and time complexity of the algorithm.In order to overcome the shortcomings of traditional chaos sequences who have simple keys and low security, the hyper chaos sequence of cellular neural network(CNN), bit plane decomposition and Arnold encryption are used for image encryption pretreatment in this article. Shearlet is used in the hiding algorithm, which is a type of transform method. The carrier image is firstly decomposed by the fourth-scale shearlet transformation, and then the low-frequency sub-band which from the shearlet transformation is decomposed by the wavelet packets. The high bit-planes of secret image are hidden into the singular values of the wavelet packets low-frequency sub-band, and the low bit-planes are hidden into the high-frequency sub-band of the shearlet. However, the second-scale shearlet transformation is used to extract the secret image because of the energy aggregation of shearlet combination.Experimental results show that, the algorithm balance the relationship between invisibility,robustness and time complexity of the algorithm in a certain extent.