Research On Digital Image Encryption And Information Hiding

With the security problem of image information as the background, this thesis isconcerned with digital image encryption and information hiding. Contributions are asfollows:Firstly, the image scrambling methods which use fixed form matrices are improved.Under the control of secret key, we present a new method for scrambling and restoringimages by means of a class of stochastic invertible matrices. This method can reachpreferable effects of encryption and decryption and it is secure.Secondly, base on AES, a new method for scrambling and restoring images isproposed, which is safe as well as simple. The scrambling effect is preferable, andit is analyzed by comparison of histograms.Thirdly, combined with image optimal block processing, we propose newmethods to evaluate image scrambling degree based on block cross-entropy andthe signal noise ratio SNR. The two methods are both effective to assess thescrambling effect, which tally with person's vision basically. Then, somecomparisons of image scrambling methods in color domain and position domain are given.The scrambling effect, computation complexity, keys and security about the two methodsare discussed.Fourthly, visual cryptography scheme with trusted authority((k, n)-TAVCS, forshort) is proposed. It retains the merit of (k, n)-VCS. This new scheme guaranteessecurity and satisfies the actual need.Fifthly, we apply the system expression of integer to image hiding. With this method,a gray image can be hided in a color image of the same size. In our vision, the quality ofthe color image is almost the same. The secret image can be restored completely.