| The content of this thesis is copyright protection technique for digital contents through the technique of digital fingerprinting based on fractional-order chaos dynamic system. Firstly, users get their initial values by the initial algorithm (you can choose the suitable algorithm according to your needs), since the fractional-order chaos is very sensitive to the initial values, so the uniqueness of fractional-order chaotic sequence for each user is guaranteed. Compared to digital video, digital images contain less data information, so the best choice of digital fingerprinting technique is simply based on fractional-order chaos. On the contrary, digital videos contain more information, although invisibility is required, the length of digital fingerprints can be extended, we need to encode the fractional-order chaotic sequence, referring to the fingerprinting algorithm based on the binary random coding methods existed. It can efficiently deter collusion attacks. Fractional-order chaos has stronger randomness than integer-order chaos, much more transformation space, and can be used by more users, so it solved the issue that copyright protection can’t be realized when the number of users is too large. Moreover, storing too much information about users is not necessary, only users’names are enough, data system space gets cut down in this situation.In addition, this thesis proposes a way dealing with the carrier images by using fractional-order WT, it is on the foundation of fractional-order Fourier Transform and Wavelet Transform. It combines their advantages, and can get much more space to encrypt, what’s more, the security is enhanced. For the carrier videos, a segmentation method based on directional empirical mode decomposition(DEMD) is used, independent component analysis (ICA) followed, then DCT is applied to dispose the independent component frames from ICA. The robustness and invisibility are improved, collusion attack is solved efficiently. |
PDF Full Text Request