The Research Of Holographic Watermark Image Encryption Technology

In recent years, with the rapid development of Internet technology and multimedia technology, especially the popularization of WIFI in recent years, People can access to the Internet whenever and wherever possible, therefore, People can get all kinds of digital products including images, video etc, However, Bringing convenience to people’s lives at the same time, some speculators also use the characteristics of easy access, easy replication and tampering of digital products to obtain improper benefits through the illegal transactions, which greatly damage the interests of the copyright owner, therefore, seeking a kind of effective method to protect the legitimate interests of the copyright owner becomes more and more urgent. Digital watermarking technology, as an effective intellectual property protection scheme, is year by year being valued by scholars at home and abroad.Compared with the conventional digital watermarking schemes, we can find that the hologram itself has the feature of encryption since its generating process is quite complicated, in addition, if a part of the hologram and even a majority of the hologram is removed, the remaining part still can restore the original information better, this feature makes it very effective to resist cropping attack. However, compared with the normal image, Since the hologram contains a lot of information, if it is embedded directly into a carrier image as a watermark, it will be not conducive to hide the watermark information, so the image carrier watermarked image quality decline, the quality of carrier image containing watermark will decrease, therefore, reducing the redundant data through hologram compression becomes more and more necessary,therefore, the paper adjust dynamically learning factor and inertia weight for the particle swarm optimization(PSO)and use it to optimize the back Propagation(BP) neural network and establish hologram compression model based on improved particle swarm optimization BP neural network, Comparing with the BP neural network compression model and particle swarm optimizing- back Propagation(PSO-BP) compression model, The comparison results show that this algorithm not only achieves better compression hologram and significantly shortens the network training time, but also significantly reduces the number of network training and significantly improves the reconstructed hologram quality.After achieving the better compression of hologram, next, it is necessary to find an effective watermark embedding and extraction algorithms, In this paper, we conduct a comprehensive analysis on discrete wavelet transform(DWT) and singular value decomposition(SVD), On the basis of the analysis, the two algorithms are combined and improved effectively, firstly, the original carrier image is transformed by DWT, then the low frequency sub image we have got through DWT is conducted by 44 ? block, next, we extract the largest singular value of each sub block SVD decomposition, then, the singular value of holographic watermark through SVD decomposition is embedded into the largest singular value of each block. By comparing with the literature [53] algorithm, we can find that this algorithm not only guarantees invisibility of the watermark, but also can obtain better robustness, it has very strong resistance to noise, filtering, cropping, JEPG compression, rotation and translation attacks.