Research On Robust Digital Watermarking Algorithm Based On SVD And Neural Network

The fast development of Internet makes it convenient to access the multimedia data, butat the same time the Internet has brought a new challenge to the copyright protection. As aneffective means for the protection of digital property rights, the technique of digitalwatermarking is an important part of the region of information hiding and attracts wideattention. Digital watermarking is a technique that embeds copyright information, such asfigures, sequence numbers, words, image logos and so on, imperceptibly into multimedia data.It can be used for tracing the distribution and usage of multimedia data. The specialapplication of digital watermarking technique requires that the embedded watermark shouldbe not only transparent to human observers, but also robust enough so that it cannot be easilydestroyed or removed after some digital image processing or attacks.The basic concept and theory of digital watermarking is first described in this paper.According to the different ways of information hiding, the existing algorithms of digitalimage watermarking are concluded and compared. Based on the shortcoming of the existingalgorithms, we propose a novel watermarking algorithm using the technique of singular valuedecomposition and the neural network. In the embedding scheme, the host image is firstdivided into un-overlapped 8×8 sub-blocks in spatial domain and performed singular valuedecomposition on each sub-block. Making use of the characteristics of singular value andvariance, we preprocess the host image and get the activity factor of each sub-block. Byselecting the suitable sub-blocks according to their value of activity factors to embedwatermarks, we can ensure the controllability of watermark embedding strength and theimperceptibility of the watermarking scheme. The use of singular value decomposition makesour scheme more robust. We simulate the algorithm in the computer and test the performanceof watermark embedding, watermark extraction and the detection, extraction and recovery ofthe embedded watermark after different kinds of attacks. Experimental results show that thealgorithm is feasible and robust.The research result of the proposed algorithm in this paper not only takes a further step tothe technique of singular value decomposition, but also remarkably enhances the ability ofresisting against attacks of the watermark. And what's more, the watermark recovery basedon the neural network fulfills the precise tracing of the multimedia data's usage anddistribution.At the end of this paper we have a prospect of the technique of digital watermarking. Theextensive application in the future of the watermarking technique also attracts morewatermarking algorithms.