An adaptive, SVM -based watermarking in frequency domain

In digital management, multimedia content and data can easily be used in an illegal way - being copied, modified and distributed again. Copyright protection, intellectual and material rights protection for authors, owners, buyers, distributors and the authenticity of content are crucial factors in solving an urgent and real problem. In this paper, we describe an algorithm of watermark-embedding in frequency domain based on the support vector machine (SVC). Because of the SVC has superiority in classified-area, so we use it to simulate the picturespsila texture-characteristic and separate the picturepsilas sub-blocks into two classes, in this way, we choose the embedding intensity adaptively. Firstly, we separate the picture into many sub-blocks (8*8) , then we use their texture characteristics as the training sample to train the SVC, the training goal is -1 and 1 (represents two kinds of different picture sub-blocks); Secondly, we use the training result to establish a SVC classification model, then we can use this model to classify the new picturepsilas blocks and decided the embedding intensity adaptively. The simulation result indicates that the watermarked-pictures with this method have a good visual sensation and the watermark is robust to many common attacks.