[[alternative]]a content base watermarking model

[[abstract]]ABSTRACT A Content Based Oblivious Watermarking Model By Fu-Hao Ya Computer network and digitized information, which bring the convenience of spreading and modifying, make it much easier to acquire multimedia data. Therefore, it becomes more and more important to affirm the legal use of multimedia data to protect the intellectual property of the creators. And Digital Watermarking, developed recently, has been proposed as an efficient way to protect copyright. The principal purpose of this research lies in the development of Oblivious watermarking, which can ensure the existence of watermarking independently without the assistance of original images, and thus can authenticate the authority of copyright. The main method we adopted is to divide the whole image into several equivalent sub-blocks, and to determine if the embedment of the watermark would cause the distortion of the image by analyzing the texture and brightness of each sub-block. Then, each sub-block is given its proper weight, by which we embed the watermark of different energy in it, and, lastly, we amend the edge of the whole image after the embedment of the watermark with a function to retrieve the distortion. The result of this experiment reveals that the method we took, as mentioned above, is able to effectively defend the attacks which other image processing techniques would cause, such as image cropping (95% cropped), lossy compression (rate of compression=12), image enhancement, multiple watermarks, etc.