A Survey Of Digital Watermarking Detection Technology Based On SVM

As an academic focus in recent years, digital watermarking technology has been applied in many fields, such as verifying the integrity and authenticity of multi-media data, secret communicating, solving disputations on copyright, designing user's marks and protecting the intellectual property rights of digital products, etc. As an important branch in digital watermarking system, digital watermarking detection technology involves detecting the existence of watermarking or further illustrating the significant and visible watermarking information.The algorithm of watermarking detection comes into nonblind watermarking detection and blind watermarking detection. The available methods for watermarking detection include related detection, maximum plausible detection and maximum posterior probability detection, etc. These methods are expected to be designed according to watermark embedding algorithm or massive samples, which sets limitations on its applications.In this paper, as an academic theory directing at minor samples, SVM (Support Vector Machine) is introduced. SVM, proposed by Vapnik and others, is a machine algorithm based on statistical theories. It is free from the disadvantages in former detecting methods that depend on the special embedding algorithm. The paper concerns a study on the applied aspects of SVM in watermarking detection and a SVM-oriented detecting platform has been set up with the discussion about its models and functions. The test shows that watermarking detection with SVM greatly promotes watermarking detection efficiency.The digital watermarking detection technology based on SVM offers a new method for the watermarking detection of digital image, which also makes beneficial attempt for image blind detections.