Image Feature Extraction And Record Method Research Used For Copyright Protection

With the awareness of the copyright ownership heightened, the image copyright protection has become a hot area of research. This kind of problem is not only used for the theoretical research, but also an important part of the actual copyright protection applications. When the image’s copyright owner needs to prove who the copyright owner is, the watermark can be extracted as a proof to defend the copyright owner’s legitimate interests.Considering the problems in image retrieve and watermark system, this paper presents a method to solve the registration issues in the images’copyright protection. This article is mainly focus on photography images, animation images, design images, brand images. The method presents in this paper extracted the images’color, shape, texture features, and the DC coefficient as the basis of the retrieval. When there is a new image needs to register its copyright, the image retrieve is needed. If there is no image has the similar content with the new image, then the new image’s features will be written to the storage device to be filed. To protect the images’copyright, the watermark service is needed. For most of the images contain lots of edge contour, the ridgelet transform and the DCT are used to embed the watermark. The core of the ridgelet transform is first performing the image with the randon transform and then the wavelet transform. The randon transform can convert the edge contour the discrete points which the wavelets transform is good at dealing with. The proceedings above can enhance the efficiency and robustness of the watermark system.The innovation of this paper is that the color, shape, texture feature is used in the image retrieval which are commonly used in pattern recognition, the DC coefficients is also added as an auxiliary. This paper presents the methods of combined ridgelet transform and the DCT transform to embed the watermark, which improves the robustness of the watermarking system.