Improvement And Research Of Digital Watermarking Algorithm Based On HVS And Ridgelet Transform

As the rapid development of network has accelerated the coming of digital era, the exchange and dissemination of digital media information has been more and quicker. But it also exposed many problems. Through the Internet, a variety of multimedia information can be easily got in the computer, mobile phone, tablet and other electronic products.More and more individuals or organizations were unauthorized and illegally copied the works with copyright information, causing tampering, infringing the copyright of digital works and illegal ling copying. Copyright protection of digital media information has become the focus of attention. Researchers have developed many mature digital watermarking algorithms in DWT domain, to solve the discontinuous or separated edge points in the image. These digital watermarking algorithms had good time-frequency characteristics. But most of the wavelet transform s can only obtain single direction information of images, and still exist great shortage in the processing of linear edge contour.In order to seek the best representation of image linear edge contour, based on in-depth research on the basic theory of ridgelet transform, characteristics of human visual system(HVS) and the odd even quantization algorithm, an improved algorithm was proposed with the characteristics of human visual system and the ridgelet transform. First, the images were uniform segmentation, and then the edge of each segmented sub block is approximately a straight line. Second, perform the ridgelet transform on the sub block. Third, determine the quantization step size using odd even quantization algorithm and human visual masking properties and the watermark was embedded into the coefficients of an important component visually. The algorithm not only solved the problem of sparse image, but also effectively expressed the image contour edge features and highlighted the advantage of the ridgelet transform compared with wavelet transform. Three typical gray images were selected as the carrier images in simulation experiments and attack test. In order to further verify the performance of the proposed algorithm, the Comparative experiments were performed between the proposed algorithm and wavelet transform. The experimental results show that, the proposed algorithm has good transparency and strong robustness to resist the common attacks such as noise, filtering, sharpening. The proposed algorithm has general performance against JPEG compression attack.