The Research And Implementation Of Image Watermarking Technique For Print-and-Scan Process

Digital watermarking is an important branch of information hiding technology. It put a special digital identification into the digital media. It can be used for assessing and identify the copyright of the digital media. Generally, the watermarked information is invisible and can be detected and extracted by special algorithm. It may be the mark of copyright, the information of the product or the series no. of the user. The problem of printing papers security is becoming more important in modern society. In this dissertation, the mainly studies is that digital watermark technology application to print papers security field, a kind of image watermarking algorithm which can resist to Print-Scan process is designed and implemented.Based on the literature review, this dissertation analyzes the pixel distortion and geometric distortion of digital image through the print-and-scan process. For pixel distortion, it can be adjusted by contrast stretching transformation and luminance transformation; for geometric distortion, it can be adjusted by Radon transformation and cubic interpolation. The experiment results that these methods can improve the accuracy rate of the watermark detection.Based on the analysis of DCT coefficients of original image and after print-and-scan. Find the area where DCT coefficients change less, and used it to be the location to embed the watermarking. Combined with the image adjustment methods, a kind of image watermarking algorithm based on DCT which can resist to print-and-sean process is designed. In this algorithm, using the adaptive method when watermarking embedding process, it solved the contradictions of robustness and imperceptible. The experiment results showed that this algorithm can better resist to the print-and-scan attacks and also have a certain degree of robustness to other attacks.