Error-correcting Code-based Digital Image Watermarking Algorithm

As an important technology in copyright protection, digital watermark has been attracted great attentions in the world. Now, there are many significant results in digital watermark research and the application of watermark technology. However, as a new area of research, there are still a lot of digital watermarking technology theories and technical problems to be solved.This paper studies an improved secure algorithm for the non-linear digital image watermark based on error correcting coding and non-linear detector. First, we modulate the watermark messages into binary pulse amplitude modulated codes, and add the same modulated codes into different sections of DCT coefficients in medium frequency domain, which have been disarranged by Arnold transform according to the features of digital image. Then, we choose a non-linear saturating detector to detect the watermark messages. This kind of non-linear detector possesses better robustness and frequency adaptability under various attacks. Finally, we use error correcting coding to improve secure algorithm. The experiment results show that the robustness and the anti-attack capability of the proposed watermarking scheme are remarkable in various attacks, such as additive noise, filtering, cropping, and compression. This study extends the application of error correcting coding and nonlinear watermark system.