数字水印技术是国际上最新的一种版权保护技术。它将数字、序列号、文字、图像标志等版权信息嵌入到多媒体数据中，以起到版权跟踪及版权保护的作用。数字水印有多种不同的实现算法，主要可以分为基于时（空）域的算法和基于频域的算法。基于时（空）域的算法直接在图像所在的二维空间进行处理，比较直观。基于频域的算法是先把图像变换到频域，对图像的变换系数进行处理后，再反变换回空间域的方法，能解决一些直观上不易理解的问题。为了将信息隐藏得更深，可以在嵌入前对水印信息进行Arnold变换置乱等预处理。本文将实现一种基于DCT域和Arnold变换的数字水印嵌入和提取算法，并提出改进方案。

Digital Watermarking is a newly arisen copyright protection technology. It can embed the copyright information such as: serial numbers, words, image and symbol into the digital media data to trace the copyright and protect the copyright. There are many algorithms to realize the digital watermarking, and mainly differed into two kinds: one based on spatial domain, and the other based on frequency domain. The algorithm based on spatial domain manipulate directly on pixels. The algorithm based on frequency domain which firstly transforms the image to frequency domain, manipulates with the transformed coefficient, and then transforms the coefficient inversely to space domain. It can solve some invisible problem. To make the information hidden deeply, we can pre-process the watermark by Arnold transform. In this article we will realize a digital watermarking embedding and extracting algorithm based on DCT domain and Arnold transform and then give the improving plan.