数字文本作为首要的信息载体，比视频、音频格式更多的被人们使用。如同纸质书籍被盗版，电子载体也存在版权保护和版权认证问题。 本文分别提出适用于Word和Adobe PDF文档的数字水印算法： 提出用于Word文档的Image-Plus-TrueType水印算法。嵌入算法为，读取水印图片数据得到“0-1”矩阵M，用TrueType新造字符“0”、“1”；将矩阵数据对应的新字符逐一嵌入到原始文档中。通过Office Word软件的“隐藏”功能和TrueType字库的显示特性实现水印嵌入。提取算法是嵌入算法的逆过程。 另提出用于Adobe PDF文档的Image-Plus-Acrobat水印算法。嵌入算法中，同上获取水印图像对应的矩阵M，再将矩阵M的元素嵌入到PDF文档中，利用Acrobat软件的TouchUp工具的“填充”和“字符间距”功能，对嵌入字符进行处理，实现隐藏。提取算法是嵌入算法的逆过程。 验证了两种算法的可行性，并在Visual Studio 2008平台上，以C++语言实现了用于Office Word 2003的水印算法。 经仿真证明，Image-Plus-TrueType和Image-Plus-Acrobat水印算法有良好的鲁棒性、透明性和安全性，可定位被篡改位置，兼容多种语言，可对Word和PDF文档进行版权保护和版权认证。然而，Image-Plus-TrueType算法受到水印图片大小的影响明显，Image-Plus-Acrobat算法的鲁棒性，可操作性仍需改进。

As the primary information carrier, digital text is more used by human than video and audio. Like piracy for printed books, there are copyright protection and copyright authentication problems for electronic carriers. This thesis proposes watermarking algorithms respectively apply to Word and Adobe PDF documents: The Image-Plus-TrueType algorithm applies to Word document. In the embed algorithm, reading the image into a “0-1” matrix and creating new characters “0” and “1” by TrueType. Embed the two new characters into the original document in the matrix order. Hide the new characters by the “hidden” function of Office Word and the display properties of TrueType font library. The extract algorithm is the reverse of embed algorithm. The Image-Plus-Acrobat algorithm apply to PDF document was also proposed. In the embed algorithm, embed the matrix into a PDF document like the primary algorithm. Then use the “fill” and “character spacing” functions of Touchup tool of Acrobat to hide the new characters. The extract algorithm is the reverse of embed algorithm. The feasibility of the two algorithm was verified, and the algorithm apply to Office Word 2003 was implemented by C++ language and Visual Studio 2008 platform. The simulation results show that both the Image-Plus-TrueType and the Image-Plus-Acrobat algorithm have great robustness, transparency and security; they can locate the tampered location and compact to variety of language; they can effectively protect the copyright of the textual documents and authenticate the copyright. However, the Image-Plus-TrueType algorithm was limited by the watermark image size; the Image-Plus-Acrobat algorithm’s robustness and feasibility need improvement.