摘要： 数字水印是一种在其它媒介上通过嵌入一些重要标识信息以保护媒介的隐私、版权等的安全技术。该文基于离散余弦变换的原理，结合JPEG标准的量化特点设计水印算法，利用图像本身特征和人的视觉特性确定水印的嵌入位置。根据嵌入算法原理，水印的检测只需要原始图像一小部分的信息，在含水印图像遭受不同程度的压缩、剪切、模糊加噪等攻击后仍能有效地检测到水印的存在。

关键词: 可见阈值, Arnold, 强度因子, DCT

Abstract: Digital watermark is an important security technique that is used widely to protect some key information such as privacy, authority through embedding watermark into the mediator. This paper designs a watermark algorithm according to the principle of the DCT transformation as well as combination with JPEG standard. While making sure of the watermark’s embedding position, this paper takes good use of the sense of visual feature of person and the characteristic of image. The examination of watermark only needs a very small part of data from the original image according to the principle of the algorithm. It can still availably examine the watermark’s esse when the image that contains watermark suffers the compression of the different degree, shearing to slice, and so on.

Key words: Threshold value of visual, Arnold, Strength factor, DCT