摘要
针对文档图像的特点提出了一种基于文档图像的盲检测水印算法。该算法能抵抗打印—扫描的攻击。首先对文档图像进行字符分割,使每块只包含一个字符;然后对每块进行线性规范化规范块大小,对规范的块进行三级离散小波变换(DWT),通过调整细节子带系数的正负号的数量来嵌入水印信息。实验结果表明,该算法适用于文档图像,能抵抗打印—扫描攻击。
This paper presented a blind detection watermarking algorithm for document image to hide information against printscan process. Considering the characteristic of the document image, firstly, cut the original document image into several blocks, each of which included only one word. Then decomposed each image by discrete wavelet transform(DWT) for three degree and embedded watermarking by modulate the count of the positive and the negative of each detail subband. Experiments show this algorithm suit for document image and it can survive the print-scan process.
出处
《计算机应用研究》
CSCD
北大核心
2007年第12期199-200,265,共3页
Application Research of Computers
基金
国家自然科学基金资助项目(60573045)
关键词
数字水印
文档图像
打印一扫描
digital watermarking
document image
printing and scanning