摘要
提出了一种基于小波域和独立分量分析的盲分离数字水印算法。该算法将水印置乱并乘以相应的系数,将此信息嵌入到经过二级小波变换的原始图像的逼近子图中,提取时利用快速独立分量分析实现了原始图像和水印的盲分离。经过大量实验证明,重构后得到的水印图像具有良好的不可见性,对叠加了水印的图像经过多种干扰及信号处理后,能够有效地检测出水印的存在。
Image matching is considered as an important technology in image processing, and it has lots of applications. Simply put, its goal is to find out the spatial relationship between two different images. The conception of earth mover＇s distance （EMD） is introduced, which can weigh comparability between two images, then a method is constructed to match image based on EMD. Experimental results indicate that image matching using this algorithm has good performance, when it is compared with other algorithms.
出处
《计算机工程与设计》
CSCD
北大核心
2008年第11期2853-2856,共4页
Computer Engineering and Design
关键词
小波域
盲分离
数字水印
图像置乱
独立分量分析
image matching
comparability measurement
EMD
feature point
wavelet transfer
