摘要
文章在分析图像的奇异值分解特性的基础上,提出了一种基于奇异值分解的载体奇异向量与水印奇异向量替换的盲水印算法。将载体图像分块进行奇异值分解,用水印分块奇异值分解的左右奇异向量和奇异值向量替换载体的对应向量。根据载体与水印的大小关系将水印循环嵌入,提取时进行平均判决。算法简单易于实现,提取的水印相似度高,可应用于高品质图像的版权保护与认证。
Based on property analysis of signal value decomposition, a singular value decomposition （SVD）-based blind watermarking technique is proposed, which replaces the signal vectors of host with the signal vectors of watermarking. By making SVD on the host image block, the singular vectors and singular values are replaced with the corresponding ones of the watermark block. According to the sizes of the host image and watermark, the watermark is embedded iteratively, and the mean value is adopted when the watermark extracted. The algorithm is simple and easy in implementation, the extracted watermark is of high similarity. This algorithm could be applied in copyright protection and authentication of the high-quality image.
出处
《通信技术》
2009年第9期94-96,共3页
Communications Technology
基金
重庆市自然科学基金资助项目(项目编号:CSTC.2007BB2105)
关键词
数字水印
奇异值分解
奇异向量
digital watermarking
singular value decomposition （SVD）
singular vector
