摘要
利用具有混沌特性的摩尔型元胞自动机(Cellular Automata,CA)对原始水印图像进行置乱加密,接着将原始载体图像进行8×8分块离散余弦变换(DCT),按预定顺序读取低频和中频系数构成与水印图像同尺寸的系数矩阵,并在该系数矩阵的奇异值分解域中完成已加密水印信息的嵌入。仿真结果表明,该算法不仅具有良好的不可见性,而且对常见攻击如JPEG压缩、滤波、噪声和几何攻击等具有较强的稳健性,在版权保护领域具有一定的应用价值。
A DCT field watermarking algorithm for digital image based on Moore-model cellular automata (CA) scrambling is proposed in this paper. Firstly, Moore-model CA with chaotic characteristics is used to :disorder the original mark. Then the original carrier image is carried on 8×8 block Discrete Cosine Transform (DCT), where low or medium frequency coefficients of the carrier image is selected in the order fixed in advance to create a matrix; with the same size as the watermarking images and the singular value decomposition (SVD) of matrix is carride out. Finally, completed encrypted watermark is embedded in decomposition domain of SVD. Experimental results show that the new watermarking scheme has good invisibility of embedded watermark image and is robust against common attacks such as JPEG compression, low-pass filtering, adding noise, and geometric distortion, it has some practical value for copyright protection.
出处
《电视技术》
北大核心
2010年第7期41-44,共4页
Video Engineering
基金
国家自然科学基金资助项目(10765002)
广西教育厅科研项目(200911MS51)
广西师范大学青年骨干教师基金项目(师政科技20097)
广西自治区高校重点建设实验室资助项目(200912)