基于图像特征和霍夫曼编码的图像水印算法  

IMAGE WATERMARKING ALGORITHM BASED ON IMAGE CHARACTERISTICS AND HUFFMAN CODING

作　　者：李亚琴[1,2] 李金祥[1,2] 梁颖红[1,2] 

机构地区：[1]江苏省现代企业信息化应用支撑软件工程技术研发中心,江苏苏州215104 [2]苏州市职业大学计算机工程系,江苏苏州215104

出　　处：《计算机应用与软件》2013年第9期128-130,140,共4页Computer Applications and Software

基　　金：江苏省普通高等学校科研成果产业化推进项目(JH09-46);江苏省现代企业信息化应用支撑软件工程技术研发中心开放基金项目(SX201105);苏州市职业大学创新基金项目(2010SZDCC03);2010苏州市职业大学创新团队项目(3100124)

摘　　要：将霍夫曼编码原理引入数字水印领域,提出一种基于图像特征的数字水印新算法,以降低数字水印算法复杂度。算法在嵌入水印时,首先将数字图像进行三级小波分解,并提取三级逼近分量的边缘特征。然后用霍夫曼编码计算边缘特征矩阵的顶端结点,再将此结点值嵌入到细节分量中。最后进行三级小波逆变换得到嵌入水印的图像。算法在提取水印时,根据嵌入水印的过程找到嵌入点,提取出待检测图像中的水印,并分析得到检测结果。仿真实验表明,该算法复杂度低、载荷小,具有很好的有效性和可行性。In this paper we introduce Huffman coding principle to digital watermarking field, and present a new image watermarking algorithm which is based on image characteristics to reduce the complexity of digital watermarking algorithm. When embedding the watermark, the algorithm conducts 3-level wavelet decomposition on the digital image first, and extracts the edge characteristics of the 3-level approximation components. Then it uses Huffman coding to calculate the vertex nodes of edge characteristics matrix, after that, it embeds the nodes into the detailed components. Finally, 3-level inverse wavelet transform is carried out to get the image with the watermark embedded. When extracting watermark, the algorithm finds the embedding point according to the process of watermark embedding, extracts the watermark from the detecting image, and analyses the detected result. Simulation experiments show that the algorithm has low complexity and small load, and has excellent effectiveness and feasibility.

关 键 词：数字水印 图像特征 霍夫曼编码 边缘特征 小波变换 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]