基于粒子群优化算法的一种智能水印技术  

AN INTELLIGENT WATERMARKING METHOD BASED ON PSO

作　　者：李伟[1] 孙云娟[1] 刘玉芳[1] 

机构地区：[1]河南师范大学物理与电子工程学院,河南新乡453007

出　　处：《计算机应用与软件》2015年第11期197-199,共3页Computer Applications and Software

基　　金：河南省自然科学基金项目(60977063)

摘　　要：提出的核心方法是在小波域利用粒子群优化算法的一种智能盲水印技术。利用粒子群的最优化算法自适应地计算出量化增量,将二进制数字水印序列0和1逐位嵌入到每一个系数模块中,解决嵌入水印图像质量和水印鲁棒性之间的矛盾。即:将原始图像按照规定阶进行离散小波变换DWT分解出高频子带,随机地将高频子带系数构成固定长度的系数模块,确定每一个模块的有效系数差,根据嵌入规则嵌入水印,按照相逆的方法提取水印。实验仿真说明了该方法的有效性。The core method presented in the paper is an intelligent blind watermarking using PSO in wavelet domain. By using PSO it can adaptively compute the quantitative increment, and embed the binary digital watermark sequence （0 or 1 ） into every coefficient module bit by bit, thus solves the contradiction between the image quality with watermark embedded and the robustness of watermark. That is, the method decomposes the original image into high frequency sub-band with DWT according to specified order, and randomly groups the high frequency sub-band coefficients into coefficient module in fixed length, and determines the difference of significant coefficient of each module, according to the embedding rules it embeds the watermark, and extracts the watermark in inverse way. Experimental simulation demonstrates the effectiveness of the method.

关 键 词：离散小波变换 粒子群算法 有效系数差 盲水印技术 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]