摘要
为了提高物流防伪标签数字水印抵抗几何攻击性能,设计了一种基于形殊点和非下采样Contourlet变换的的鲁棒数字水印算法。首先利用Canny算子提取图像边缘信息,其次将物流标签图像边缘分块并计算分块子图的形殊点坐标,最后水印信息自适应量化嵌入特征区域的非下采样Contourlet变换的低频变换系数中。检测水印时利用随机抽样一致性算法估计几何攻击前后图像的仿射变换参数。大量仿真实验结果证明,本文算法不仅具有良好的不可见性,同时对常见物流标签图像处理攻击和几何攻击均具有较强的鲁棒性。
In this paper, in order to improve the geometric attack resistance of the watermarks of the logistics anti-counterfeiting labels, we designed a robust algorithm for the digital watermarks based on reference points extraction(RPE) and non-sampling Contourlet transformation(NSCT). Next we introduced the working process of the algorithm and then through a large quantity of simulation experiments, demonstrated that the algorithm was of good invisibility and was robust to the common attacks and geometric attacks in logistics label image processing.
出处
《物流技术》
北大核心
2014年第3期445-448,共4页
Logistics Technology
基金
泰安市科技发展计划"泰山历史文化资源信息化管理查询系统"(20102065)