空域和频域相结合的矢量空间数据多重盲水印算法  

Multiple Blind Watermarking Algorithm Based on Spatial-Frequency Domain for Vector Map Data

作　　者：张黎明 闫浩文[1,2] 齐建勋[3] 张永忠[3] 

机构地区：[1]兰州交通大学测绘与地理信息学院,甘肃兰州730070 [2]甘肃省地理国情监测工程实验室,甘肃兰州730070 [3]兰州市勘察测绘研究院,甘肃兰州730050

出　　处：《测绘地理信息》2016年第5期32-36,共5页Journal of Geomatics

基　　金：国家自然科学基金资助项目(41371435;71563025);甘肃省科技支撑计划项目(1304GKCA009);甘肃省自然基金项目(148RJZA041;148RJZA028);甘肃省财政厅基本科研业务费(214146)~~

摘　　要：针对单一水印算法难以抵抗多种水印攻击的问题,分析了矢量空间数据多重水印的特点,提出了一种空域和频域相结合的矢量空间数据多重水印算法。该算法通过混沌置乱水印图像,建立水印与载体数据之间的Hash单向映射函数,使用两种不同的嵌入方法,水印先后嵌入到空域和DFT域。水印被多次嵌入,实现了水印的盲提取。试验结果表明,该方法引起空间数据误差很小,同时对几何攻击、増删点、裁剪、压缩、要素排序、数据格式转换等攻击具有较好的鲁棒性。Single watermarking algorithm can＇t resist more attacks.Aimed at this issue,the feature of multiple watermarking is analyzed,and we present a multiple watermarking algorithm for vector data of spatial-frequency domain.Through the shuffle of watermark images,a one-way Hash mapping function between the watermark and vector data is established.By using two different approaches,watermark is embedded firstly in spatial domain and then in DFT domain.Through the several embedding of the watermarks,we can realize blind extraction of watermark.Experimental results show that spatial accuracy of the watermarked data is acceptable and the algorithm is robust against a various of attacks such as geometric transformation,vertex insertion and removal,cropping,compression,reordering,data format conversion.

关 键 词：矢量空间数据 多重水印 盲水印 DFT 

分 类 号：P237.3[天文地球—摄影测量与遥感]