矢量地图水印技术在地理信息管理中的应用研究

发布时间：2018-06-04 11:57

本文选题：矢量地图 + 地理信息系统　；参考：《延边大学》2017年硕士论文

【摘要】：近年来,数字矢量地图已在地理信息系统、数字化国防建设、智能交通运输系统等领域得到广泛的应用。由于数字矢量地图具有制作成本极高、经济及军事战略意义突出、可重复利用价值巨大等特性,针对矢量地图的非法拷贝、恶意篡改等行为日渐严重。为了维护矢量地图开发者的合法权益和保障矢量地图的内容安全,矢量地图的版权保护已经成为企业和国家发展的关键命题,而数字矢量地图水印技术是实现数字矢量地图版权保护的有效手段之一。传统的矢量地图数字水印技术会对地图数据产生扰动,不适用于高保真的场合,因此本学位论文以实用性为目标,提出了一种基于矢量地图空间特征的零水印算法,为矢量地图版权保护提供一套可行性方案。本学位论文主要根据地图上的空间拓扑关系模型提取矢量地图的特征信息。为了提高算法抗攻击的鲁棒性,在特征提取时,首先采用Douglas-Peucker压缩算法提取地图线图元上的特征节点以提高抗压缩攻击的能力;其次选取地图点图元上的关键节点作为中心点,并根据K-means聚类的原理对特征节点进行分块以提高抗裁剪攻击的能力;最后通过调整中心点坐标精度的方式以提高抗精度约减攻击的能力。特征提取完成后,通过与水印图像进行相互映射以融入版权所有者的身份信息,本论文引入的是有意义的二值图像作为水印图像。为了与地图的特征信息相匹配,采用了基于完全互补码的扩频技术以扩大水印容量;为了保障算法的安全性,采用了基于Logistic映射的混沌置乱技术对水印图像进行加密;最后,结合分块重构技术,将加密后的水印信息与矢量地图的特征信息进行异或运算得到具有版权认证功能的零水印,并将其注册入IPR数据信息库中,同时需添加时间戳等属性信息以抵抗解释攻击。测试结果表明,本文提出的基于矢量地图空间特征的零水印算法适用于高保真矢量地图的版权保护,有效地解决了传统水印算法中地图精度与水印鲁棒性之间的矛盾,在抗矢量地图常见攻击,如几何变换、数据压缩、数据扰乱、精度约减等攻击时表现出很强的鲁棒性,并设置了多重秘钥,使得算法具有很高的安全性。
[Abstract]:In recent years, digital vector map has been widely used in geographic information system, digital national defense construction, intelligent transportation system and so on. Because digital vector map has the characteristics of extremely high production cost, prominent economic and military strategic significance and huge reusable value, the illegal copy of vector map and malicious tampering are becoming more and more serious. In order to safeguard the legitimate rights and interests of vector map developers and ensure the content security of vector map, the copyright protection of vector map has become a key proposition for the development of enterprises and countries. Digital vector map watermarking technology is one of the effective means to realize copyright protection of digital vector map. The traditional vector map digital watermarking technology will disturb the map data, so it is not suitable for the high-fidelity situation. Therefore, a zero-watermarking algorithm based on the spatial feature of vector map is proposed in this thesis, aiming at practicability. It provides a set of feasible scheme for vector map copyright protection. In this thesis, the feature information of vector map is extracted according to the spatial topological relation model on the map. In order to improve the robustness of the algorithm, in feature extraction, Douglas-Peucker compression algorithm is used to extract the feature nodes on the map line graph element to improve the ability of resisting compression attack, and the key nodes on the map point element are selected as the center points. According to the principle of K-means clustering, the feature nodes are divided into blocks to improve the ability of resisting clipping attack, and finally, the ability of anti-precision reduction is improved by adjusting the precision of center coordinates. After the feature extraction is completed, by mapping with the watermark image to incorporate the identity information of the copyright owner, this paper introduces a meaningful binary image as the watermark image. In order to match the feature information of the map, the spread spectrum technology based on the complete complementary code is used to enlarge the watermark capacity. In order to ensure the security of the algorithm, chaotic scrambling technology based on Logistic mapping is used to encrypt the watermark image. Combined with block reconstruction technology, the encrypted watermark information and the feature information of vector map are used for XOR operation to obtain the zero watermark with copyright authentication function, and the watermark is registered into the IPR data database. At the same time, we need to add time stamp and other attribute information to resist the interpretation attack. The test results show that the proposed zero-watermarking algorithm based on the spatial features of vector maps is suitable for copyright protection of high-fidelity vector maps, and effectively solves the contradiction between map accuracy and watermark robustness in traditional watermarking algorithms. The algorithm is robust against common attacks such as geometric transformation, data compression, data scrambling, precision reduction and so on, and sets up multiple secret keys, which makes the algorithm highly secure.
【学位授予单位】：延边大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P208;TP309.7

【相似文献】