【作者】 夏松竹；

【导师】 张健沛；

【作者基本信息】 哈尔滨工程大学 ， 计算机应用技术， 2010， 博士

【摘要】 随着信息处理技术和网络技术的快速发展,数字地图已逐步成为地理信息系统、智能交通运输系统、数字化城市和数字化国防建设等方面必需的保障资源。数字矢量地图的地理信息详实准确,具有巨大的经济价值和战略意义。同时,数字地图制作成本极高,针对数字地图的非法拷贝、窃取等侵权行为也日益猖獗。对于数字地图的数据安全防范措施稍有不当,就会给企业造成损失,甚至会对国家安全造成严重的威胁。为此,数字地图的内容安全问题已受到我国政府、军队和科研院所以及相应企业的关注；数字地图水印技术是一种解决数字矢量地图版权保护问题的常用手段。虽然人们对数字地图水印技术进行了近二十年的研究,提出了若干技术并试图解决数字地图在应用中遇到的各类安全问题,但这些技术并没有很好地解决数字地图的.无损信息隐藏问题。数字地图加入隐藏信息后或多或少的会对地图有损害。本文以数字地图的两大类别：数字矢量地图和数字影像地图为主要研究对象,以无损信息隐藏为研究内容,从数字地图的实用角度出发,深入研究了可满足数字矢量地图内容无损要求、抵抗多种攻击方式的数字地图无损信息隐藏水印技术。本文的创新工作主要包括以下几个方面：首先,结合数字矢量地图的特点以及小波分形理论,本文提出了基于小波分形无损的零水印算法,并且详细介绍了基于小波分形无损的零水印算法的技术实现过程,通过有效性、鲁棒性、IBM攻击测试等实验表明：该技术对数字矢量地图的精度无损,且具有较高鲁棒性,结合第三方认证机制,该技术可以完全抵制IBM攻击,使该水印技术具备真正的可行性和实用性。其次,针对空间域算法主要是通过修改像素信息来嵌入水印,技术实现简单,鲁棒性差,对图像的损伤也较大,无法抵抗多种攻击等缺点。本文提出了基于波段融合的数字影像地图无损信息隐藏技术,以IMG格式的数字影像地图为例详细介绍了技术的实现流程。经过抗攻击性能测试、不可见性检测与分析等方式,验证结论表明该技术达到了水印无损嵌入和提取的目标,且具有较高的鲁棒性和隐蔽性。最后,在数字矢量地图零水印技术和数字影像地图无损信息隐藏技术研究的基础上,针对上述无损信息隐藏技术地图格式依赖性强,自适应性差的特点,本文提出了格式无关的数字地图可逆数字水印技术。该技术运用云模型的不确定性特征,实现了对主流格式数字地图的无损信息隐藏,相继介绍了云水印的生成、嵌入和提取方法,并就此进行了大量的实验验证。依据云的模糊性和随机性理论,当部分云滴水印受到攻击后,云模型也不会有太大变化,因而具有较好的鲁棒性。并且该算法可适应于数字矢量地图、数字影像地图等多种格式的数字地图。更多还原

【Abstract】 With the fast development of information processing technology and network techniques, digital vector map has becoming the necessary resource being widely used in GIS, intelligent transport system, digital city, digital national defense construction and so on. Digital vector map has great economical value and strategy significance for its accurate content. And its fabricating cost is huge. On the other hand, illegal copying and stealing of digital vector map are becoming increasingly rampant. A lower level security of it will cause enterprises’losses and even bring risks to national security. As a result, the security problem has attracted our government, military and science research institution’s attention. Digital watermarking technology is a common scheme to secure the copyright of digital vector map.The research on digital watermarking for digital vector map has lasted for about 20 years and many schemes have been proposed to solve some kinds of security problems of it. But these schemes have not fully considered the data characteristic of digital vector map and the particular operations on it.The two kinds of digital map, digital vector map and digital image map has been discussed in the paper. The author has studied lossless information hiding technology based on the application that can losslessly recover waterwark information, resist some attacks and trace the source maps have been proposed. The main innovative points are the following:Firstly, this paper presents a lossless zero watermarking algorithm based on wavelet fractal, which is the combination of fractal, wavelet, multiple image fractals, and wavelet fractal and so on, and the technology and implementation procedure of lossless zero watermarking algorithm based on wavelet fractal has been presented in detail. The strong robustness and lossless to vector map is proved by validity test, robustness test and IBM attack test. This model absolutely resists IBM attack and has the real value of practicability and feasibility with the third-party authentication mechanism.Secondly, spatial domain algorithm is realized by embedding watermark through changing pixel information, so its robustness is weak, its image is hurt mostly and cannot resist many kinds of attack although the implementation is simple. This paper presents the zero watermarking based on wave band superposition, of which image map is an example. What’s more, the paper discusses the implementation and test of algorithm using IMG format image map as an example. This algorithm absolutely has lossless watermark embedding and watermark extracting, also has strong robustness and concealment, which is tested by anti-attack and multi-attack experiment.Finally, this paper presents independent format lossless watermarking algorithm of digital map based on digital vector map watermarking algorithm and image map watermarking algorithm because of their weak self-adaption. The algorithm combining with the advantage of cloud model and digital watermark presents generating, embedding and extracting method of cloud watermark, and do many test based on the algorithm. Cloud model has not much change although some cloud droplets is attacked because of cloud’s fuzziness and randomness. The algorithm has strong robustness. In addition, the algorithm adapts to digital vector map, image map and so on, the algorithm has nothing to do with digital map’s format.更多还原