海底地形三维可视化技术研究

发布时间：2018-03-15 12:39

本文选题：自然邻点插值　切入点：海底地形　出处：《哈尔滨工程大学》2013年硕士论文　论文类型：学位论文

【摘要】：随着国家经济、国防事业的发展，海洋的开发及其利用对国家越来越重要，海洋三维模型相比海洋二维模型展现出更大的优势和发展前景，建立真实的海洋三维模型的紧迫性得到人们的逐步认识。同时随着计算机软硬件的快速发展，三维可视化技术和虚拟现实技术的也得到了广泛的发展应用。这一切都为海底三维可视化技术的发展提供了有利条件。本文在这种背景下开展了对海底三维可视化技术的研究。首先，本文分析了海底地形建模相关基础，从海洋地形数据的获取方法、海底地形数据的存储模型、海底地形数据的插值方法三个方面进行总结分析，从而选出将电子海图作为数据源，将规则网格模型作为海底地形数据的存储模型，将自然邻点插值作为本文海底地形数据的插值方法。其次，本文介绍了基于Voronoi图和Delaunay三角形的自然邻近插值方法，比较了Sibson插值和Laplace插值的优缺点，，使用了快速去寻找自然邻点的方法从而避免去生成Voronoi图和Delaunay三角形。最后分析了Laplace插值的局限性，Laplace插值无法在地形数据形成的凸壳上和外部进行插值，基于此本文提出了加权平均法和Laplace插值混合插值的方法。同时通过仿真，从插值速度、插值精度及其插值效果上与加权平均法及传统自然邻点插值法进行了比较分析。再次，本文介绍了基于二叉树的ROAM技术基础，从中具体分析了当前视截体裁剪方式的优缺点，提出了基于视截体的快速粗糙裁剪和子块内视截体的细致裁剪的改进方法。使用混合裁剪方法，不仅加快了裁剪速度，减少了三角形冗余，同时也保留了地形的细节。最后，从电子海图中提出离散的海底地形数据，使用加权平均法和Laplace混合插值算法生成规则网格数据，结合改进的ROAM算法，用OpenGL图形API函数对海底三维地形进行了漫游测试。从海底地形算法漫游帧率、生成的三角形个数技术指标上与全分辨率显示算法、传统ROAM算法进行了比较分析。
[Abstract]:With the development of national economy and national defense, the development and utilization of ocean is becoming more and more important to the country. The urgency of building a real three-dimensional ocean model has been gradually recognized. At the same time, with the rapid development of computer software and hardware, 3D visualization technology and virtual reality technology have also been widely developed and applied, all of which provide favorable conditions for the development of seabed 3D visualization technology. In this paper, the research of seabed 3D visualization technology has been carried out under this background. First of all, this paper analyzes the relevant foundation of submarine terrain modeling, summarizes and analyzes the acquisition method of ocean terrain data, the storage model of submarine terrain data, and the interpolation method of submarine terrain data. Thus, the electronic chart is chosen as the data source, the regular grid model is used as the storage model of submarine topographic data, and the natural neighbor interpolation is used as the interpolation method for the topographic data of this paper. Secondly, this paper introduces the natural neighbor interpolation method based on Voronoi graph and Delaunay triangle, and compares the advantages and disadvantages of Sibson interpolation and Laplace interpolation. In order to avoid the generation of Voronoi diagrams and Delaunay triangles, the method of finding natural neighbor points quickly is used. Finally, the limitation of Laplace interpolation is analyzed, which can not interpolate on and outside convex shells formed by topographic data. In this paper, the weighted average method and the Laplace interpolation mixed interpolation method are proposed, and the simulation results are compared with the weighted average method and the traditional natural neighbor interpolation method in terms of interpolation speed, interpolation accuracy and interpolation effect. Thirdly, this paper introduces the technology foundation of ROAM based on binary tree, and analyzes the advantages and disadvantages of the current cutting method. This paper presents an improved method of fast rough clipping based on visual cutlery and meticulous clipping of subblock. Using mixed clipping method, not only the cutting speed is accelerated, triangulation redundancy is reduced, but also terrain details are preserved. Finally, discrete submarine topographic data are proposed from electronic charts, and regular grid data are generated by using weighted average method and Laplace mixed interpolation algorithm, and combined with improved ROAM algorithm. The roaming test of submarine 3D terrain is carried out by using OpenGL graphic API function, and compared with the full resolution display algorithm and traditional ROAM algorithm in terms of roaming frame rate of submarine terrain algorithm and the number of triangles generated.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P229.1

【参考文献】