激光扫描数据论文：基于激光扫描数据点集的三维重建算法研究.doc

激光扫描数据论文：基于激光扫描数据点集的三维重建算法研究 【中文摘要】随着激光扫描技术和三维表面重建技术的迅速发展,基于激光扫描数据点集的三维表面重建在科学计算可视化、虚拟现实、医学图像、数字娱乐、文物保护等领域发挥了越来越重要的作用,使其有了非常高的现实研究价值,越来越多的引起人们的关注;另外三维表面重建过程涉及到计算几何、拓扑学以及计算机图形学等多个领域的知识,因此对数据点集的三维表面重建技术的研究具有重大的现实意义和理论意义。本文主要围绕激光扫描数据点集的三维重建的算法展开,首先介绍了三维表面重建中比较常用的曲面以及具体算法;接着针对表面三角网格化,介绍了逐点插入法、三角网生成法、分治算法这三种常用的Delaunay三角剖分算法的构网过程,并对算法的时间复杂度进行分析比较。在对Delaunay三角剖分算法进行深入研究的基础上,综合各种算法的优缺点,提出了一种基于二维数据点重建的新算法。首先将激光扫描得到三维深度数据进行投影变换,把数据点从三维转化到二维平面域,然后对二维数据按改进的Delaunay剖分算法进行三角剖分,该算法分为初始三角网生成和三角网优化两部分,按坐标顺序排列数据点,并直接生成三角网,再根据拥有公共边的四边形的凹凸性对三角网进行优化,然后转换到三维曲面,最后对曲面进行消隐处理,得到三维立体图。有效的提高了Delaunay三角剖分的效率,缩短了剖分所用时间。对MFC框架编程、OpenGL以及计算机图形学技术中的光照模型、纹理映射模型作了详细介绍,最后在VC平台上采用MFC和OpenGL混合编程技术,以激光扫描煤场得到的数据为例,使用本文提出的算法实现了激光扫描数据点集的表面重建,并在此基础上加入了OpenGL的光照处理函数提高了显示效果,证实了算法的有效性。 【英文摘要】With the rapid development of the three-dimensional reconstruction and laser scanning technology, the technology of 3D reconstruction based on laser scanning data plays an very important role in more and more fields, such as scientific visualization, virtual reality, medical imaging, digital entertainment and heritage, so it has very high researching value, and more and more people pay attention to it. The process of three-dimensional surface reconstruction involves several fields, such as computational geometry, topology and computer graphics, so the investigation of 3D surface reconstruction based on laser scanning data has great practical and theoretical significance.This paper is written focus on the algorithm of 3D surface reconstruction based on laser scanning data. Firstly, the more commonly used surface and specific algorithm in the process of reconstruction are introduced; then three methods of Delaunay triangulation are described: one by one point insert, triangular mesh generation method and divide and conquer algorithm, the time complexity among the three algorithms is analysis and compared.The algorithm of Delaunay triangulation has been studied. A new algorithm for two-dimensional data which is based on the D