项目名称: 面向DLG制作的LiDAR数据处理关键技术研究
项目编号: No.41204033
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 地球物理学和空间物理学
项目作者: 林明华
作者单位: 江苏师范大学
项目金额: 20万元
中文摘要: LiDAR技术能直接获取地表的三维信息,在DEM、DOM制作及三维建模上已得到广泛应用,在测制大比例尺地形图(如1:2000)上应用是否可行?项目围绕这个问题,详细探讨面向DLG制作的LiDAR数据处理的点云滤波、插值和简化三项关键技术。LiDAR数据的滤波中,通过引进双谐样条函数,利用丰富的插值理论研究点云数据的滤波问题;LiDAR的插值,借助于具有良好预测能力的支持向量机(SVM),将点云的插值问题转化为预测问题,以其获得良好的插值精度;在LiDAR数据的简化中,建立可以根据点云的地形、几何特征进行均匀简化和非均匀简化的算法,进一步提高利用点云进行地形成图的效率。通过上述三项关键技术的研究将为点云到DLG成图的软件研制奠定坚实的理论基础,为充分利用LiDAR数据进行高效快速的地形成图发挥重要作用。
中文关键词: 机载激光雷达扫描;点云滤波;点云修复;双谐样条函数插值;支持向量机
英文摘要: LiDAR technology can directly obtain the surface of the three-dimensional information, DEM, DOM production and three-dimensional modeling has been widely applied in determining the system of large-scale topographic maps (eg, 1:2000) on the application,which is feasible? This paper focuses on the problem, discussed in detail for the production of DLG data processing of LiDAR point cloud filtering, interpolation, and simplify the three key technologies. LiDAR data filtering, through the introduction of dual-harmonic spline function interpolation theory with rich point cloud data filtering problem; LiDAR interpolation, by means of a good predictive ability of support vector machine (SVM), the point cloud interpolation issue into a prediction, with its interpolation to obtain a good accuracy; simplified in LiDAR data, create point cloud according to the terrain, to simplify the geometric characteristics of uniform and non-uniform simplified algorithm to further improve the use of point cloud to form a map of efficiency. Through the three key technologies for the point cloud into a map to the DLG software development,it plays an important role to lay a solid theoretical foundation for the full use of LiDAR data for fast and efficient way.
英文关键词: LiDAR;Point Cloud Filter;Point Cloud repairing;Biharmonic Spline Interpolation;SVR