顾及语义的机载LiDAR点云格网化方法  

LiDAR Data Cloud Gridding Based on Semantics

作　　者：邬建伟[1] 马洪超[1] 李奇[1] 

机构地区：[1]武汉大学遥感信息工程学院,湖北武汉430079

出　　处：《测绘科学技术学报》2008年第2期87-89,93,共4页Journal of Geomatics Science and Technology

基　　金：国家863计划项目资助(2006AA12Z101)

摘　　要：传统的点云格网化方法采用纯数字拟合内插的方法,基本不顾及平面坐标和高程之间存在的潜在关系———语义信息。将语义信息纳入点云格网化生成的高程格网具有更高的语义一致性,更适合用于3维可视化和真正射影像的生成。以静态水体为例,研究了顾及语义的机载L iDAR点云格网化方法。最后通过实验验证了该方法的有效性。The conventional gridding methods of point clouds were based on the mathematic principles of fitting and interpolation, whereas the potential relationships（ semantic information） between the horizontal and vertical positions were seldom considered. When incorporating the semantic information, we could produce height grids with more se- mantic consistency, which was helpful for 3D visualization and true ortho-photo generation. Taking the water body as example, gridding methods considering semantic information were researched. At last, an experiment was conducted to demonstrate the efficiency of the proposed method.

关 键 词：LIDAR点云 点云格网化 语义信息 数字表面模型 

分 类 号：P237[天文地球—摄影测量与遥感]