摘要
结合洞穴类目标的特殊几何结构,对居民窑洞、天然洞窟两种典型洞穴目标进行点云数据模拟生成,并基于模拟点云数据研究目标的快速识别方法,首先分析机载LiDAR系统成像方式,以参数模型和地形TIN模型两种方式分别模拟生成两类洞穴的点云数据;以激光点与LiDAR传感器之间的距离突变为依据实现目标的快速识别,通过限定目标区域搜索范围及设定突变点个数阈值进一步提高识别精度;并完成对目标正面轮廓与内部结构的正确分类。
Based on the special geometry of cave targets, a method for LiDAR point cloud data simulation and fast recognition of residents cave and natural cave is proposed. Firstly the LiDAR working principle is analyzed, and the simulation of airborne LiDAR point cloud is carried out in two ways： based on parametric model and terrain TIN. Through calculating the distance between LiDAR points and sensor position, the target can be quickly identi-fied if there is a range mutation. In order to improve the accuracy of target recognition, the number threshold of mutational points and region-wide threshold are introduced, and the recognized target contour points can be further distinguished between front and inside of the target.
出处
《测绘科学技术学报》
CSCD
北大核心
2014年第3期278-282,共5页
Journal of Geomatics Science and Technology
