摘要
新一代对地观测星载激光雷达卫星ICESat-2采用光子计数体制,获得的单光子点云数据呈现噪声高、沿轨方向剖面式分布的特点,传统激光点云滤波算法不能直接处理。本文提出一种基于空间密度自适应的单光子澈光点云去噪算法,统计局部点密度并转换为密度直方图,利用双高斯函数拟合并精确求解去噪密度阈值。通过WIABEL数据测试表明:本文提出的去噪算法有校强鲁棒性,能有效识别噪声,整体精度达到98%。
The new generation of Earth observation satellite ICESat-2 with spacebonie LiDAR adopts the photon counting system.The obtained single photon point cloud data feature high noise and profile distribution along the orbital direction,which cannot be directly processed by the traditional laser point cloud filtering algorithm.In this paper,a denoising algorithm based on adaptive spatial density for single photon laser point cloud is proposed.The local point density is calculated and converted into density histogram.Then the threshold of density denoising is solved through double Gaussian function fitting.The MABEL data test shows that the denoising algorithm proposed in this paper is robust,and can effectively identify noise with an overall accuracy of 98%.
作者
曹彬才
方勇
江振治
高力
胡海彦
CAO Bincai;FANG Yong;JIANG Zhcngzhi;GAO Li;HU Haiyan(Xi'an Research Institute of Surveying and Mapping,Xi'an 710054,China;State Key Laboratory of(zeo-Iiifomiation Engineering,Xi'an 710054,China)
出处
《测绘科学与工程》
2019年第4期13-17,共5页
Geomatics Science and Engineering
关键词
单光子激光雷达
空间密度
自适应
高斯拟合
去噪算法
single photon lidar
spatial density
adaptive
Gaussian fitting
denoising algorithm
