基于改进DBSCAN的单光子激光点云去噪算法  

Single-photon laser point cloud noise reduction algorithm based on improved DBSCAN

作　　者：张少华 施银迪 谭莲红 张译方 Zhang Shaohua;Shi Yindi;Tan Lianhong;Zhang Yifang(Shandong No.3 Exploration Institute of Geology and Mineral Resources,Yantai 264000,China;Huai’an Water Conservancy Survey and Design Institute Co.,Ltd.,Huai’an 223005,China;Yantai Taochu Municipal Engineering Co.,Ltd.,Yantai 265500,China;Shandong Yantai Geological Engineering Survey Institute,Yantai 264011,China)

机构地区：[1]山东省第三地质矿产勘查院,山东烟台264000 [2]淮安市水利勘测设计研究院有限公司,江苏淮安223005 [3]烟台陶褚市政工程有限公司,山东烟台265500 [4]山东省烟台地质工程勘察院,山东烟台264011

出　　处：《矿山测量》2022年第1期32-37,共6页Mine Surveying

摘　　要：针对单光子激光雷达系统观测数据中背景噪声过多的问题,文中提出了一种基于改进DBSCAN的单光子激光点云去噪算法。以ICESat-2卫星在工作运行期间所采集的ATL03数据为实验数据,通过设置合理的阈值进行粗去噪,然后基于改进DBSCAN聚类的算法对点云数据进行精去噪,并进行理论分析和实验验证。结果表明,实验区目标点云去噪精度达到99.11%,性能优于传统滤波算法。Aiming at the problem of the excessive background noise in single-photon lidar system observation data,in this paper,a single-photon laser point cloud noise reduction algorithm was proposed based on improved DBSCAN.Taking ATL03 data collected during the operation of ICESat-2 satellite as the experimental data,the coarse noise reduction was performed by setting a reasonable threshold,and then the fine noise reduction of the point cloud data was performed based on the improved DBSCAN clustering algorithm,and theoretical analysis and experimental verification were carried out.Results showed that the noise reduction accuracy of the target point cloud reached 99.11%in the experimental area,which was better than the traditional filtering algorithm.

关 键 词：单光子激光雷达系统 粗去噪 精去噪 DBSCAN聚类 

分 类 号：P225[天文地球—大地测量学与测量工程] TN958.98[天文地球—测绘科学与技术]