低矮植被的无人机激光雷达测高精度分析  

High precision analysis of low vegetation for UAV LiDAR

作　　者：杨凡[1,2] 杨贵军[2,3,4] 李振海[2,3] 孟鲁闽 赵晓庆[2,5] 

机构地区：[1]西安科技大学测绘科学与技术学院,西安710054 [2]北京农业信息技术研究中心,北京100097 [3]国家农业信息化工程技术研究中心,北京100097 [4]农业部农业信息技术重点实验室,北京100097 [5]北京市农业物联网工程技术研究中心,北京100097

出　　处：《测绘科学》2017年第9期151-157,共7页Science of Surveying and Mapping

基　　金：国家自然科学基金项目(41271345;41471351);北京市自然科学基金项目(4141001);北京市农林科学院科技创新能力建设项目(KJCX20140417);测绘地理信息公益性行业科研专项项目(201512010)

摘　　要：针对无人机激光雷达估测低矮植被高度的精度大小,本文以3m以下低矮树木为研究对象,通过数值模拟方法得到不同航高、不同扫描角情况下激光脚点坐标和点云估测单一树木高度的最大测量误差值;对比实测树高,分析了激光点云估测树高的精度。结果表明,在航高为30m、扫描范围为(-50°,5°)的情况下,无人机激光雷达获取的激光脚点坐标误差和由激光点云估测低矮树高(3m以下)的误差均可以达到cm级;激光点云估测单一树木高度与实测高度的决定系数为0.977,均方根差为5cm,标准均方根差为4%。因此,应用无人机激光雷达数据可以快速、精确获取低矮植被高度信息,进而为反演植被生物量和植被长势信息监测提供重要依据。Aiming at the precision of the low vegetation height estimation of the unmanned aerial vehi- cle（UAV）laser radar, the maximum error value of the laser foot point coordinate and point cloud estima- tion single tree height under different flight height and different scanning angle were obtained by the nu- merical simulation method with the low tree height below 3 m. Compared with the measured tree height, the accuracy of the laser point cloud estimation tree is analyzed. The results show that under the condition that the height is 30 m and the scanning range is -50°to 50, the error of the laser foot point coordinate obtained by the UAV laser radar and the error of the low tree height（below 3 m） estimated by the laser point cloud can reach the centimeter level. The determination coefficient of single tree height and measured height of laser point cloud is 0. 977, the root mean square difference is 5 cm, and the standard root mean square difference is 4 %. Therefore, the application of UAV lidar data can quickly and accurately obtain the low vegetation height information, and thus provide an important basis for the retrieval of vegetation biomass and vegetation long potential information monitoring.

关 键 词：无人机激光雷达 低矮植被 高度 精度 

分 类 号：P232[天文地球—摄影测量与遥感]