基于ArboLiDAR的林分自动分割研究与应用

摘要

森林作为陆地生态系统的主体和自然界功能最完善的资源库,不仅为人类的生存发展提供了物质保障,也在维护陆地乃至全球生态平衡上发挥着重要作用.LIDAR技术作为一种主动遥感手段,在对森林植被空间结构的探测和参数反演方面具有显著的优势.而在森林参数反演之前,对研究区进行合理的林分分割很有必要.利用祁连山大野口林区2008年的激光雷达点云数据,首次引进国外先进的LiDAR点云数据处理软件ArboLiDAR,经过前期的数据分类和相关预处理操作,将数据代入ArboLiDAR软件,通过多次设定相关分割参数,选择最优参数设置,以完成对研究区林分分割操作.并通过对该地区进行了林分平均树高的提取与精度评价,来估测林分自动分割结果,以期提高其他林分参数的估测精度.结果表明:LiDAR数据对林分平均高估测的相关性系数达到0.807,RMSE为1.12 m,精度较高,表明ArboLiDAR平台可以对LIDAR数据做较高精度的林分自动分割,为林分水平上的其他森林参数提取奠定基础,也为LIDAR数据的林分区划提供新的思路.%Forests are the main body of terrestrial ecosystem and the most complete resource library of nature features.They can not only provide the materials for human survival and development,but also play an important role in the maintenance of the land and the global ecological balance.As a kind of active remote sensing technology,light detection and ranging technology has obvious advantages in the detection of the spatial structure of forest vegetation and the inversion of forest parameters.Reasonable forest stand segmentation is the key to the study area before the inversion of forest parameters.This paper introduces the foreign advanced LiDAR point cloud data processing software ArboLiDAR for the first time.First of all,the LiDAR point cloud data using in the Qilian Mountains Dayekou area in 2008 were classified and related preprocessing operated.Then the optimal parameter settings are selected by setting relevant segmentation parameters,and the operation of the forest segmentation in the study area is completed in ArboLiDAR.Last,the result of the automatic stand segmentation was evaluated through the extraction and validation of the stand average height(HGW) in the area,so as to provide reference for estimating other stand parameters.According to the estimation results show,the correlation coefficient of average HGW of stand was 0.807,RMSE to 1.12 m,the precision is high.The result indicated that ArboLiDAR platform can do the automatic segmentation of LiDAR data with high accuracy,lay the foundation for other forest parameters extraction on the stand level,and provide new ideas for the forest regionalization by using LiDAR data.