基于机载LiDAR的伪大光斑波形数据反演林分平均高  

Retrieval of mean canopy height based on pseudo large footprint data from airborne LiDAR

作　　者：闫灿[1] 邢艳秋[1] 高金萍 辛颖[1] 田昕 YAN Can;XING Yanqiu;GAO Jinping;XIN Ying;TIAN Xin(Center for Research Institute of Forest Operations and Environment,Northeast Forestry University,Harbin 150040,Heilongjiang,China;National Forestry and Grassland Bureau Investigation and Planning Institute,Beijing 100714,China;Heilongjiang Provincial Research Institute of Surveying and Mapping,Harbin 150040,Heilongjiang,China)

机构地区：[1]东北林业大学森林作业与环境研究中心,黑龙江哈尔滨150040 [2]国家林业与草原局调查规划设计院,北京100714 [3]黑龙江省测绘科学研究所,黑龙江哈尔滨150040

出　　处：《中南林业科技大学学报》2019年第12期62-69,共8页Journal of Central South University of Forestry & Technology

基　　金：国家林业公益性行业科研专项(201504319)

摘　　要：为降低小光斑机载激光雷达因光斑直径太小而导致的脉冲首次回波无法代表冠层高度的影响,以进一步提高小光斑机载激光雷达波形数据在森林结构参数估测中的应用潜能。以内蒙古依根地区为研究区,以机载激光雷达波形数据为基础数据,在波形数据高斯分解的基础上提出一种基于小光斑波形形成伪大光斑波形数据的方法。通过计算样地内各高斯分量脉冲能量占总脉冲能量的比例,将其视为各高斯分量特征参数对应权数,分别求特征参数振幅、位置和半波宽的加权平均数,即为样地伪大光斑波形数据对应高斯函数的特征值。基于小光斑波形数据和伪大光斑波形数据提取特征参数,分别结合野外样方实测平均树高建立回归模型,并进行比较分析。结果小光斑波形反演模型的决定系数R2为0.47,总体平均精度P为78.19%,伪大光斑反演模型的决定系数R2为0.61,估测林分平均高总体平均精度P为90.65%。结果表明,伪大光斑模型反演精度高于小光斑波形反演模型,降低了小光斑LiDAR因光斑直径过小带来的影响,挖掘了小光斑机载LiDAR波形数据的应用潜力。Due to the spot diameter of small footprint LiDAR is too small,the canopy height can not be represented by the pulse first echo.To lower the impact and further improve the application potential of small footprint airborne LiDAR full-waveform data in forest structural parameter estimation.In this study,a new method based on Gaussian decomposition by waveform fitting was proposed.Additionally,the airborne LiDAR full-waveform data of Inner Mongolian Irogen district was obtained and processed.By calculating the ratio of the Gaussian component pulse energy to total pulse energy in the sample,it was regarded as the corresponding weights of each Gaussian component feature parameter,and the weighted average of the amplitude,position and half-wave width of the feature parameter was obtained respectively,which was the the feature value corresponding to the Gauss function of pseudo large footprint waveform data.Based on the waveform data and the pseudo large footprint data extraction characteristic parameters,the regression model was established by combining the field measured Lorey’s average tree height respectively,and the comparative analysis was carried out.The model determination correlation R2 of the waveform inversion model was 0.47,the average overall precision P was 78.19%,the model determination correlation R2 of the pseudo large footprint inversion model was 0.61,and the average overall precision P was 90.65%.It was concluded that the inversion accuracy of the pseudo large footprint model is higher than the waveform inversion model,which reduces the effect brought from the small footprint diameter of LiDAR,and excavates the application potential of the airborne LiDAR waveform data of small spot.

关 键 词：机载LIDAR 波形数据 伪大光斑 林分平均高 

分 类 号：S758.54[农业科学—森林经理学]