基于机载LiDAR数据的农作物叶面积指数估算方法研究

Estimation Method of Crop Leaf Area Index Based on Airborne LiDAR Data

叶面积指数(LAI)是农作物长势监测及估产的重要参数,激光雷达能够提供精确的农作物冠层结构信息,可弥补光学遥感在提取冠层结构信息方面的不足。因此,本文旨在挖掘激光雷达所能提取的农作物垂直结构信息,并研究冠层结构参数与农作物叶面积指数之间的关系,从而估算整个研究区的叶面积指数。首先,基于机载激光雷达数据提取平均高度(Hmean)、最大高度(H_(max))、最小高度(H_(min))、高度百分位数(H_(25th)、H_(50th)、H_(75th)、H_(90th))、激光穿透力指数(LPI)、回波点云密度、孔隙率(fgap)、叶倾角(MTA)等结构参数;然后,利用Pearson相关性分析法对以上参数与地面实测LAI进行相关性分析,并选择与LAI相关性高的参数;最后,对选择的敏感性参数进行回归分析,构建激光雷达参数与实测LAI的LiDAR-LAI估算模型,估算整个研究区的农作物冠层LAI。精度评价结果表明:预测LAI与实测LAI之间的相关系数为0.79,均方根误差为0.47,说明激光雷达所提取的农作物冠层结构参数可用于估算空间上连续、大面积的农作物LAI。

Leaf area index（ LAI） is an important parameter in crop growth monitoring and crop yield estimation. However,optical remote sensing cannot extract the structural information. Light detection and ranging（ LIDAR） can provide accurate crop structural information,so Li DAR can make up the shortage of optical remote sensing. Therefore,the purpose of this research is to study the vertical structure information of crops which can be extracted by Li DAR,analyze the correlation of LiDAR vertical metrics and LAI of crop,and estimate LAI of the whole study area. First,the metrics were extracted based on LiDAR data,including mean height above ground of all first returns（ Hmean）,maximum height above ground of all first returns（ Hmax）,minimum height above ground of all first returns（ H_（min））,the percentiles of the canopy height distributions（ H_（25th）,H_（50th）,H_（75 th）,H_（90th））,laser penetration index（ LPI）,density of points,porosity and leaf angle. Then,Pearson correlation analysis was used to filter Li DAR metrics which are better related to LAI measured data. Last,regression analysis of selected sensitive parameters was carried out on setting up Li DAR-LAI estimation model,and the LAI estimated result of the whole study area was calculated. The result shows that correlation coefficient between estimated LAI and field measured LAI is 0. 79,and RMSE is 0. 47. It shows that crop canopy structure parameters extracted by Li DAR can be used to estimate the spatial continuous and large area of LAI of crops.