Automatic Estimation Of Forest Plots Parametei Based On TLS And UAV Image Point Clouds

Forest resource investigation is an important way to clearly understand the situation of growth and change of national forest.Tree height,diameter at breast height(DBH)and crown width,and canopy density and volume of forest plots are the key parameters of forest resource survey.The method of obtaining forest parameters from traditional field measurement and remote sensing image of limited resolution can not meet the requirements of precision forestry and information development.Currently,estimating forest key parameters based on 3D point cloud data has become a research hotspot in forestry remote sensing.In this paper,for a comprehensive description of forest structure,and improve the estimation accuracy and efficiency of forest key parameters,forest plots in Jiangle County of Fujian Province as the representative area,study the methods of extracting forest key parameters based on terrestrial laser scannin(TLS)point cloud and unmanned aerial vehicle(UAV)image point cloud data.The main contents and conclusions are as follows:(1)Research on filtering algorithm of plot point cloud data.The irregular grid is used to separate the ground points and the non-ground points,and the digital elevation model(DEM)precision satisfies the requirement of parameter extraction.This is the basis for obtaining the digital canopy height model(CHM)of the plot.(2)Extraction of key parameters of forest plots based on TLS single point cloud data.Based on 27 TLS single scanning point cloud data,the accuracy of extraction of tree height,DBH of the individual tree by the fitting cylindrical algorithm and height threshold method respectively,and estimation of plot volume were analyzed on the individual tree scale.Constructed the Height-DBH prediction model,effectively improve the estimation accuracy of individual tree height of crown point cloud lack.evaluated the accuracy of direct estimation method using the ratio the area of the CHM and the plot.The results showed that there was a clear linear correlation between the estimated average tree height,average DBH,volume,canopy density and the field measured average values of plots.It is not only efficient,but also overall accuracy,which can meet the accuracy requirement of forestry plot survey.(3)Based on the TLS single point cloud automatic extraction tree height and DBH of multi-tree of coniferous forest plots.In order to speed up the extraction efficiency of forest plots parameters,select 14 point cloud of good quality coniferous forest plots for the test plots,study the minimum hierarchical segmentation algorithm,achieved the individual tree of plots automatic segmentation and tree height and DBH automatic extraction at the same time.The experimental results show that the accuracy of the parameter estimation is about 90%for the correct identification of the single trees.(4)Fusion UAV image point cloud and TLS point cloud data,and extract tree height and DBH,crown width and crown density of forest plots.For the low canopy density plots,effectively extraction of individual tree height and crown width base on single UAV image cloud data by local maxima,minimum algorithm.In order to solve the tree height extraction problem of high canopy forest plot,fusion TLS point cloud and UAV image point cloud data,and unified the coordinate of two data by feature points.The results show that the precision of parameter estimation is significantly higher based on the fusion data than that single data.This method can describe the forest plots from two aspects of the ground and the air simultaneously,providing a new way for more comprehensive and accurate estimation of the parameters of high canopy density forest plots in subtropical.