Estimation Of Tree Height Of Chinese Fir Based On UAV Remote Sensing

Tree height is an important factor in forest resource investigation and monitoring.To collect accurate tree height data,to better understand the growth status of trees,so as to develop a scientific forest management program to promote the forest ecology,economic efficiency and effective play.At present,the development of satellite remote sensing and aerial photogrammetry and its application in forestry have greatly improved the time-consuming and labor-intensive situation in the traditional artificial forest resources survey.However,due to the low spatial resolution of such remote sensing images,the interval period is long and the high cost of acquisition is obtained,and the remote sensing measurement technology of tree height has not made significant progress.UAV with low-altitude flexible flight,low-cost fast access to images and other technical advantages.It provides an important platform for obtaining remote sensing data with high precision and high temporal resolution,and efficient and accurate remote sensing of tree height.In this study,Baiyun Mountain Forest Farm in Minqing County of Fujian Province was the research area.Based on the UAV aerial multi-spectral images in this area,to produce high-precision digital surface model(DSM)and digital orthophoto map(DOM).Then,the digital elevation model(DEM)and the digital canopy height model(CHM)were fabricated by analyzing interpolation and grid calculation.Using object-oriented image information classification and GIS spatial analysis techniques to extract crown vertices,and the identification of canopy was optimized by using random forest classification algorithm.Based on the field survey data,the UAV aerial survey model was established,and the application of monitoring,the development of forestry information to provide a scientific reference.The main contents and conclusions of the study are as follows:(1)Use Pix4D Mapper software to stitch,correct aerial multi-s pectral images,and create high-precision digital orthophoto(DOM),digital surface model(DSM)and other basic data.The latent tree v ertices were extracted by object-oriented classification,neighborhood analysis and raster calculation.The stochastic forest algorithm was a pplied to the sorting and classification of tree vertex features.The r esults showed that the classification accuracy of tree vertex was 93.46%,among which the normalized vegetation index(NDVI),near r ed band(NIR),ratio vegetation index(SR),atmospheric impedance vegetation index(ARVI)Classification plays an important role.(2)Using GIS software to extract tree vertex tree height from t he CHM layer,construct the UAV remote sensing tree height estim ation model:y=1.3896x0.8931(where y is the estimated tree height,x is the extraction tree High),the height of Chinese fir trees in the study area was estimated.The estimation accuracy of single tree hei ght is 88.66%,and the average tree height of Chinese fir stands is 89%.