摘要
以广西壮族自治区南宁市树木园坛里管理区为研究区域、以桉树(Eucalyptus spp.)人工林为研究对象,在坛里管理区内设置6块20 m×20 m的典型样地,实地测量单木树高、胸径(1.3 m高)、冠幅直径,获得287株桉树样木数据;样地调查同时进行无人机航拍数据采集,获取分辨率为8 cm像素影像(110张),通过正射校正、倾斜校正、投影差校正对无人机影像进行处理,消除环境因子的影响,采用面向对象的影像分析方法对遥感影像上的桉树林提取林分单木冠幅;根据样地实测数据建立冠幅-胸径模型、冠幅-树高模型,将无人机影像提取的修正后的桉树单木冠幅数据代入构建的模型中,选择相关系数最高的模型推算胸径和树高,利用桉树二元材积公式估算样地的蓄积量,分析依据无人机遥感影像提取桉树单木冠幅数据估算林分蓄积量的可行性与精确度。结果表明:利用无人机影像提取冠幅与实测冠幅之间存在显著正相关,提取的平均精度为90.85%,建立的桉树冠幅-胸径曲线估计模型,其中对数函数方程拟合效果最好(决定系数为0.799);桉树冠幅-树高模型,拟合效果最好的是三次方函数方程(决定系数为0.755)。影像提取的单木冠幅,通过模型得到的蓄积量精度为93.32%,通过模型计算的样地蓄积量与实测蓄积量比对后进行t检验(P>0.05),2组数据差异不显著,模型拟合度较好。采用面向对象的影像分析方法对桉树单木冠幅信息的提取,效果较好;利用影像提取的冠幅,通过实测冠幅-胸径模型、冠幅-树高模型,得到胸径和树高参数,推算林分蓄积量,方法可行,能够快速获取到森林资源数据情况,精度能够满足森林资源调查的应用。
Taking the Tanli Management Area of Nanning Arboretum in Guangxi Zhuang Autonomous Region as the research area and the Eucalyptus spp.plantation as the research object,six typical plots of 20 m×20 m were set up.The height,DBH(1.3 m height)and crown diameter of individual trees were measured on the spot,and 287 Eucalyptus spp.sample trees were obtained.UAV data acquisition was carried out to obtain 8 cm pixel images(110 images).The UAV images were processed by orthorectification,tilt correction,and projection difference correction to eliminate the influence of environmental factors.The object-oriented image analysis method was used to extract the single tree crown of a stand of Eucalyptus spp.from a remote sensing image.According to the measured data of the sample plot,the crown-diameter model and the crown-height model were established.The modified crown data of Eucalyptus spp.single trees extracted by UAV images were substituted into the constructed model,and the model with the highest correlation coefficient was selected to calculate the DBH and tree height.The volume of the sample plot was estimated by using the binary volume formula of Eucalyptus spp.,and the feasibility and accuracy of estimating the stand volume by extracting the crown data of Eucalyptus spp.single tree from UAV remote sensing images were analyzed.There was a significant positive correlation between the crown width extracted from UAV images and the measured crown width,and the average accuracy of extraction was 90.85%.The established crown width-diameter curve estimation model of Eucalyptus spp.had the best fitting effect of the logarithmic function equation(the coefficient of determination was 0.799).The best-fitting effect of Eucalyptus spp.crown width-tree height model is the cubic function equation(the coefficient of determination is 0.755).The accuracy of the volume obtained by the model was 93.32%.The t-test(P>0.05)was performed after comparing the plot volume calculated by the model with the measured volume.The difference between the two groups of data was not significant,and the model fit was good.The object-oriented image analysis method was used to extract the crown width information of Eucalyptus spp.individual tree,and the effect was better.Using the crown width extracted from the image and the measured crown width-diameter model and crown width-tree height model,the DBH and tree height parameters are obtained,and the stand volume is calculated.The method is feasible and can quickly obtain forest resource data.The accuracy can meet the application of a forest resource survey.
作者
覃静婷
刘世男
张丽琼
覃永华
Qin Jingting;Liu Shinan;Zhang Liqiong;Qin Yonghua(Guangxi University,Nanning 530004,P.R.China;Nanning Arboretum;Guangxi Forest Inventory and Planning Institute)
出处
《东北林业大学学报》
CAS
CSCD
北大核心
2023年第6期96-102,共7页
Journal of Northeast Forestry University
基金
广西林业科技推广示范项目(gl2020kt01)。
关键词
无人机影像
桉树
冠幅
蓄积量
UAV(unmanned aerial vehicle)imagery
Eucalyptus spp.
Crown width
Volume
作者简介
第一作者:覃静婷,女,1994年10月生,广西大学林学院,硕士研究生。E-mail:1814203060@qq.com;通信作者:刘世男,广西大学林学院,讲师。E-mail:lsn_smile8866@126.com。
