基于无人机高光谱影像和深度学习算法的长白山针阔混交林优势树种分类  

Classification of dominant species in coniferous and broad-leaved mixed forest on Changbai Mountain based on UAV-based hyperspectral image and deep learning algorithm

作　　者：郑迪 沈国春[3] 王舶鉴 戴冠华[1] 蔺菲[1] 胡家瑞 叶吉[1] 房帅[1] 郝占庆 王绪高[1] 原作强[1] ZHENG Di;SHEN Guo-chun;WANG Bo-jian;DAI Guan-hua;LIN Fei;HU Jia-rui;YE Ji;FANG Shuai;HAO Zhan-qing;WANG Xu-gao;YUAN Zuo-qiang(C4S Key Laboratory of Forest Ecology and Management,Institute of Applied Ecology,Chinese Academy of Sciences,Shenyang 110016,China;University of Chinese Academy of Sciences,Beijing 100049,China;School of Ecological and Environmental Sciences,East China Normed University,Shanghai 200241,China;School of Ecology and Environment,Northwestern Polytechnical University,Xi'an 710129,China)

机构地区：[1]中国科学院沈阳应用生态研究所,中国科学院森林生态与管理重点实验室,沈阳110016 [2]中国科学院大学,北京100049 [3]华东师范大学生态与环境科学学院,上海200241 [4]西北工业大学生态环境学院,西安710129

出　　处：《生态学杂志》2022年第5期1024-1032,共9页Chinese Journal of Ecology

基　　金：美丽中国生态文明建设科技工程专项(XDA 23080301);国家重点研发计划项目(2016YFC0500300);国家自然科学基金项目(31971439和31961133027);中国科学院青年创新促进会项目(2017241)资助。

摘　　要：快速、准确识别树种及其分布格局是森林资源经营管理和生物多样性保护的基础和前提。与传统实地调查的方法相比,近年来飞速发展的近地面遥感技术可以灵活、高效和便捷地获取高分辨率高光谱遥感影像,而如何从包含丰富信息的诸多特征中选择信息量大且冗余度低的特征进行树种自动识别,是当前研究亟待解决的问题。本研究以长白山25 hm^(2)温带针阔混交样地为主要研究平台,于2019年8月使用无人机搭载的光谱传感器获取面积为6 hm^(2)的高光谱影像,选择红松、春榆、蒙古栎、水曲柳、大青杨和紫椴6种林冠层树种作为实地标记树种,使用实时载波相位差分技术对所选目标树种进行精确定位,结合2019年样地复查结果对研究区的影像进行目视解译,分别使用卷积神经网络法、最大似然法和马氏距离3种分类方法进行冠层树种的自动分类研究。结果表明:(1)卷积神经网络的树种分类总体精度和Kappa系数(99.85%、0.998)优于最大似然法(89.11%、0.86)和马氏距离法(79.65%、0.75)。(2)在3种分类方法中,单个优势树种分类精度均在卷积神经网络中为最高精度,红松、春榆、蒙古栎、水曲柳、大青杨和紫椴的最高分类精度分别为100%、99.9%、99.9%、99.8%、99.8%和99.5%。(3)从分类效果看,卷积神经网络混分程度最低,马氏距离法混分程度最严重。研究表明,基于深度学习的卷积神经网络模型能够完成对温带天然林林冠树种的准确高效分类,在树种多样性监测和林业资源调查应用中具有较大潜力。Rapid and accurate identification of tree species and their distribution patterns is the basis and premise for forest resource management and biodiversity conservation.Compared with the traditional field investigation methods,the rapid development of the surface of remote sensing technology in recent years can obtain high resolution hyperspectral remote sensing images flexibly,efficiently,and conveniently.However,how to select features with large amount of information and low redundancy from many features containing rich information for automatic tree species identification is an urgent problem.We used spectral sensors carried by unmanned aerial vehicles to take hyperspectral images,which covered an area of 6 hm^(2)of the 25-hm^(2)temperate mixed conifer-broadleaf plot in Changbai Mountain.Six canopy tree species,includingPinus koraiensis,Ulmus davidiana,Quercus mongolica,Fraxinus mandshurica,Populus ussuriensis,andTilia amurensis,were selected as the field labeled tree species.Real-time Kinematic Phase Difference (RTK) technology was used to accurately capture the position of those target tree species.In addition,visual interpretation of the image of the study area was performed using ArcGIS based on the forest re-inventory results in 2019.Three classification methods,including convolutional neural network,maximum likelihood and Mahalanobis distance,were used to analyze the automatic classification of canopy tree species.Our results showed that:(1) The overall accuracy and Kappa coefficient of tree species classification of convolutional neural network (99.85%,0.998) were better than maximum likelihood (89.11%,0.86) and Maharanobis distance method (79.65%,0.75);(2) Among the three classification methods,the classification accuracy of single dominant tree species was the highest when using the convolve neural network,and the highest classification accuracy ofP.koraiensis,U.davidiana,Q.mongolica,F.mandshurica,P.ussuriensisandT.amurensiswere100%,99.9%,99.9%,99.8%,99.8%and 99.5%,respectively.(3) Overall,convolutiona

关 键 词：天然林 高光谱 卷积神经网络法 最大似然法 马氏距离法 

分 类 号：S725.2[农业科学—林木遗传育种] S771.8[农业科学—林学]