基于机载PHI高光谱数据的森林优势树种分类研究

发布时间：2018-04-11 07:41

本文选题：高光谱数据 + PHI　；参考：《国土资源遥感》2017年02期

【摘要】：近年来,高光谱遥感在林业方面的应用越来越广泛,尤其在分类方面居多。但机载PHI高光谱数据通常用于农业病虫害监测、海洋悬浮物颗粒监测等,在林业方面的应用较少。以湖北省荆门市东宝区为研究区,以机载PHI高光谱遥感数据为数据源,对森林优势树种进行了分类研究。首先采用独立成分分析法(independent component analysis,ICA)对裁剪后的PHI数据进行降噪,并利用自适应波段选择法(adaptive band selection,ABS)进行降维,再采用归一化植被指数(normalized difference vegetation index,NDVI)区分林地与非林地,最后利用支持向量机法(support vector machine,SVM)进行森林优势树种监督分类。研究结果表明,分类精度可达80.70%,Kappa系数达到0.75;分块处理PHI数据以及采用NDVI区分林地与非林地,对于减弱"同物异谱"和"异物同谱"现象有较好的作用;ABS与SVM相结合的分类方法,较适用于PHI数据在树种识别方面的应用探索,具有重要意义。
[Abstract]:In recent years, hyperspectral remote sensing has been widely used in forestry, especially in classification.However, airborne PHI hyperspectral data are usually used to monitor agricultural diseases and insect pests and marine suspended matter particles, but are rarely used in forestry.Based on the airborne PHI hyperspectral remote sensing data, the forest dominant tree species were classified in Dongbao District, Jingmen City, Hubei Province.The independent component analysis (ICA) method was used to reduce the noise of the PHI data, and the adaptive band selection method was used to reduce the dimension. Then the normalized difference vegetation index (NDVI) was used to distinguish the forest land from the non-forest land.Finally, support vector machine (SVM) was used to monitor and classify forest dominant tree species.The results show that the classification accuracy can reach 80.70 and the Kappa coefficient can reach 0.75. The classification method of combining PHI data with SVM can be used to reduce the phenomenon of "isomorphism" and "foreign body isospectral" by using NDVI to distinguish woodland and non-forest land.It is of great significance to explore the application of PHI data in tree species identification.
【作者单位】：中国林业科学研究院资源信息研究所;
【基金】：高分辨率对地观测系统重大专项项目(编号:30-Y20A37-9003-15/17) 国家自然科学基金青年科学基金项目“机载激光雷达探测森林冠层高度的机理模型研究”(编号:41201334) 国家高技术研究发展计划(863计划)子课题“全球林业定量遥感专题产品生产体系(二)”(编号:2013AA12A302)共同资助
【分类号】：S771;TP751

【相似文献】