不同森林植被的高光谱特征分析  

Hyperspectral Characteristic Analysis of Different Forest Vegetation

作　　者：潘佩芬[1] 杨武年[1,2] 戴晓爱[1,2] 郑菠 

机构地区：[1]成都理工大学地质灾害防治与地质环境保护国家重点实验室地学空间信息技术国土资源部重点实验室,四川成都610059 [2]成都理工大学地球科学学院,四川成都610059

出　　处：《遥感技术与应用》2013年第6期1000-1005,共6页Remote Sensing Technology and Application

基　　金：国家自然科学基金资助项目(41071265);2010年度高等学校博士学科点专项科研基金(2010512211 0006)

摘　　要：“生态水（层）”富水特征特殊，各信息指标参数难以用常规方法进行量化和反演，高光谱遥感由于其波段多、光谱信息丰富的优点为生态水（层）各信息指标参数的量化反演提供有效的数据源及方法。利用高光谱遥感技术进行植被分析时，其光谱特征的分析和敏感波段提取非常重要。针对“生态水”信息指标植被参数有关量化反演需要，对研究区部分典型植被叶片进行了光谱采集，利用微分方法对光谱数据进行处理，分析了不同植被叶片光谱的原始、一阶微分和二阶微分光谱曲线，从中提取差异大的波段区分不同植被。同时，采用距离统计分析方法对所选择的不同波段进行有效性验证。研究结果表明：虽然3种方法提取的波段有差异，但存在共同点；选择的光谱特征波段可有效地区分不同植被，在近红外波段尤为明显，分别是1814～1823nm、1874～1883nm和1890～1899nm附近。＂Eco-water（layer）＂ was provided with special watery features, whose parameters are difficult to be quantified and inverted by common methods,but hyperspectral remote sensing with plenty bands and a- bundant spectral information provide efficacious data source and method to invert the parameters of Eco- water. Using hyperspectral remotely sensing technology to analyze spectral characters and extracting sensi- tive bands of vegetation were very important. To meet the need of quantifying the eco-water＇s parameters, in this paper,the spectral information of partial vegetation leafs in the study area was collected,the spectral data was derived, the original, the first derived and the second spectral curves of different vegetation leafs were analyzed,and the more different bands of different vegetation were extracted to distinguish the vege- tations. Meanwhile, distance statistical method was to verify the validity of different bands. The results show that, there were some differences for extracted bands by three methods, but there were some same points,that is the extracted bands can effectively distinguish the different vegetations,especially the near- infrared bands which was 1 814～1 823 nm,1 874～1 883 nm and 1 890～1 899 nm.

关 键 词：“生态水” 信息指标参数 森林植被 高光谱特征 微分法 近红外 

分 类 号：Q149[生物学—生态学]