植物生化组分遥感探测的光谱统计参数比较  

Comparison of Spectral Statistical Parameters about Plant Biochemical Concentration Determination with Remote Sensing

作　　者：阮伟利[1] 颜春燕[1] 牛铮[1] 

机构地区：[1]中国科学院遥感应用研究所遥感信息科学重点实验室,北京100101

出　　处：《遥感技术与应用》2003年第4期233-236,共4页Remote Sensing Technology and Application

基　　金：中国科学院知识创新工程重大项目(KZCX1-SW-01);国家重点基础研究发展规划项目(G2000077900);国家自然科学基金资助项目(40271086)资助。

摘　　要：利用统计分析方法,分析了叶片叶绿素和氮含量与其光谱特性的统计关系,分别建立了log(1/R)、反射率一阶导数(FDS)、对波深中心归一化(BNC)及对波深面积归一化(BNA)的光谱形式与叶片叶绿素和氮含量的统计方程,并对这4个指标的性能进行了比较和评价。结果表明采用描述光谱吸收特征的参数BNC和BNA,能够提高遥感估算植物生化组分含量的效果,特别是对氮的估算,其预测值与测量值的相关系数R达到0.960。Based on analyzing the concentration of two biochemical components, including total chlorophyll and nitrogen, we establish the statistical relationships between the concentration and inverse\|log spectra(log 1/R)through the stepwise multiple regression method. So do the relationships between the concentration and several transformations of reflectance such as standard first derivative reflectance spectra (FDS), absorption band depths following continuum removal and normalization against band depth at the center of the absorption feature (BNC), absorption band depths following continuum removal and normalization against the area of the absorption feature(BNA).The results show very different performances among these four spectral statistical parameters ,including the selections of bands, the results of stepwise multiple regression and the effects of the statistical equations. This just illustrates uncertainty when using statistical methods to determine plant biochemical concentration. However, we still get good prediction performance for chlorophyll and nitrogen with BNC and BNA. Especially, R value of the correlation between estimated and observed nitrogen concentration is 0.960. We get these results on a spectra/biochemical data set from dried and ground Douglas Fir. Further research work should be done to determine foliar biochemical concentration from field, airborne and spaceborne spectra.

关 键 词：生化组分 波深中心归一化 波深面积归一化 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]