CAJ | 학술논문

利用红边参数反演作物参数是定量感研究的一个热点,红边参数中红边位置与作物生化组分强相关,为监测作物胁迫提供了一个非常敏感的指标.准确估测植被绿素含量,对研究森林健康和胁迫、森林生产力的估计,碳循环的研究有着重要的意义.介绍几种红边位置算法,并对这些算法及其应用进行了比较,通过选取红边位置的不同敏感波段来估测植被片绿素含量.经室内光谱获取片的光谱数据,采用一阶光谱导数法、平滑处理后一阶光谱导数法、线性四点内插法、五次多项式拟合法四种算法处理光谱数据,获得红边位置变量,并与绿素含量进行拟合,构建估测木荷片绿素含量的回归模型.结果表明:各种算法获取的红边位置变量所构建的回归模型估测绿素含量是可行的;五次多项式拟合法估算精度是最高的,其获取红边位置计算相对复杂;线性四点内插法估算精度次之,但计算较简便.
이용홍변삼수반연작물삼수시정량감연구적일개열점,홍변삼수중홍변위치여작물생화조분강상관,위감측작물협박제공료일개비상민감적지표.준학고측식피록소함량,대연구삼림건강화협박、삼림생산력적고계,탄순배적연구유착중요적의의.개소궤충홍변위치산법,병대저사산법급기응용진행료비교,통과선취홍변위치적불동민감파단래고측식피편록소함량.경실내광보획취편적광보수거,채용일계광보도수법、평활처리후일계광보도수법、선성사점내삽법、오차다항식의합법사충산법처리광보수거,획득홍변위치변량,병여록소함량진행의합,구건고측목하편록소함량적회귀모형.결과표명:각충산법획취적홍변위치변량소구건적회귀모형고측록소함량시가행적;오차다항식의합법고산정도시최고적,기획취홍변위치계산상대복잡;선성사점내삽법고산정도차지,단계산교간편.
Red edge parameters are widely used to inv ersely deduce crop parameters in quantitative remote sensing studies. Among them, the red edge position, as a very sensitive indicator for monitoring crop stress, is strongly correlated with crop biochemical components. Accurate estimation of the chlorophyll content of vegetation is of importance for studies on forest health, stress, and productivity estimation, as well as carbon cycle. In this article, several algorithms of red edge position were introduced firstly, their applications were compared, and the leaf chlorophyll content of vegetation was estimated through selecting its different sensitive bands. Then leaf spectral data from indoor spectra were extracted, four algorithms were used (first -order derivative spectrometry, first -order derivative spectrometry after smoothing, four point interpolation, and quintic polynomial fitting) to process spectral data and obtain red edge position variables. Finally the obtained variables were used to fit the chlorophyll content, and various regression models of these algorithms for estimating leaf chlorophyll content were established. The results show that all these established models are feasible to estimate chlorophyll content. Among them, the quintic polynomial fitting method is most accurate, but highly complex in obtaining the red edge position while the four point linear interpolation is next to it in accuracy, but simpler.