基于遥感图像的地物识别需要大量地物波谱数据的支持,而现有地物波谱库中收集到的同一地物的波谱,由于其测量尺度和方法的差异,波谱也存在很大差异.以冬小麦为例,首先介绍了材料波谱、端元波谱和像元波谱这3种不同尺度波谱的概念,并以实验数据分析了不同测量尺度下波谱的差别,以此说明波谱尺度转换的必要性.然后利用物理模型和统计模型建立不同测量尺度下的波谱转换关系.分别验证了SAILH模型和线性光谱混合模型在波谱转换中的精度.研究表明,在大尺度上采用统计模型,在小尺度上采用非线性的物理模型可以解释不同尺度观测植被波谱之间的差异.

Remote sensing mi agery provides vast information about the land surface, and the spatialdistribution of land cover types by classification could be obtained. Moreover, the spectrum of land surface objects is useful to mi prove the accuracy of mi age classification. However, the spectrum of the same objectmay be differentwhen they aremeasured at differentmeasuring scale andwith differentmethod. Forexample, the spectrum ofwinter-wheatextracted from LandsatTM andmeasured in the field are different. So it is mi portant to study the scale effectand scalingmethod on the spectrum at differentmeasuring scale.In this paper, we took thewinterwheatas example, and selected Shunyi region in Beijing as our study area. Firstly the definition of three-scale spectrum was explained, then we analyzed the discrmi ination using the measuring data to highlight the mi portance of the scale transformation of spectrum. The collected data included: fieldmeasured spectrum of leaf and canopy and the hyper-spectrum high-resolution remote sensed mi ageryOMIS, LandsatTM andMODIS data. We compared the winter-wheat spectrums and calculated the slope of“red-edge”at differentmeasuring scales. We also studied the scaling-up method of the spectrum, and the physicalmodel (SAILH) and statisticalmodel (Linearmixing model) were used to describe the relationship between the spectrum at differentmeasuring scales. SAILH is a typical radiation transfermode,l which can be used to smi ulate the canopy spectrum by taking the leaf spectrum, some structural parameters and environmental variables as inputs. In this expermi ent, the input parameters were acquired with high accuracy, so the error of smi ulation result is very smal:l 8. 45%. Linear mixing model was used to describe the relationship between endmember spectrum and pixel spectrum. The resolution ofMODIS mi agery (visible and infrared bands) is 250m, whichwas taken as pixel spectrum, and the endmember one can be gotbymultiplemethods, herewe adopt twomethods: Broadmanmethod fromMODIS mi agery and aggregation one from TM mi agery. The unmixing results were compared and analyzed, and the linearmixingmodelwas validated.Through the spectrum data in the study area, the winter-wheat spectrum of lea,f canopy and OMIS mi agery is\ndifferent and the character reflecting the plantgrowing status is also varied. As for the scalingmethod, we found statistical models and physicalmodels are fit on the three research scales respectively. However, the endmember selectingmethod\nfrom the mi agery alsoneedsmore mi provement, andmore statisticalmodelsorcouplingphysicalmodels should be explored\nin the furtherwork.