基于核驱动模型参数反演的作物株型遥感识别

传统的单一方向遥感只能得到地面目标一个方向的辐射量,缺乏足够信息来同时推断像元的波谱和空间结构.多角度的反射光谱对于植被结构特征的估算及类型鉴别比垂直光谱具有明显的优越性.将多角度观测数据和核驱动模型结合,构建了基于核参数各向同性核所占比例(fiso),几何光学核所占比例(fgeo)和体散射核所占比例(fvol)的植被结构参数敏感指数(SPEI)来进行作物株型的遥感识别,并将SPEI与结构散射指数(SSI)和归一化差异参数指数(NDFI)等对紧凑型品种京411和披散型品种中优9507的实测多角度冠层方向反射率数据进行了核参数的反演,不同指数对作物株型结构表现出的敏感性依次为:SPEI＞SSI＞NDFI.SPEI较以往学者构建的指数对不同株型品种遥感识别具有更好的敏感性,该方法可以在其他作物上拓展应用,具有很好的理论基础和应用前景.

Crop Geometry Identification Based on Inversion of Semiempirical BRDF Models

Investigations have been made on identification of erective and horizontal varieties by bidirectional canopy reflected spectrum and semi-empirical bidirectional reflectance distribution function(BRDF)models.The qualitative effect of leaf area index(LAI)and average leaf angle(ALA)on crop canopy reflected spectrum was studied.The structure parameter sensitive index(SPEI)based on the weight for the volumetric kernel(fvol),the weight for the geometric kernel(fgeo),and the weight for constant corresponding to isotropic reflectance(fiso),was defined in the present study for crop geometry identification.However,the weights associated with the kernels of semi-empirical BRDF model do not have a direct relationship with measurable biophysical parameters.Therefore,efforts have focused on trying to find the relation between these semi-empirical BRDF kernel weights and various vegetation structures.SPEI was proved to be more sensitive to identify crop geometry structures than structural scattering index(SSI)and normalized difference f-index(NDFI),SPEI could be used to distinguish erective and horizontal geometry varieties.So,it is feasible to identify horizontal and erective varieties of wheat by bidirectional canopy reflected spectrum.