【摘要】 蛋白质含量和粗淀粉含量是水稻品质的两个重要指标。通过大田和室内试验,测定了5个品种、3个供氮水平处理的水稻冠层和稻穗在成熟过程中不同时期的高光谱反射率及稻谷的高光谱,分析了稻穗稻谷的粗蛋白及粗淀粉含量。结果表明,稻穗粗蛋白及粗淀粉含量与冠层光谱及其一阶导数光谱在某些波段达到显著相关、与稻穗粉末样光谱及其一阶导数光谱在某些波段也达到显著相关,其中决定系数R2约0.7;稻谷粗蛋白含量与粗淀粉含量呈负相关,它们与稻谷光谱及其一阶导数光谱在某些波段达到极显著相关水平。表明能用高光谱方法估算稻穗和穗谷的粗蛋白及粗淀粉含量,为水稻品质遥感监测提供依据。更多还原

【Abstract】 The contents of protein and starch in rice are two of the indices to evaluate the nutrition and taste quality of rice. Normal determinations of their contents by chemical methods are of high expense and time consuming. The hyperspectral reflectances of the canopy and panicle of 5 rice varieties were measured by a ASD FieldSpec Pro FRTM in field and indoor under 3 nitrogen application levels in mature process, and that of paddy measured too. The contents of crude protein and crude starch in panicle and paddy were determined. The contents of crude protein and crude starch in panicle were significantly correlated to the canopy spectra and their first derivative spectra at some wave bands, and also significantly correlated to the spectra of panicle powder and their first derivative spectra at some wave bands. The squared multiple correlation coefficients (R2 ) were about 0.7. The crude starch content was negative correlative to the crude protein contents in paddy. They were very significantly correlated to the paddy spectra and their first derivative spectra at some wave bands. This indicated that it was feasible for estimating the contents of crude protein and crude starch in panicle and paddy by hyperspectral remote sensing. It would provide a basis for monitoring rice quality by remote sensing.更多还原