论文部分内容阅读
通过2个小麦品种4个氮素水平的大田小区试验,及光谱仪传感器在冬小麦群体侧面水平测定不同叶层反射光谱,分析不同叶层光谱特征参量与冠层氮素分布、籽粒蛋白质含量的定量关系,建立籽粒蛋白质含量预测的分层光谱模型。结果表明,2个小麦品种冠层内氮素分布特点不同,普通蛋白质含量小麦京冬8号上、下叶层对不同氮素处理反应敏感,而高蛋白质含量小麦中优9507反应不敏感。京冬8号开花期叶层氮素含量梯度(ΔLNC)与籽粒蛋白质含量(GPC)极显著相关,在本文选择的6个光谱特征参量中,叶层比值植被指数梯度ΔRVI[670,890]与籽粒蛋白质含量(GPC)极显著相关;中优9507乳熟期籽粒蛋白质含量和叶层光谱参量梯度ΔRVI[670,890]极显著负相关。籽粒蛋白质含量和分层光谱参量梯度ΔRVI[670,890]统计模型,决定系数高于传统垂直测定模型,京冬8号模型RMSE为0.7500,中优9507模型RMSE为0.6461,在不同生育时期建立的分层光谱模型可以对不同小麦籽粒蛋白质含量进行预测。
The field reflectance spectra of four wheat cultivars with four nitrogen levels and the spectral reflectance spectra of different leaf layers were measured by spectrometer sensors. The quantitative relationship between spectral characteristic parameters and canopy nitrogen distribution and grain protein content was analyzed. , A stratified spectral model of grain protein content was established. The results showed that the nitrogen distribution characteristics of the two wheat cultivars were different. The upper and lower leaves of common wheat protein Jingdong 8 were sensitive to different nitrogen treatments, while the high protein content wheat Zhongyou 9507 was not sensitive. The leaf nitrogen content gradient (ΔLNC) at flowering stage of Jingdong 8 had a significant correlation with grain protein content (GPC). Among the six spectral parameters selected in this study, leaf gradient RAVI [670,890] and grain protein (GPC). The grain protein content of Zhongyou 9507 at the milk ripening stage was significantly and negatively correlated with the leaf layer spectral parameter gradient ΔRVI [670,890]. Grain protein content and stratified spectral parametric gradient ΔRVI [670,890], the coefficient of determination was higher than that of the traditional vertical model, the RMSE of Jingdong 8 model was 0.7500, the RMSE of Zhongyou 9507 model was 0.6461, the stratification established at different growth stages Spectral models predict the protein content of different wheat grains.