中文摘要：

使用4波段（550nm、650nm、766nm和850nm）便携式作物反射光谱测量仪对泰农18型冬小麦分蘖状态进行自动监测与建模，通过分析植被指数与分蘖数的相关关系实现了对分蘖数的建模预测。首先利用仪器获得小麦冠层在4个波段的反射信号，计算对应波段的作物冠层反射率，经校正后计算得到OSAVI、MSAVI、SAVI、EVI2、TVI、NDGI、NDVI、RVI和DVI9种多波段组合的植被指数。然后分析以上9种植被指数与小麦分蘖数之间的相关关系，确定了可用于该类型小麦分蘖状态监测和评价的植被指数类型。2013--2014年在山东省淄博市和桓台县开展了田间试验，计算了不同氮素水平下泰农18型小麦返青期和起身期分蘖数以及其两个生育期分蘖数与9种植被指数之间的相关系数，OSAVI（650，850）指数与返青期茎蘖数相关系数最高，决定系数最高为0．85，均方根误差为118．93；EVI2（650，850）指数与起身期茎蘖数相关系数最高，决定系数最高为0．84，均方根误差为73．04；以上试验结果表明，在冬小麦返青期和起身期利用OSAVI（650，850）和EVI2（650，850）两种植被指数可以快速预测小麦分蘖状态，可为田间精细管理提供科学依据。

英文摘要：

The number of tillers has a significant effect on the winter wheat field management and the prediction of winter wheat yield. However, the traditional manual counting method of the tiller counting is inefficient. With the development of spectral technology and the application of low altitude remote sensing technology in agriculture, a method was provided for monitoring the number of tillers and growth of the winter wheat by calculating crop canopy reflectance and vegetation index. A 4-waveband crop monitor with spectral reflectance was used to carry on the experiment （Tainong 18）. The instrument can obtain the crop canopy reflecting signals at 550 nm, 650 nm, 766 nm and 850 nm simultaneously. After that the crop canopy reflectance was first calculated and then nine vegetation indexes： OSAVI, MSAVI, SAVI, EVI2, TVI, NDGI,NDVI, RVI and DVI, were also calculated. The relationship between the tillering of winter wheat and each index of nine vegetation indexes was analyzed in both regreening and erecting stages. In regreening stage, the correlation between OSAVI（650,850）and tillers was the highest （R^2 is 0. 85, RMSE is 118.93 ）, while in erecting stage, the correlation between EVI2 （650,850）and tillers was the highest （R^2 is 0.84, RMSE is 73.04）. The results of the test showed that there was a significant relationship between the winter wheat tillers and the two vegetation indexes. This may help the development of the instrument for winter wheat tillers counting based on canopy spectral reflection. The conclusions can be used in rapid predicting of wheat tillering and giving suggestions to field precision management.