基于近红外光谱技术的叶面药液浓度检测  

Measurement of Pesticide Concentration on the Leaf Based on Vis/NIR Spectroscopy

作　　者：邱白晶[1] 殷磊[1] 

机构地区：[1]江苏大学现代农业装备与技术省部共建教育部重点实验室,镇江212013

出　　处：《农业机械学报》2012年第9期197-201,208,共6页Transactions of the Chinese Society for Agricultural Machinery

基　　金：江苏省高校自然科学研究重大项目(10KJA470007);高等学校博士学科点专项科研基金资助项目(20093227110008);江苏省农机三项工程项目(NJ2010-15;NJ2010-12);江苏高校优势学科建设工程资助项目(苏财教[2011]8号);江苏省农业装备与智能化高技术研究重点实验室资助项目(BM2009703)

摘　　要：提出了一种应用近红外光谱技术快速检测叶面药液浓度的方法。采用漫反射测量方式获取了叶面药液的近红外光谱。选用标准偏差归一化、三点滑动平均滤波和一阶导数为最优组合预处理。通过7种波段方案的对比,得出最优波段为350～1 900 nm。采用偏最小二乘法建立了叶面药液质量浓度与光谱反射率的定量分析模型。其预测集相关系数为0.994,预测均方根误差为0.039。结果表明,利用近红外光谱技术检测叶面药液浓度具有实际指导意义。A method based on Vis/NIR spectral analysis technology was applied to the rapid measurement of pesticide contentration on the leaf. Near infrared spectral data of pesticide concentration on the leaf was acquired by adopting diffuse reflectance spectroscopy. Normalization of standard deviation, moving average smoothing with the segment size 3 and the first derivative combination was the optimal combination pretreatment method. From the seven schemes in contrast, the optimal hands for 350 ~ 1 900 nm were obtained. Partial least squares method was used to develop the quantitative analysis model for pesticide concentration with diffuse reflectance spectroscopy. The correlation coefficient between the prediction values and the truth values in the prediction set was 0. 994, the root mean square error of cross-validation of the prediction set was 0. 039. The result indicated that it is feasible to measure pesticide concentration on the leaf based on the near infrared （NIR） spectroscopy, and the accurate actual pesticide concentration on the leaf can be obtained after spraying pesticides.

关 键 词：叶面喷雾 药液浓度 漫反射 偏最小二乘法 近红外光谱 

分 类 号：O657.33[理学—分析化学] S481.9[理学—化学]