摘要：本实验提出了一种用近红外光谱技术建立桑品种快速检测模型新方法，首先 用近红外线光谱仪对典型的三个桑树品种的桑叶进行光谱采集，获取桑叶在 900-1700nm 不同波长下的近红外光谱图，并生成数据。在偏最小二乘判别分析 方法（pls-da）下再结合四种不同的预处理方法进行品种鉴别。结果表明，有的 预处理方法只对特定品种有较高的识别率，而对有些品种识别率较差，总体准确 度不高。四种预处理方法相比较，1st Derivation –SNV-Mean center 预处理方法对数据的处理结果总体识别率最好，且不存在只对特定品种有较高的识别率，识别 情况较平均，但品种识别准确率总体达到 85％以上，达到预期效果。说明近红 外光谱技术对于桑树品种有鉴别作用，为桑树的品种鉴别提供了一种新方法。
关键字：桑树；近红外光谱；预处理；识别率
Abstract：In this study, a new method for rapid detection of mulberry varieties was established by near-infrared spectroscopy. First, the mulberry leaves of three typical mulberry trees were collected by near-infrared spectroscopy to obtain the mulberry leaves at 900-1700 nm Near infrared spectroscopy, and generate data..The partial least squares discriminant analysis method (pls-da) under the combination of four different pretreatment methods was used for identification of species The results showed that some pretreatment methods only had higher recognition rate for specific varieties, while the recognition rate of some varieties was poor and the overall accuracy was not high. Compared with the four pretreatment methods, the 1st Derivation-SNV-Mean center pretreatment method has the best recognition rate of the data, and there is no higher recognition rate only for specific species, and the recognition is more average, but the variety Recognition accuracy rate of 85% overall, to achieve the desired results. The results show that the near infrared spectroscopy has a new method for the identification of mulberry varieties.
Keywords: mulberry;Near Infrared Spectroscopy;Pretreatment;Recognition rate
目 录
第一章 绪论 1
1.1 桑品种的经济效益 1
1.1.1 桑品种概述 1
1.1.2 不同桑品种效益区别 1
1.2.1 形态特征鉴别法 3
1.2.2 DNA 鉴定技术体系 4
1.2.3 基于 DNA 序列分析的条形码技术 4
1.2.4 基于 PCR 的分子鉴定技术 5
1.2.5 色谱指纹图谱 5
1.3 近红外光谱法 7
1.3.1 多元散射校正（MultiplicativeScatteringCorrection，MSC） 8
1.3.2 标准正态能量变换（StandardNormalVariateTransformation，SNV） 8
1.3.3 偏最小二乘法判别分析（PLS-DA） 8
1.4 近红外光谱分析技术的应用 8
1.4.1 近红外光谱在农业领域的应用 9
1.4.2 近红外光谱在食品领域的应用 10
1.4.3 近红外光谱在医药领域的应用 10
1.5 展望与期待 11
第二章 实验材料与方法 近红外光谱法建立桑品种快速检测模型:http://www.youerw.com/shengwu/lunwen_203689.html