基于傅里叶变换红外光谱的葡萄果苗鉴别研究

发布时间：2018-03-31 04:24

本文选题：傅里叶变换红外光谱　切入点：葡萄果苗　出处：《云南师范大学》2015年硕士论文

【摘要】：葡萄是我国云南省主要的经济作物之一,以其香甜味美、营养价值高而受大众喜爱。葡萄种类繁多、种质资源丰富,一般很难通过外观对其区分。本文运用傅里叶变换红外光谱(FTIR)技术结合化学计量学统计分析方法对葡萄果苗进行分类鉴别研究。葡萄果苗的红外光谱主要由蛋白质、多糖和脂类等吸收带组成;各种葡萄果苗的红外光谱差异不大;但在1800~750 cm-1波数范围内差异显著,用1800~750cm-1范围二阶导数光谱进行主成分分析和系统聚类分析;前3个主成分累计方差贡献率达到94.9%,分类正确率达100%;系统聚类分析的正确率达到96%,能够很好地鉴别五个不同品种的葡萄果苗。将傅里叶变换红外光谱(FTIR)与Morlet小波主成分分析和偏最小二乘法判别分析(PLS-DA)相结合对红地球和皇家秋天进行了研究。原始光谱整体相似,仅在1800~750 cm-1范围有微小差异。选取该波段第7尺度的Morlet小波系数和原始光谱进行主成分分析;以及对该波段进行20尺度的一维连续小波变换,并将变换结果用于偏最小二乘法判别(PLS-DA)。结果表明Morlet小波主成分分析和PLS-DA都能很好地鉴别两个品种的葡萄果苗,其中Morlet小波主成分分析的正确率为100%;PLS-DA在隐含潜变量为9时,红地球和皇家秋天果苗的分类正确率均达到100%。研究结果表明,傅里叶变换红外光谱结合化学计量学统计分析方法能够准确地分类鉴别不同品种的葡萄果苗,为葡萄果苗的分类鉴别研究提供了快速和准确的方法。
[Abstract]:Grape is one of the main cash crops in Yunnan Province of China. It is loved by the public for its sweet taste and high nutritional value. There are many kinds of grapes and abundant germplasm resources. It is difficult to distinguish grape fruit by its appearance. In this paper, Fourier transform infrared spectroscopy (FTIR) and chemometrics were used to classify and identify grape fruit seedlings. The absorption bands of polysaccharides and lipids were not different from each other, but the difference was significant in the range of 1800,750 cm-1 wave number. Principal component analysis and systematic cluster analysis were carried out by using second-derivative spectra in 1800~750cm-1 range. The cumulative variance contribution rate of the first three principal components was 94.9%, the classification accuracy was 100%, and the correct rate of systematic cluster analysis was 96%, which could be used to identify the grape fruit seedlings of five different varieties. Fourier transform infrared spectroscopy (FTIR) and Morlet wavelet were used to identify grape fruit plantlets. Principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) were used to study red earth and royal autumn. There is only a slight difference in the range of 1800 ~ 750 cm-1. The Morlet wavelet coefficients and the original spectrum of the seventh scale of the band are selected for principal component analysis, and the one-dimensional continuous wavelet transform of 20 scales is carried out on the band. The results show that Morlet wavelet principal component analysis and PLS-DA can identify the grape fruit plantlets of two varieties well, and the correct rate of Morlet wavelet principal component analysis is 100% and the latent variable is 9%. The classification accuracy of red earth and royal autumn fruit seedlings is 100. The results show that Fourier transform infrared spectroscopy combined with chemometrics statistical analysis method can accurately classify and identify grape fruit seedlings of different varieties. It provides a rapid and accurate method for the classification and identification of grape fruit seedlings.
【学位授予单位】：云南师范大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：S663.1;TN219

【参考文献】