摘要
采用近红外高光谱成像系统对3类不同产地的圆枣进行判别分析,快速鉴别圆枣产地。应用近红外高光谱获取3类圆枣样本的光谱数据,对光谱采用标准正则变换(SNV)方法预处理,从原始光谱数据中提取特征波长,并建立全波段与特征波段下的线性判别模型来判别3类圆枣的产地。结果表明,近红外光谱结合线性判别法对圆枣产地鉴别的特征波段模型可有效替代全波段模型,模型准确率均大于99%,为实现农产品产地鉴别和自动分类提供理论依据。
In this study, the near-infrared hyperspectral imaging system was used for discriminant analysis of three kinds of round jujubes came from different origins, identified the origin of round jujubes rapidly. Obtained spectral data of three kinds of round jujubes by near-infrared hyperspectraL Used the Standard Normal Variate transformation （SNV） to pretreatment spectral, extracted characteristic wavelength from original spectral data and established linear discriminant model of full-band and characteristic-band to distinguish the origin of round jujubes. The results showed that near-infrared spectroscopy combined with linear discriminant analysis method to identify the origin of the round jujubes, the model of characteristic- band could replace full-band model effectively replace full-band, which accurately rate of the model greater than 99%, providing a theoretical basis for the realization of automatic identification and classification of agricultural oriclin.
出处
《食品科技》
CAS
北大核心
2015年第6期344-347,共4页
Food Science and Technology
基金
2011年度宁夏回族自治区科技攻关计划项目(2011HZF05J01)
关键词
近红外光谱
圆枣
PCA
SNV
LDA
判别
near-infrared spectral
round jujube
PCA
SNV
LDA
discrimination
