近红外光谱对磨盘柿货架期的定性判别  

Qualitative Identification of Mopan Persimmon Shelf-life by Near Infrared Spectroscopy

作　　者：王丹[1] 鲁晓翔[1] 张鹏[2] 陈绍慧[2] 李江阔[2] 

机构地区：[1]天津商业大学生物技术与食品科学学院,天津市食品生物技术重点实验室,天津300134 [2]国家农产品保鲜工程技术研究中心,天津市农产品采后生理与贮藏保鲜重点实验室,天津300384

出　　处：《食品科学》2013年第22期185-188,共4页Food Science

基　　金："十二五"国家科技支撑计划项目(2012BAD38B01);天津市重点科技攻关项目(11ZCKFNC01900);天津市农业科学院院长基金项目(12004)

摘　　要：通过近红外漫反射光谱技术对磨盘柿的货架期进行了定性判别，比较不同光谱预处理方法和不同波段，发现在1100～1800nm波段范围内，采用二阶导数结合标准正常化和去散射（SNVD）处理的光谱预处理方法最好。其判别模型的正确率达到95．4％，预测准确率达到93．3％。因此，近红外光谱技术对磨盘柿货架期的检测具有应用价值。Near-infrared spectroscopy was used for qualitative identification of Mopan persimmon shelf-life. Different spectral pre-treatment methods and different spectral ranges for the shelf-life of Mopan persimmon were compared. The best results of spectral pretreatment were obtained by using secondary derivative model combined with standard normal variate and detrend （SNVD） in the wavelength range of 1100-1800 nm. The percentage of correct classification of samples by using the developed discriminant model was 95.4% with a forecast accuracy of 93.3%. Therefore, near-infrared spectroscopy is a useful technique to predict the shelf life of Mopan persimmon.

关 键 词：近红外光谱 磨盘柿 货架期 判别 预测 

分 类 号：S665.2[农业科学—果树学]