基于高光谱技术的覆盖保鲜膜菠菜货架期预测研究
发布时间：2019-08-15 09:00
【摘要】：保鲜膜能提高果蔬保水性,隔绝外界细菌侵染,延长货架期。为了准确估测覆盖保鲜膜果蔬品质的优劣,对其货架期进行预测具有重要意义。应用高光谱技术结合化学计量学方法对同等贮藏条件下覆膜新鲜菠菜叶片的货架期进行了预测。先采集五个不同贮藏时间下75盘共300片菠菜样本在可见-近红外(VisNIR,380~1 030nm)与近红外(NIR,874~1 734nm)波段的高光谱数据,然后测定不同贮藏时间下菠菜叶片叶绿素含量。提取300片覆膜菠菜叶片的平均光谱(200个为建模集,100个为预测集)后,对建模集光谱进行主成分分析(principal component analysis,PCA),发现不同贮藏期内叶片光谱数据在前3个主成分空间有一定的聚类。根据建模集光谱信息与预先赋予的不同贮藏期虚拟等级分别建立偏最小二乘判别分析(partial least squares discriminant analysis,PLS-DA)模型,得到预测集样本的贮藏期总的判别准确率分别为83%(Vis-NIR)和81%(NIR)。表明,高光谱技术结合化学计量学方法能够实现对新鲜菠菜货架期的分类和预测,为消费者正确评价覆盖保鲜膜的菠菜品质提供了理论指导,也为后期果蔬货架期检测仪器的开发提供了技术支持。
[Abstract]:Fresh-keeping film can improve the water retention of fruits and vegetables, isolate the infection of external bacteria, and prolong the shelf life. In order to accurately estimate the quality of fruits and vegetables covered with fresh-keeping film, it is of great significance to predict the shelf life of fruits and vegetables covered with fresh-keeping film. The shelf life of fresh spinach leaves covered with film under the same storage conditions was predicted by hyperspectral technique combined with chemometrics. The hyperspectral data of 75 spinach samples in visible-near-infrared (VisNIR,380~1 030nm) and near-infrared (NIR,874~1 734nm) bands were collected at five different storage times, and then the chlorophyll content of spinach leaves was measured at different storage times. After extracting the average spectrum of 300 spinach leaves covered with film (200 as modeling set and 100 as prediction set), the principal component analysis (principal component analysis,PCA) of the modeling set spectrum was carried out. It was found that the leaf spectral data had certain clustering in the first three principal component spaces during different storage periods. According to the spectral information of the modeling set and the virtual grades given in advance for different storage periods, the partial least squares discriminant analysis (partial least squares discriminant analysis,PLS-DA) models are established respectively. the total discriminant accuracy of the predicted set samples is 83% (Vis-NIR) and 81% (NIR)., respectively. The results show that hyperspectral technology combined with chemometrics can realize the classification and prediction of shelf life of fresh spinach, which provides theoretical guidance for consumers to correctly evaluate the quality of spinach covered with fresh-keeping film, and also provides technical support for the development of shelf life testing instruments for fruits and vegetables in the later period.
【作者单位】： 浙江大学生物系统工程与食品科学学院;
【基金】：国家自然科学基金项目(61273062) 国家高科技研究发展计划“863”项目(2013AA102301)资助
【分类号】：O657.3;TS255.7
本文编号：2526898
[Abstract]:Fresh-keeping film can improve the water retention of fruits and vegetables, isolate the infection of external bacteria, and prolong the shelf life. In order to accurately estimate the quality of fruits and vegetables covered with fresh-keeping film, it is of great significance to predict the shelf life of fruits and vegetables covered with fresh-keeping film. The shelf life of fresh spinach leaves covered with film under the same storage conditions was predicted by hyperspectral technique combined with chemometrics. The hyperspectral data of 75 spinach samples in visible-near-infrared (VisNIR,380~1 030nm) and near-infrared (NIR,874~1 734nm) bands were collected at five different storage times, and then the chlorophyll content of spinach leaves was measured at different storage times. After extracting the average spectrum of 300 spinach leaves covered with film (200 as modeling set and 100 as prediction set), the principal component analysis (principal component analysis,PCA) of the modeling set spectrum was carried out. It was found that the leaf spectral data had certain clustering in the first three principal component spaces during different storage periods. According to the spectral information of the modeling set and the virtual grades given in advance for different storage periods, the partial least squares discriminant analysis (partial least squares discriminant analysis,PLS-DA) models are established respectively. the total discriminant accuracy of the predicted set samples is 83% (Vis-NIR) and 81% (NIR)., respectively. The results show that hyperspectral technology combined with chemometrics can realize the classification and prediction of shelf life of fresh spinach, which provides theoretical guidance for consumers to correctly evaluate the quality of spinach covered with fresh-keeping film, and also provides technical support for the development of shelf life testing instruments for fruits and vegetables in the later period.
【作者单位】： 浙江大学生物系统工程与食品科学学院;
【基金】：国家自然科学基金项目(61273062) 国家高科技研究发展计划“863”项目(2013AA102301)资助
【分类号】：O657.3;TS255.7
【相似文献】
相关期刊论文 前3条
1 赵娟;彭彦昆;赵松玮;宋育霖;;基于高光谱技术检测苹果外观缺陷[J];食品安全质量检测学报;2012年06期
2 彭彦昆;李永玉;赵娟;单佳佳;;基于高光谱技术苹果硬度快速无损检测方法的建立[J];食品安全质量检测学报;2012年06期
3 ;[J];;年期
相关会议论文 前1条
1 蒋桂英;李少昆;官春云;陈金湘;;高光谱技术在农业上的应用(综述)[A];21世纪作物科技与生产发展学术讨论会论文集[C];2002年
相关硕士学位论文 前1条
1 金灵;基于高光谱技术的宁夏滩羊肉内部品质检测方法研究[D];宁夏大学;2014年
，本文编号：2526898
本文链接：https://www.wllwen.com/kejilunwen/huaxue/2526898.html
教材专著
