摘要
提出一种用近红外光谱技术快速鉴别原料肉和注水肉的新方法。首先以原料肉和注水肉为原料,利用近红外光谱仪测定其漫反射光谱曲线,然后选取二阶导数+25点平滑方法进行预处理,再应用主成分分析结合人工神经网络技术对其进行判别分析。结果表明,前5个主成分的累计贡献率已达99.626%,以前5个主成分作为人工神经网络的输入,对应的肉种类(原料肉与注水肉)作为输出,建立了一个三层BP神经网络模型,模型对建模集109个样本的鉴别率为91.74%,对预测集30个样本的鉴别率为90%。说明利用近红外光谱分析技术对原料肉注水进行快速鉴别是可行的。
A new method for the discrimination of raw meat and water-injected meat by means of near infrared spectroscopy（NIRS） was developed.First,raw meat and water-injected meat were collected.Using Near-infrared Spectroscopy to scan the samples and get the spectrum data.The principal component analysis（PCA） and A three-layer back-propagation neural network（BP-ANN） was developed for qualitative discrimination,which had been preprocessed with the second derivative and 25 point smoothing.Results show that the accumulative reliabilities of the first fifth components were more than 99.626 %,and chose ANN-BP as further research method.The first five components were then applied as ANN-BP inputs and the values of the type of meat were applied as the outputs.A three-layer back propagation neural network model was developed for classification.Finally,the result indicated the distinguishing rate of 109 calibration samples is 91.74 % and the distinguishing rate of 30 unknown test samples is 90 %.This model is reliable and practicable.So the present paper could offer a new approach to the fast discrimination of water-injected meat.
出处
《食品研究与开发》
CAS
北大核心
2012年第5期118-120,128,共4页
Food Research and Development
基金
陕西生猪产业科技创新体系基金(K336020902)
公益性行业(农业)科研专项经费项目(3-45)
西北农林科技大学校青年学术骨干支持计划资助
关键词
近红外光谱
原料肉
注水肉
主成分分析
BP人工神经网络
near infrared spectroscopy
raw meat
water-injected meat
principal component analysis（PCA）
BP-artificial neural network（BP-ANN）
作者简介
杨志敏（1985-），女（汉），硕士研究生，主要从事畜产品深加工的研究
通信作者：丁武（1971-），男（汉），副教授，博士，主要从事畜产品加工和食品安全方向的研究
