食品科学 ›› 2020, Vol. 41 ›› Issue (6): 298-303.doi: 10.7506/spkx1002-6630-20190125-325

• 安全检测 • 上一篇    下一篇

基于深度信念网络的近红外光谱鉴别莲子粉掺假

胡仁伟,俞玥,倪明龙,俞娇,周俊伟,朱诚,李占明   

  1. (1.中国计量大学生命科学学院,浙江 杭州 310018;2.湖州职业技术学院,浙江 湖州 313000;3.广东食品药品职业学院,广东 广州 510520)
  • 出版日期:2020-03-25 发布日期:2020-03-23
  • 基金资助:
    浙江省自然科学基金项目(LQ17C200002);湖州市公益性研究项目(2018GZ28); 福州市科技计划项目(2018K0049);广东省医学科学技术研究基金项目(B2018171); 广东食品药品职业学院自然科学研究项目青年项目(2016YZ033)

Identification of Lotus Seed Flour Adulteration Based on Near-Infrared Spectroscopy Combined with Deep Belief Network

HU Renwei, YU Yue, NI Minglong, YU Jiao, ZHOU Junwei, ZHU Cheng, LI Zhanming   

  1. (1. College of Life Sciences, China Jiliang University, Hangzhou 310018, China; 2. Huzhou Vocational and Technical College, Huzhou 313000, China; 3. Guangdong Food and Drug Vocational College, Guangzhou 510520, China)
  • Online:2020-03-25 Published:2020-03-23

摘要: 为快速鉴别莲子粉真伪,利用近红外光谱技术对莲子粉掺杂进行鉴别。基于已知利用支持向量机(support vector machine,SVM)对光谱数据分类的结果,在未知样品类别的情况下使用基于深度信念网络(deep belief network,DBN)进行判别。结果表明,当训练集数目达到600时,SVM模型对掺入不同比例各类其他作物粉的平均识别率达到98%;基于DBN模型能够有效识别掺杂了各类其他作物粉的莲子粉,极个别掺杂比例的平均识别率在96%左右。采用DBN算法避免了当前深层神经网络易陷入局部最优和无大量标签样本的情况。近红外光谱技术结合DBN为农产制品掺假的快速检测提供了新的尝试。

关键词: 莲子粉, 掺假, 近红外光谱, 支持向量机, 深度信念网络

Abstract: We developed a near-infrared (NIR) spectroscopic method for rapidly identifying the adulteration of lotus seed flour. The spectral data of lotus seed flours mixed with different other crop flours at different levels were classified using support vector machine (SVM). On this basis, deep belief network (DBN) was employed for discrimination of lotus seed flours with unknown adulterants. The results showed that the average recognition rate of the SVM model for adulteration of the other crop flours was 98% (training numbers ≥ 600). Based on the DBN model, lotus seed flour adultered with the other crops could be effectively identified and the average recognition rate for very few adulteration levels was about 96% (training numbers ≥ 600). The DBN model avoided the disadvantages of the current deep neural networks such as local optimization and not using massive labelled samples. NIRs combined with DBN provides a new alternative for the rapid detection of adulteration of agricultural products.

Key words: lotus seed flour, adulteration, near-infrared spectroscopy, support vector machine, deep belief network

中图分类号: