食品科学 ›› 2019, Vol. 40 ›› Issue (8): 300-305.doi: 10.7506/spkx1002-6630-20171124-305

• 安全检测 • 上一篇    下一篇

便携式葡萄专用可见-近红外光谱检测仪器开发与实验

肖 慧,孙 柯,屠 康,潘磊庆   

  1. 南京农业大学食品科技学院,江苏 南京 210095
  • 出版日期:2019-04-25 发布日期:2019-05-05
  • 基金资助:
    “十二五”国家科技支撑计划项目(2015BAD19B03)

Development and Application of a Specialized Portable Visible and Near-Infrared Instrument for Grape Quality Detection

XIAO Hui, SUN Ke, TU Kang, PAN Leiqing   

  1. College of Food Science and Technology, Nanjing Agricultural University, Nanjing 210095, China
  • Online:2019-04-25 Published:2019-05-05

摘要: 基于可见-近红外光谱技术,研发低成本、便携式葡萄专用多参数检测仪器,用于满足葡萄采后品质快速、无损的检测需求。本仪器选用凹面全息光栅搭配电荷耦合器件的光谱仪作为核心器件,用于获取样品400~1 100 nm的漫反射光谱数据;选用卤素灯作为稳定可靠的光源、低OH的Y型石英光纤作为光传输的可靠媒介,并设计可调型样品池满足不同大小和品种样品的需求,基于Windows系统采用C#撰写的软件性能稳定,便于模型的更新操作。以“美人指”、“白玉霓”两个葡萄品种进行实验,指标参数包含CIE L*a*b*、可溶性固形物含量及总酚含量。结果显示,本仪器基于最小二乘-支持向量机模型对2 个品种的a*值、可溶性固形物、总酚有较好的建模效果,“美人指”各指标的模型决定系数(Rc 2 )分别为0.91、0.94和0.90;“白玉霓”各指标的模型决定系数分别为0.96、0.99和0.95。最后,利用70 个非建模葡萄样本模型进行外部测试,结果表明“美人指”的a*值、可溶性固形物、总酚3 个指标的预测根均方误差分别为3.15、1.39 °Brix和0.24 g/kg;“白玉霓”3 个指标的预测根均方误差分别为0.78、1.56 °Brix和0.22 g/kg。结果表明,本仪器能完成对葡萄多个理化指标的建模预测,同时样品池的设计能够满足不同品种葡萄的需求。本研究为果蔬专用型近红外仪器的开发提供技术参考。

关键词: 可见-近红外光谱, 凹面全息光栅, 电荷耦合检测器, CIE L*a*b*, 可溶性固形物含量, 总酚

Abstract: In this study, a specialized portable low-cost visible and near-infrared (Vis-NIR) instrument for the evaluation of grape quality was designed to meet the need for fast and non-destructive detection in the grape industry. A holographic concave diffractive grating combined with a charge-coupled device was used as the core component of this system to obtain diffuse reflectance spectral data of grape samples in the wavelength range between 400 and 1 100 nm, and the combined application of a halogen light source with low-OH Y-shaped quartz optical fiber as a reliable optical transmission medium was proposed to provide a stable Vis-NIR light source. In order to meet the requirements of different varieties with fruits of different sizes, an adjustable sample cuvette with both internal and external threads was designed. The software, written in C# based on windows operating system, was stable and could facilitate model updating. Two Vitis vinifera cultivars, Manicure Finger and Ugni Blanc, were detected using this instrument. CIE L*, a* and b* values, total solid soluble content (SSC) and total phenolic (TP) content were taken into consideration in the development of predictive models using least squares support vector machine (LS-SVM). The results showed that for both cultivars a* value, SSC and TP could be well predicted by the proposed models with a coefficient of determination for calibration (Rc 2 ) of 0.91, 0.94 and 0.90, respectively. In the external validation using 70 unknown samples, the root mean square error of prediction (RMSEP) of a* value, SSC and TP were 3.15, 1.39 °Brix and 0.24 g/kg for Manicure Finger; and 0.78, 1.56 °Brix and 0.22 g/kg for Ugni Blanc, respectively. This study has confirmed the feasibility of this Vis-NIR instrument for grape quality detection and can provide a technical basis for developing specialized Vis-NIR instruments to detect fruit and vegetable quality.

Key words: visible and near-infrared (Vis-NIR) spectroscopy, holographic concave diffractive grating, charge-coupled device, CIE L*, a* and b*, total solid soluble content, total phenolic content

中图分类号: