论文部分内容阅读
目的检验自行搭建的半透射光谱采集平台检测水果中可溶性固形物含量的可行性,并比较不同光谱采集方式对光谱模型的影响。方法以红富士苹果为检测对象,光谱采集平台中的USB2000+光谱仪采集半透射光谱数据,AntarisⅡFT-NIR光谱仪采集漫反射光谱数据,同标准法检测得到的苹果可溶性固形物含量建立偏最小二乘(PLS)模型,并结合不同的预处理方式优化近红外光谱模型。结果比较发现采用半透射的光谱采集方式优于漫反射方式。半透射光谱采用平滑处理后模型预测性能最佳,对样本预测得到相关系数为0.937,均方根误差为0.517。结论自行搭建的光谱采集平台可行,为今后检测水果的光谱采集方式提供参考。
Objective To test the feasibility of semi-permeable spectrum acquisition platform constructed by ourselves to detect the content of soluble solids in fruits and to compare the effects of different spectral acquisition methods on spectral models. Methods The samples collected from Fuji apple were collected, and the semi-transmissive spectra were collected with USB2000 + spectrometer. The data of diffuse reflectance spectra were collected with Antaris Ⅱ FT-NIR spectrometer. The content of soluble solids in apple was determined by PLS ) Model, combined with different pretreatment methods to optimize the near-infrared spectroscopy model. The results show that the use of semi-transmissive spectral acquisition method is superior to the diffuse reflection method. After the semi-transmissive spectrum is smoothed, the prediction model is the best, the correlation coefficient is 0.937 and the root mean square error is 0.517. Conclusion The self-built spectral acquisition platform is feasible and provides a reference for the future detection of fruit spectral acquisition methods.