可见/近红外光谱半透射法检测苹果中可溶性固形物含量

Non-destructive analysis of soluble solids content in apple by VIS/NIR semi-transmittance

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 The feasibility of self-built platform by semi-transmittance mode acquired spectrum in detection of the soluble solid content(SSC) in apples was studied,and the difference of two spectral acquisitions on the model of spectral dataset was also compared.Methods Fuji apples were used as detected object.The semi-transmitted and diffuse reflectance spectral datasets were acquired by the USB2000+ and AntarisⅡ FT-NIR spectroscopes respectively,and were combined with the apples’ SSC determined by national standard method,to establish the partial least squares(PLS) models,which was optimized by several pretreatments.Results After comparison of the models,it was found that the spectrum acquired by semi-transmittance mode was better than diffuse reflectance mode,while the performance of the model showed the best after spectral smoothing,with the Rp of 0.937,as well as the RMSEP value of 0.517 for external samples.Conclusion The study showed that the self-built platform is useful for detecting apple fruit and can be a good reference for spectral acquisition of non-destructive determination in future.