摘要
针对传统方法测定梨可溶性糖含量耗时长、成本高、步骤繁琐的问题,本研究提出一种通过光谱扫描梨果检测其可溶性糖含量的新方法。以河北省威县龙集梨园的‘新梨7号’为试材,利用光谱仪扫描果实获取其光谱反射率,使用蒽酮比色法测定梨可溶性糖含量,采用SG平滑滤波法、标准正态变量(SNV)和正交信号校正(OSC)对数据进行预处理,使用主成分分析(PCA)和连续投影算法(SPA)对光谱数据进行降维处理,利用多元线性回归(MLR)、偏最小二乘法(PLSR)和支持向量机回归(SVR)建立预测模型,最终选择使用OSC+PCA+SVR建立基于光谱反射率的梨可溶性糖含量的预测模型,其训练集RC^(2)=0.85,RMSE=2.21,测试集RV^(2)=0.78,RMSE=1.48,实现了对梨可溶性糖含量高效、快速、低成本、无破坏性测定。
Conventional methods of soluble sugar detection were time consuming and costly due to tedious steps.To improve the experimental effeciency,a new method based on spectral scanning was proposed in this paper for detecting soluble sugar content of pear.The pear variety'Xinli No.7'was used as the experimental materials which were picked from the longji pear orchard in Wei county of Hebei Province.The spectral reflection factor was obtained by scanning the pear with a spectrometer and soluble sugar contents were examined using Anthrone-Sulfuric acid colorimetry.The data was preprocessed by SG smoothing filtering method,Standard Normal Variables(SNV)and Orthogonal Signal Correction(OSC)followed by Principal Component Analysis(PCA)and Continuous Projection Algorithm(SPA)to reduce the dimension of spectral data.Then the prediction model of the soluble sugar content was established by Multiple Linear Regression(MLR),Partial Least Squares Regression(PLSR)and Support Vector Machine Regression(SVR).Finally,OSC+PCA+SVR were adopted to establish the prediction model of pear soluble sugar content with its training set as RC^(2)=0.85,RMSE=2.21 and testing set as RV^(2)=0.78,RMSE=1.48.The new method was fast and low-cost with high-efficiency in non-destructive determination of pear soluble sugar content.
作者
张铭
王文芳
郜一川
吴悦菊
王国英
ZHANG Ming;WANG Wenfang;GAO Yichuan;WU Yueju;WANG Guoying(College of Horticulture/Pear Engineer and Technology Research Center of Hebei,Hebei Agriculture University,Baoding 071000,China)
出处
《河北农业大学学报》
CAS
CSCD
北大核心
2022年第1期79-84,共6页
Journal of Hebei Agricultural University
基金
国家梨产业技术体系建设专项(CARS-28-09)
河北农业大学科研发展基金计划项目—梨园营养元素供给和分配研究.
关键词
梨
可溶性糖
光谱
机器学习
数据降维
Pear
Soluble sugar
Spectrum
Machine learning
Data dimensionality reduction
