论文部分内容阅读
研究了利用近红外漫反射无损检测海沃德猕猴桃可溶性固形物含量和pH的方法。以45个湘西海沃德猕猴桃为标准样本,采集1 000~1 800 nm范围的近红外光光谱,光谱采用9点滑动窗口平滑处理、一阶微分和多元散射校正分别进行预处理,然后采用偏最小二乘算法(PLS)、主成分回归算法(PCR)和多元线性回归算法(MLR)分别建立预测模型。经比较,采用一阶微分的PLS模型预测性能最好。
The method of non-destructive detection of soluble solids content and pH of Hayward kiwifruit using near-infrared diffuse reflectance spectroscopy was studied. 45 Xiangxi Hayward kiwifruit as a standard sample collected near 1 000 ~ 1 800 nm near-infrared spectrum, the spectrum using 9 sliding window smoothing, first-order differential and multiple scattering correction were pre-treatment, and then use the partial Least squares algorithm (PLS), principal component regression (PCR) and multiple linear regression (MLR) were used to establish the prediction model. By comparison, the first-order PLS model predicts the best performance.