苹果梨单果重的近红外无损检测研究

Research on the Near-infrared Nondestructive Detection on Single Fruit Weight of Apple-pear

[目的]探究应用近红外技术进行相关性状分析。[方法]以苹果梨为试材,利用近红外透射光谱仪扫描完整苹果梨的短波近红外透射光谱,光谱经预处理后,采用偏最小二乘法(PLSR)建立单果重的数学模型。[结果]光谱经归一化处理后建立的模型稳定性最好,相关参数为:SEP=18.01,RC=0.70,RMSEC=18.68,PC=5。[结论]利用近红外技术可以预测苹果梨的单果重,若增加建模集样本数,模型的相关系数可能有所升高。

[ Objective] The aim was to explore applying near-infrared technology in the analysis of relative characters. [ Method] With apple-pear as tested material, the near-infrared transmission spectrometer was used to scan short wave near-infrared transmission spectrum of whole apple-pear and the spectrum was pretreated, and then the mathematic model of single fruit weight was established by partial least square method （PLSR）. [Result] After the spectrum was standardized, the stability ＇of established model was best. The relative par,ameters were as follows： SEP = 18.01, Rc =0.70, RMSEC = 18.68 and PC = 5. [ Conclusion] The single fruit weight of apple-pear could be predicted by near-infrared technology. If the sample number in modeling set is increased, the correlation coefficient of model may have some increase.