论文部分内容阅读
为检测苹果品质并依据擦伤进行分级,研发了一个基于拉曼光谱的实时无损自动检测分类虚拟仪器分级系统样机。采用一台Nicolet傅氏变换拉曼光谱仪进行苹果擦伤光谱检测。测试集和训练集的苹果光谱用WinDAS的典型变量分析(CVA)和主成分分析法(PCA)进行分类处理。分析得出的模型经UNEQ分类检验,’马氏平方图’和χ2检验结果该分类模型。其次,应用LabVIEW设计苹果虚拟仪器分级控制系统,并制作了样机。试验结果表明拉曼光谱分析能用于苹果擦伤无损检测和类别确定;虚拟仪器分级系统能对苹果进行准确分级处理。
In order to test the quality of apples and grading them according to the abrasion, a prototype of real-time non-destructive automatic classification system based on Raman spectroscopy was developed. A Nicolet Fourier transform Raman spectrometer was used to detect the apple scratch spectrum. Apple spectra of the test set and training set were categorized using WinDAS’s CVA and PCA. The analysis of the model by the UNEQ classification test, ’Markov square’ and χ2 test results of the classification model. Secondly, the application of LabVIEW design Apple virtual instrument grading control system, and made a prototype. The experimental results show that Raman spectroscopy can be used for nondestructive testing and classification of apple scratches. The virtual instrument grading system can accurately classify apples.