论文部分内容阅读
提出一种利用近红外光谱技术进行鸡蛋种类快速、无损鉴别的新方法。选用7500—4000cm-1的光谱,采用标准正态变量变换(SNV)后作主成分分析(PCA),选取前10个主成分作为模型输入,种类类别作为模型输出,分别建立了3种鸡蛋种类的线性判别法(LDA)和支持向量机(SVM)鉴别模型,所建模型均能较好的对鸡蛋种类进行鉴别,SVM模型效果优于LDA模型,其预测集正确识别率达97.44%。结果表明,近红外光谱技术可用于鸡蛋种类的快速、无损鉴别。
A new method for rapid and nondestructive identification of egg types using near infrared spectroscopy is proposed. The spectrum of 7500-4000cm-1 was selected. The principal component analysis (PCA) was performed after using standard normal variable transformation (SNV). The top 10 principal components were selected as the model inputs and the species categories were used as model outputs. Three kinds of egg types (LDA) and support vector machine (SVM) discriminant models. All the models were able to identify the egg types better. The SVM model was better than the LDA model, and the correct recognition rate of the prediction set was 97.44%. The results show that NIR spectroscopy can be used for rapid and nondestructive identification of egg types.