论文部分内容阅读
为提高激光诱导击穿光谱技术(Laser-induced breakdown spectroscopy,LIBS)对鲜肉品种的识别率,采用支持向量机结合主成分分析算法辅助LIBS技术对鲜肉品种进行识别。对鲜肉切片用载玻片压平,采用LIBS技术对鲜肉组织(猪肉、牛肉和鸡肉)表面进行光谱数据的采集,每种鲜肉采集150幅光谱并进行随机排列,取前75幅光谱作为训练集建立模型,后75幅作为测试集测试建模结果。研究选取K、Ca、Na、Mg、Al、H、O等元素的49条归一化谱线数据进行主成分分析,并用所得数据建立支持向量机分类模型。结果表明,通过主成分分析降维,输入变量从49个优化减少到18个,模型建模速度从88.91 s降至55.52 s,提高了支持向量机的建模效率;并使预测集的平均识别率提高到89.11%。本研究为激光诱导击穿光谱技术在鲜肉品种快速分类领域提供了方法和数据参考。
In order to improve the recognition rate of fresh meat varieties by Laser-induced breakdown spectroscopy (LIBS), Support Vector Machine (SVM) and Principal Component Analysis (LPCS) were used to identify the fresh meat varieties with LIBS. The slices of fresh meat were flattened with glass slides and the surface data of fresh meat (pork, beef and chicken) were collected by LIBS. 150 spectra were collected and randomly arranged for each fresh meat. The first 75 spectra As a training set to establish the model, after 75 as a test set test modeling results. 49 normalized spectral data of K, Ca, Na, Mg, Al, H, O and other elements were selected for principal component analysis, and the support vector machine classification model was established with the obtained data. The results showed that the principal component analysis reduced the input variables from 49 to 18 and the model modeling speed dropped from 88.91 s to 55.52 s, which improved the modeling efficiency of SVM. Rate increased to 89.11%. This study provides a method and data reference for the rapid classification of fresh meat species by laser-induced breakdown spectroscopy.