基于红外光谱主成分分析的易混毛皮的鉴别研究
|
摘要
文章对60组牛毛皮和60组马毛皮红外光谱数据采用spss进行了主成分分析;以100组数据为建模样本(牛毛皮和马毛皮各50组)采用spss进行了多类判别分析,建立了典型判别函数和牛毛皮、马毛皮分类函数,并进行了回代验证;以20组数据验证样本(牛毛皮和马毛皮各10组)对所建立典型判别函数和牛毛皮、马毛皮分类函数进行了验证。研究结果表明:采用主成分分析能够有效降维,将原谱图2696个波长变量降至9个变量,新变量的累积贡献率可达到99.89%;建立的典型判别函数回代正确率达100%,验证正确率达100%;建立的牛毛皮、马毛皮分类函数回代聚类图聚类良好,验证聚类图聚类正确率达100%。
The principal component analysis was performed on 60 groups of cattle fur and 60 groups of horse fur infrared spectral data using SPSS. With 100 groups of data as modeling samples(50 groups of cattle fur and horse hair), multiple discriminant analysis was carried out by SPSS. The typical discriminant function, cattle fur and horse fur classification function were established, and the back generation verification was carried out. The canonical discriminant function and the classification function of cattle fur were verified by 20 sets of data validation samples(10 groups of cattle fur and horse fur). The results show that the principal component analysis can effectively reduce the dimension and reduce the original spectrum to 9 variables, and the cumulative contribution rate of the new variable can reach 99.89%. The correct rate of the typical discriminant function is 100%, the correct rate is 100%, and the clustering graph of the cattle fur and the horse hair skin classification function is good, and the accuracy of clustering graph clustering is 100%.
The principal component analysis was performed on 60 groups of cattle fur and 60 groups of horse fur infrared spectral data using SPSS. With 100 groups of data as modeling samples(50 groups of cattle fur and horse hair), multiple discriminant analysis was carried out by SPSS. The typical discriminant function, cattle fur and horse fur classification function were established, and the back generation verification was carried out. The canonical discriminant function and the classification function of cattle fur were verified by 20 sets of data validation samples(10 groups of cattle fur and horse fur). The results show that the principal component analysis can effectively reduce the dimension and reduce the original spectrum to 9 variables, and the cumulative contribution rate of the new variable can reach 99.89%. The correct rate of the typical discriminant function is 100%, the correct rate is 100%, and the clustering graph of the cattle fur and the horse hair skin classification function is good, and the accuracy of clustering graph clustering is 100%.
引文
[1]陈磊,徐有强,冯军芳,等.红外光谱和光学显微镜两步法鉴别皮革的研究[J].中国皮革,2016,45(11):1-4.
[2]李晓丽,胡兴越,何勇,等.基于主成分和多类判别分析可见-红外光谱水蜜桃品种鉴别新方法[J].红外和毫米波学报,2006,25(6):417-420.
[3]谢玮,严守雷,赵雯霖.基于果胶FTIR结合主成分分析鉴别不同品种莲藕[J].食品工业,2016,37(12):272-275.
[4]梁奇峰,陈秀玲,侯红娜.红外光谱结合主成分分析识别不同产地单枞茶[J].广东化工,2015,21(42):33-34.
[5]袁玉峰,陶站华,刘军贤,等.红外光谱结合主成分分析鉴别不同产地黄柏[J].光谱学与光谱分析,2011,31(5):1258-1261.
[6]罗庇荣,刘刚,时有明,等.FTIR结合主成分分析对杜鹃花植物鉴别分类研究[J].红外技术,2009,32(1):39-43.
[7]赵小蓉,胡宗智,顾彦.傅里叶变换红外光谱法识别分析四种天然皮革材质[J].理化检验-化学分册,2008,(44):543-547.
[8]周子立,蒋璐璐,谈黎虹,等.基于光谱技术鉴别机油品种的新方法[J].光学学报,2009,29(8):2203-2207.
[9]李智勇,匡纲要,郁文贤,等.基于高光谱图像主成分分量的小目标检测算法研究[J].红外与毫米波学报,2004,23(4):286-290.
[2]李晓丽,胡兴越,何勇,等.基于主成分和多类判别分析可见-红外光谱水蜜桃品种鉴别新方法[J].红外和毫米波学报,2006,25(6):417-420.
[3]谢玮,严守雷,赵雯霖.基于果胶FTIR结合主成分分析鉴别不同品种莲藕[J].食品工业,2016,37(12):272-275.
[4]梁奇峰,陈秀玲,侯红娜.红外光谱结合主成分分析识别不同产地单枞茶[J].广东化工,2015,21(42):33-34.
[5]袁玉峰,陶站华,刘军贤,等.红外光谱结合主成分分析鉴别不同产地黄柏[J].光谱学与光谱分析,2011,31(5):1258-1261.
[6]罗庇荣,刘刚,时有明,等.FTIR结合主成分分析对杜鹃花植物鉴别分类研究[J].红外技术,2009,32(1):39-43.
[7]赵小蓉,胡宗智,顾彦.傅里叶变换红外光谱法识别分析四种天然皮革材质[J].理化检验-化学分册,2008,(44):543-547.
[8]周子立,蒋璐璐,谈黎虹,等.基于光谱技术鉴别机油品种的新方法[J].光学学报,2009,29(8):2203-2207.
[9]李智勇,匡纲要,郁文贤,等.基于高光谱图像主成分分量的小目标检测算法研究[J].红外与毫米波学报,2004,23(4):286-290.