基于主成分分析的反射光谱在岩石学中的应用——以哈密地区为例  

Application of rock laboratorial reflectance spectra in Hami area based on principal component analysis.

作　　者：吴昀昭[1] 田庆久[2] 陈骏[1] 季峻峰[1] 杜凤兰[2] 

机构地区：[1]南京大学地球科学系,南京210093 [2]南京大学国际地球系统科学研究所,南京210093

出　　处：《岩石学报》2003年第4期761-766,共6页Acta Petrologica Sinica

基　　金：国土资源部项目:资源卫星应用系统研究与开发(编号20010203)的成果

摘　　要：本文首先总结并分析了岩矿常见光谱特征的波长位置及其产生原因。然后基于主成分分析技术对哈密42条岩石实验室光谱进行了定量分析。前两个主成分共包含了96.7％的信息量,代表了原始信息。第一主成分反映了岩石总体反射率大小,称为反照率因子;第二主成分反映了光谱曲线斜率变化,称为形状因子。使用这两个主成分可以区分该区主要岩类,达到了数据压缩和分类的目的。第三和第四主成分尽管所占信息很少,但反映了岩石蚀变信息。为了突出蚀变岩石光谱吸收特征,进而又对经过连续统去除后的光谱进行主成分分析,结果发现,在新生成的第三和第四主成分图上,4种矿石被清晰区分开来。Reflectance spectroscopy shows a wealth of information about minerals and rocks, and it has been widely used in the research of petrology. Firstly the absorption features of minerals and rocks in the Visible-near-infrared region is analyzed. Secondly Principal component analysis (PCA) is used to deal with 42 rock laboratorial reflectance spectra from Hami Area. The first principal component (PC1) which represents most of the data variability, is related to the albedo variation, and the second principal component (PC2), with 8% of the total variance, is related to the slope of the curves. Based on PC1 and PC2, most of the rocks are distinguished. It demonstrates that Principal component analysis can be used for data compression and classification. To enhance spectra absorption features, a second PCA phase is developed. Firstly the original data are processed by the continuum removed method, and then these worked data are further processed also by the principal component analysis. The rocks related to mineralization are recognised on the scatter plot of PC3 and PC4. It should be emphasized that these results show that the reflective properties of rock are different from the properties of petrologic classification.

关 键 词：主成分分析 反射光谱 分类 哈密 

分 类 号：P584[天文地球—岩石学]