首页|Hierarchical classification and matching of mid-infrared spectra of paint samples for forensic applications
Hierarchical classification and matching of mid-infrared spectra of paint samples for forensic applications
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NSTL
Elsevier
A novel fast and automatic methodology for the hierarchical classification and similarity matching of mid-infrared spectra of paint samples based on the principles of Soft Independent Modelling of Class Analogy (SIMCA) and on the definition and properties of the Mahalanobis distance is here proposed. This approach was tested in a so-called market study (i.e., targeting products largely accessible to the general public and conceived for a considerably wide range of usages) conducted across the surroundings of the city of Lille, in France, and has permitted not only to successfully achieve the chemical characterisation of most of the analysed samples but also to discover specific commonality patterns among specimens sharing the same chemical features.
ClassificationMatchingHierarchical methodsSoft independent modelling of class analogy(SIMCA)Hotelling'sT2PaintsGREEN SPRAY PAINTSINFRARED-SPECTROSCOPYLEAST-SQUARESIDENTIFICATIONREDMICROSPECTROPHOTOMETRYSIMCAFTIR
NSTL
Elsevier
