| Summary: | Attenuated total reflectance infrared spectroscopy was employed to characterise the clear coats from a collection of automotive paint samples representing 130 vehicles, encompassing a range of Australian and international vehicle manufacturers. Principal component analysis revealed nine distinct classes, which were associated with the country of vehicle manufacture. Further statistical analysis identified variations in the samples from Australian and Japanese vehicles, which ultimately provided information regarding the manufacturer. Linear discriminant analysis gave excellent differentiation between the classes, with 100% of the calibration and test set samples being correctly classified. This ability to rapidly provide information concerning the vehicle origin and manufacturer will facilitate the procuring of investigative leads from questioned paint samples located at crime scenes. Although presented here in an Australian context, the strategy employed is universal and by extension could be utilised in other jurisdictions if they were to generate statistically significant data sets.
|
|---|