|
|
|
|
Application of neural network technique to the forensic analysis of infrared spectra of car paint samples was examined. Infrared spectral data of 222 paint samples were classified by the Cohonen self-organizing feature map network. As a result, 222 spectra were classified into six groups. According to the result of this classification, the two-step neural network system was trained. Test was carried out using paint samples gathered at real car accident places. Finally, it was concluded that the development of the practical system for the forensic analysis of car paint is hopeful.