Search prefilters for mid‐infrared absorbance spectra of clear coat automotive paint smears using stacked and linear classifiers

Abstract

By using stacked partial least squares classifiers and genetic algorithms for feature selection and classification, it is demonstrated that search prefilters can be developed to extract investigative lead information from clear coat paint smears. The results obtained in this study also show that identifying specific wavelengths or wavelet coefficients in IR spectral data is superior to identifying informative wavelength windows when applying pattern recognition techniques to IR spectra from the paint data query (PDQ) database when differentiating paint samples by assembly plant. Search prefilters developed using specific wavelengths or wavelet coefficients outperformed search prefilters that utilized spectral regions. Clear coat paint spectra from the PDQ database may not be well suited for stacking as there are few spectral intervals that can reliably distinguish the different sample groups (i.e., assembly plants) in the data. The information contained in the IR spectra about assembly plant may not be highly compartmentalized in an interval, which also works against stacking. The similarity of the IR spectra within a plant group and the noise present in the IR spectra may also be obscuring information present in spectral intervals.

The effect of environmental degradation on the characterisation of automotive clear coats by infrared spectroscopy

“…Several publications by Lavine [14][15][16][17][18][19][20]. However, as the samples utilised in this study originated solely from North American manufacturing plants, the methodology developed currently has limited applicability within an international context, and additionally relies upon a matching IR spectrum being present within the PDQ database.…”

Forensic analysis of automotive paints using a pattern recognition assisted infrared library searching system: Ford (2000–2006)

“…However, these search prefilters [4][5][6][7][8][9][10] were limited to a single manufacturer To perform spectral library searching, the cross correlated spectra were divided into windows. For the forward search, the original spectra were also divided into three regions.…”

Characterisation and classification of automotive clear coats with Raman spectroscopy and chemometrics for forensic purposes

“…Cluster analysis and PCA performed on the resultant data revealed three distinct groupings; however, no correlation could be made between the groupings as a function of the make, model or year of the corresponding vehicles, thus limiting the applicability of the statistical model for procuring investigative leads . Lavine and co‐workers have demonstrated the utility of pattern recognition techniques in conjunction with the PDQ database to successfully extract information from FTIR spectra in the database to aid in the generation of investigative leads . A recent study conducted by the authors used attenuated total reflectance (ATR) IR spectroscopy to characterise clear coats of over 100 vehicles and then used a combination of PCA and LDA to examine the diversity in the sample population.…”