Description of original award (Fiscal Year 2012, $439,842)
New techniques for pattern recognition assisted infrared (IR) library searching of the Paint Data Query (PDQ) automotive paint database will be developed to determine the make, model, and year of an automobile from an unknown paint sample recovered at a crime scene. Modern automotive paints have a thin color coat which on a microscopic fragment, is often too thin to obtain accurate chemical information. The small size of the fragment also makes it difficult to accurately compare it with manufacturers paint color standards. Since adhesion between paint layers is very strong, both primer paint layers are often transferred during a collision if the clear coat and color coat layers are also transferred. As primer and clear coat paint layers are usually unique to the automotive manufacturing plant where these layers were applied, combining chemical information obtained from the Fourier Transform (FT) IR spectra of the two primer layers and from the clear coat layer should make it possible to rapidly and accurately identify the make, model, and specify even certain years of manufacture of an automobile from its paint system alone. The development of search prefilters and library searching algorithms for the PDQ database, which is a major thrust of this research proposal, is needed to extract investigative lead information from clear coat and primer layer paint smears. An added advantage of using this approach to identify paint samples is an increase in accuracy because spectra from the entire database are searched. Information derived from the proposed pattern recognition searches will serve to quantify the general discrimination power of original automotive paint comparisons encountered in casework, and to succinctly communicate the significance of the evidence to the courts. Addressing these concerns is a direct response to Recommendation 3 of the National Academies February 2009 report, Strengthening Forensic Science in the United States: A Path Forward.. To maintain relevancy of the newly designed pattern recognition techniques, the analysis of additional paint samples will be simultaneously undertaken to populate the database with samples from production years where there are insufficient data. A correction algorithm to allow attenuated total reflectance (ATR) spectra to be searched using the IR transmission spectra of the PDQ database will also be developed as part of the proposed research. Currently, ATR is a widely used sampling technique in FT-IR spectroscopy because little expertise is required on the part of the user. For this reason, forensic laboratories are using this sampling technique with increasing frequency to collect FT-IR spectra of automotive paints. Since the optics involved in ATR are quite different from those used in transmission experiments, the infrared spectrum of a sample obtained by ATR will exhibit differences that prevent direct comparisons with its transmission counterpart. The proposed correction algorithm to convert transmission spectra in the PDQ library to ATR spectra will address ATR distortions issues such as the relative intensities and broadening of the bands, and the introduction of a wavelength shift at lower frequencies, which prevent library searching of ATR spectra using archived transmission data. This algorithm is crucial for the scientists who rely on the chemical information contained in the clear coat and primer layers when attempting to identify a motor vehicle from a sample of paint.