Abstract

When a firearm is fired emerge waste discharge known as Gunshot residue (GSR), these residues are particles produced by combustion. In this paper, using a hyperspectral imaging system in the near-infrared (NIR), hyperspectral data cubes are obtained from scanned fabric samples with gunshot residues. These fabrics have been penetrated by a bullet, where the distribution of waste released by the firearm is not homogeneous. Three samples of fabrics with GSR were used. The principal objective was to obtain particular spectral characteristics of gunshot residues. To obtain these characteristics, the reflectance spectrum for three selected pixel sets from three images of the fabric samples with GSR was built in. Each selected pixel contains only information from the residue compound and not from the fabric. The fabrics were illuminated at the same lamp power. One fabric without GSR was also characterized for comparison between the waste particles and the clean fabric itself. To compare the results, the statistical average was applied to each of the generated reflectance spectra for every set of selected pixels. The Experimental results have shown that analysis of the GSR fabric samples using hyperspectral imaging systems in the NIR spectrum allowed to characterize and compared the fabric samples with and without GSR. In particular, it is possible to establish a clear spectral difference between the fabrics with and without GSR in the wavelengths between 950 and 1350 nm.