ABSTRACT
Clothing articles are important pieces of evidence in criminal, search and rescue, and search and recovery investigations. Hyperspectral remote sensing of clothing will be an important tool for supporting such investigations in the near future. This study investigated over 300 items of clothing that varied in fabric type, texture, color, and pattern. Clothing items were analyzed using an ASD FieldSpec 4 High Resolution spectroradiometer with a contact probe attachment. Of the clothing items analyzed, there were 141 having endmember fabrics (100% single fabric type composition): 89 were cotton, 39 were polyester, 5 were wool, 1 was cashmere, 3 were acrylic, 1 was leather, and 3 were rayon. The remaining 164 clothing items were various fabric blends. Spectral features relating to different fabric types exhibit sufficient differences that allow them to be discriminated from the surrounding environment, as well as from one another in many, but not all, cases. Cotton and polyester, in particular, two of the most widely-used fabrics, exhibit numerous features in the near infrared (NIR) and shortwave infrared (SWIR) that would allow them to easily be distinguished from geologic materials in the environment such as rocks and soil. Plant based fibers, especially cotton, possess similar reflectance features to vegetation owing to their cellulose content. Outdoor aging experiments were conducted for 19 weeks on selected fabrics. Although significant changes were observed in aged garments, the variability observed in the reflectance of the aged garments does not support the derivation of a metric for aging, at least over the relatively short time scale of this effort. Results from this study should support numerous forensic efforts globally for non-destructive investigation of clothing items in the field and in lab settings with a spectroradiometer, enhance the potential for remote sensing searches, and in the future, potentially documenting crime scenes with hyperspectral imaging.
