PROGRESS IN DEVELOPING A HYPERSPECTRAL REMOTE SENSING LIBRARY AND SOFTWARE TOOL FOR FINDING PERSONS IN THE ENVIRONMENT

Hyperspectral remote sensing continues to emerge as a multidisciplinary tool. One area of intersection is forensics and geology for locating humans in the environment. This is a complicated subject area owing to the natural variability of geologic materials, human materials and anthropogenic confuser materials. Reflective spectra from 350 to 2500 nm have been collected on >25 human subjects, 300 clothing articles, >50 geologic substrates, and >75 apatite and bone samples. Additionally spectra from preserved animal specimens, plastics and other items have been collected. Extensive materials characterization such as X-ray diffraction, scanning electron microscopy as well as item documentation such as composition, brand and make, provides potential users with a high level of contextual information for interpretation.

Findings thus far indicate demonstrable spectral differences occur between the vast majority of human skin, human hair, clothing and geologic materials. Spectral variation amongst human skin appears to modestly correlate to age. Human and animal hair are similar. Vegetation and cotton-based clothing commonly have shortwave infrared (SWIR) features that overlap owing to cellulose in both materials. Synthetic fiber-based clothing is commonly distinct from geologic materials and vegetation, even when wet with water. Although useful for search and rescue, it appears that common red hunter’s vests may provide challenges in detection using hyperspectral imaging owing to overlap of some features in the near infrared. Integration with existing USGS hyperspectral data is ongoing in order to produce a first generation tool for finding humans in the environment. Furthermore, meta-analysis of data is ongoing to evaluate causes of variation amongst and between spectra. This project has search and rescue implications for finding lost hikers and missing persons as well as criminal justice implications for crime scene investigation and locating fugitives and victims.