The analysis of colors is omnipresent in forensics. While their use for macroscopic evidence such as soil sees a large amount of research, their correct handling is more challenging when observed under a microscope. Either for hairs or textile fibers, their color (from melanin, dyes or for birefringence analysis) is crucial in being able to compare between fibers found at a crime scene. Typically, color is quantified using a color space, such as RGB, XYZ, L*a*b*, and L*u*v*. Unfortunately, color is very subjective and often difficult to quantify accurately, even using digital detectors. Finding true matches of colors has been limited by factors including illumination, optics transmission function, and in many cases the human eye. The colors produced by the birefringence of fibers under analysis using polarized light microscopy (PLM) are known to provide quick information on the fiber identification. Nonetheless, the exact determination of the accurate colors is still either subjective to the operator or the correct calibration of the software. In this thesis, calibration of the color response of a polarizing light microscope was performed using a UV-Vis broadband spectrometer to characterize its illuminant and camera. All the color images obtained with the camera were then transformed from the common RGB encoding to L*a*b* color space using MATLAB. An empirical Michel-Lévy chart was also designed using a quartz wedge in order to calibrate retardation with empirical L*a*b* color values, also as a teaching tool for the UCF undergraduate class "Forensic Microscopy", and with a qualitative comparison with common textile fibers. The discriminatory power and uncertainty of color perception of the microscope camera and of human vision were measured and will be discussed.
| Identifer | oai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:etd2020-2197 |
| Date | 01 January 2022 |
| Creators | Feild, Olivia |
| Publisher | STARS |
| Source Sets | University of Central Florida |
| Language | English |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Electronic Theses and Dissertations, 2020- |
Page generated in 0.0018 seconds