Honors Program
Date of this Version
Spring 5-2019
Document Type
Thesis
Citation
Douglas, A. 2019. Analysis of Error within Forensic Measurements and Photogrammetry Programs. Undergraduate Honors Thesis. University of Nebraska - Lincoln.
Abstract
Digital images and photography have been considered commonplace in forensic science since the late 1990’s. Photogrammetry is a tool used in forensic science to measure objects within photographs that contain a scale or programs that are able to measure items in 3D images. Items at crime scenes often need to be measured and forensic scientists may not know what the best option is for their situation. This research will be conducted to show which type of measurement technique is most accurate, calculated against NIST traceable measurements, at various sizes of objects. The measurement techniques that will be tested are standard, commercially available, hand scale measurements, Photoshop, GIMP, IrfanView, and FARO. The hypothesis of the experiment is that FARO will have the lowest percent error for all sizes of objects. It was discovered that Photoshop, GIMP, and IrfanView had the lowest percent errors for small objects (0.43%), while for small objects, FARO had the highest percent error (10.61%). For medium objects, Photoshop again had the lowest percent error (0.44%) but standard measurements had the highest error (1.74%). For large objects, FARO had the lowest percent error (0.28%), and IrfanView had the highest percent error (1.85%). However, most methods had percent errors equal to or less than the acceptable five percent. Some methods were preferred for different sizes of objects, but all programs were useful as long as the limitations were considered.
Comments
Copyright Amy Douglas 2019.