Thesis Defense: Testing the Capability of Close-Range Photogrammetry to Document Outdoor Forensic Scenes With Skeletal Remains Using Mock Scenarios
June 15, 2018 @ 02:00 PM - 04:00 PM
Announcing the Final Examination of Kevin Gidusko for the degree of Master of Arts in Anthropology
More rigorous methodological protocols are needed to document outdoor forensic scenes containing skeletal remains. However, law enforcement protocols rarely provide specific guidelines for processing these scenes. Regardless, the need to preserve contextual information at crime scenes is of paramount importance and it is worth exploring new technological applications that will allow for better documentation. Close-range photogrammetry (CRP) is one option for outdoor scene documentation, more prominently utilized in archaeological contexts, that may provide forensic archaeologists with a tool to better document these scenarios via 3D modeling. To test the efficacy of CRP as documentation tool three mock scenarios representing common outdoor scenes were created using faux osteological material: a close scatter of osteological remains in a pine flatwood setting, a wide scatter of osteological remains in the same setting, and the partial excavation of skeletonized remains. Images were collected using a digital camera and processed using Agisoft Photoscan Professional. A series of variables were tested in successive iterations of data capture for each scenario to determine best practices for overall accuracy: camera images captured by hand versus fixed to a tripod, scale bar positioning, and number of images captured. Accuracy was determined via final root mean square error values and through a comparison between real-world to virtual measurements. Results show that CRP is a cost and time-effective method of documenting contextual data at a scene via the creation of 3D models and scaled orthomosaic images. This method is most useful for the documentation of excavations owing to the controlled and contrasted sub-surface in comparison to the subject material. The two scatter scenarios offered additional challenges due to the complexity of the ground covering, however models nonetheless provided accurate contextual detail and errors may be mitigated through proper data capture. There was little difference in the variables for image capture, scale bar placement, or number of images. Instead, the quality of images, image capture method, and post-processing operations proved to be more important. Due to the ease of use and the ability to convey best practices for data capture, the utilization of CRP for outdoor scene documentation is recommended as a valuable addition to current forensic documentation protocols. Future research should focus on the utilization of actual osteological material as a proxy for forensic scenarios as well as study the applicability of CRP to assist in documenting taphonomic modifications.
Committee in Charge: John J. Schultz (Chair), John Walker, Scott Branting