Abstract
Photogrammetry from images captured by terrestrial cameras and manned aircraft has been used for many years to model objects, create scale diagrams and measure distances for use in traffic accident investigation and reconstruction. Due to increasing capability and availability, Unmanned Aircraft Systems (UAS), including small UAS (SUAS), are becoming a valuable, cost effective tool for collecting aerial images for photogrammetric analysis. The metric accuracy of scale accident scene diagrams created from SUAS imagery has yet to be compared to conventional measurement methods, such as total station and laser measurement systems, which are widely used by public safety officials and private consultants.
For this study, two different SUAS were used to collect aerial imagery for photogrammetric processing using PhotoModeler software. A high-resolution consumer grade camera as well as a lower-resolution integrated camera was used as payload to determine the effect of camera resolution on the photogrammetric accuracy of two mock accident scenes. Using a Nikon total station as measurement control, SUAS photogrammetry from both cameras was compared using established targets. As a subjective comparison, the roadway layout was measured without targets and the resultant diagrams were superimposed over the total station control. Results are analyzed and presented in tables and diagrams. The results show the photogrammetric measurement of an accident scene from SUAS aerial imagery provides measurements with errors well below generally accepted ranges for accident reconstruction.